Citation
Alternate work schedules and their effect on productivity in the National Park Service

Material Information

Title:
Alternate work schedules and their effect on productivity in the National Park Service
Creator:
Danz, Harold P
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
x, 236 leaves : ill. ; 29 cm.

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of Public Administration)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public Administration
Committee Chair:
Gage, Robert W.
Committee Members:
deLeon, Linda
Overman, E. Samuel
Finkelmeier, Robert L.

Subjects

Subjects / Keywords:
Flextime -- United States ( lcsh )
Job satisfaction -- Testing -- United States ( lcsh )
Employees ( fast )
Flextime ( fast )
Job satisfaction -- Testing ( fast )
United States ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (D.P.A.)--University of Colorado at Denver, 1992. Public administration
Bibliography:
Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Public Administration.
General Note:
School of Public Affairs
Statement of Responsibility:
by Harold P. Danz.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
28163862 ( OCLC )
ocm28163862

Downloads

This item has the following downloads:


Full Text
ALTERNATE WORK SCHEDULES AND THEIR
EFFECT UPON PRODUCTIVITY IN THE
NATIONAL PARK SERVICE
by
Harold P. Danz
M.A., University of Northern Colorado, 1981
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Public Administration
1992


This thesis for the Doctor of Public Administration
degree by
Harold P, Danz
has been approved for the
Graduate School of Public Affairs
by
Linda deLeon
Robert L. Finkelmeier


Danz, Harold P. (D.P.A.)
Alternate Work Schedules and Their Effect Upon
Productivity in the National Park Service
Thesis directed by Associate Professor Robert W, Gage
The National Park Service is the largest bureau
office within the Department of interior. When Congress
permanently consented to the authority to approve
flexible and compressed schedules in 1982, each of the 10
Regional Offices in the National Park: Service developed a
formal or, in some instances, an informal policy
regarding use of alternate work schedules* These
policies ranged from very liberal to ultra conservative.
Although there have been several efforts to observe
and experiment with alternate work schedules to determine
organizational consequences, much of the subject
literature is anecdotal in nature. Those studies that
did make use of experimental and control groups primarily
reported mixed results or suggested that no measurable
effect resulted from the experiments.
Six experiments were conducted in the National Park
Service, and, in addition, a questionnaire was mailed to
433 randomly selected permanent employees. The
objectives of the experiments and questionnaire were


to measure employee productivity, performancef use of
leave, and level of satisfaction, and to determine if
there were significant differences between employees
working under a traditional work schedule and those
working under alternate work schedule(s).
Six scientific hypotheses were formulated around the
above research objectives, and through a series of tests
and analyses, it was determined through the use of
control and experimental groups that ¢1)employees
working under a compressed work schedule demonstrated
somewhat lower productivity levels during the pretest
posttest time periods, and (2) employees working under a
flexible work schedule demonstrated marginally higher
productivity levels. However, in each of the above
circumstances the difference was not statistically
significant. Use of unscheduled annual leave was greater
for both compressed and flexible work groups than for
those groups working under traditional work schedules.
Use of sick leave was marginally greater but did not
reach statistical significance for these groups*
Employee turnover in the National Park Service is
extremely low, and no measurable termination/hire effect
was associated with the availability of alternate work
schedules. Performance evaluation ratings and employee
iv


satisfaction levels were higher for both alternate work
schedule employees and traditional work schedule
employees. No significant statistical difference was
detected.
The form and content of this abstract are approved. I
recommend its publication.
Signed
Robert W, Gage
v


TABLE OF CONTENTS
Figures ix
Tables x
CHAPTER
I. INTRODUCTION .................................... 1
II. THE PROBLEM .................................... 14
Statement of the Problem....................... 14
Elements, Hypothesis, or Research Questions 17
Delimitations and Limitations of the Study 24
Limitations............................... 24
Delimitations .............................. 27
Definition of Terms............................ 28
Summary........................................ 36
III. REVIEW OF THE LITERATURE........................ 40
Theoretical Review ............................ 40
Historical Analysis............................ 45
Alternate Work Schedules Definitions
and Usage...................................... 52
Flexitime................................... 54
Compressed Schedule ........................ 61
Flexible Workplace Arrangements ............... 69
Productivity .................................. 74
Summary...................................... 93


IV. RESEARCH DESIGN................................ 97
General Structure and Methodology Framing 97
Specific Procedures ......................... 103
Research Population ......................... 112
Instrumentation.............................. 115
Data Collection.............................. 119
Treatment of the Data........................ 122
Summary...................................... 124
V. ANALYSIS AND RESULTS.......................... 126
Study Variables.............................. 127
The Antecent Variables ................. 127
The Intervening Variables ................ 129
The Independent Variables ................ 135
The Dependent Variables .................. 137
Initial Analysis ............................ 142
Results of Preliminary Analysis ............ 142
Comparability of Demographic Variables
for Experimental and Control Groups . 142
The Effect of Alternative Work Schedules
on Productivity........................... 148
Questionnaire/Survey Responses - 157
Hypothesis Testing .......................... 163
Hypothesis Testing,
for Hypotheses 1,2, 3, and 4............ 163
Hypothesis Testing,
for Hypotheses 5 and 6.................... 176
Summary of the Findings...................... 184
vii


VI. CONCLUSIONS AND SUMMARY............ 187
Work Schedule Productivity Results .... 189
Usage of Sick and Unscheduled Annual Leave 190
Employee Turnover in Employment .............. 192
Employee Performance Evaluation and Morale 192
APPENDIX
A. Reviewer^s Guide to Questionnaire .............. 200
B. Questionnaire......... 205
C. Questionnaire Significance Response Testing 212
D. Comparisons of Equality of Means
Experimental Groups ............................ 216
BIBLIOGRAPHY ........................................ 221
viii


FIGURES
Figure
4-1 Variable Elaboration Model..................105
5.1 Age Distribution Human Resource Groups,.. 144
5.2 Age Distribution Human Resource Groups
Compressed AWS Experiments..................145
5.3 Age Distribution MARO Human Resource Groups
Flexitime Work Schedule Experiments .......... 145
5.4 Age Distribution RMNP Maintenance Group
Compressed AWS Experiment .................... 146
5.5 Age and Productivity Correlation, Human
Resource Control Groups ...................... 147
5.6 Age Productivity Correlation, Human
Resource Experimental Groups ................. 148
5 7 Productivity Human Resource Groups .... 154
5.8 Productivity Compressed AWS
Human Resource Experiments ................... 155
5.9 Productivity MARO Flexitime
Experiment..................................156
5.10 MARO Productivity Pretest.................170
5.11 MARO Productivity Posttest................171
5.12 MARO Productivity Increase/Decreases , * 171
ix


2S
43
107
136
144
149
150
152
159
162
167
170
172
175
177
180
181
TABLES
Number of National Park Service Employees and
Authorized Compressed Work Schedules ....
Findings of Studies/Research Experiments,..
AWS Human Resource Distribution by Regions..
Compressed AWS by Regions and Offices
Age Comparisions for Experimental and
Control Groups F-Ratio Analysis ...........
Mean Productivity by Organization and
Experimental/Control Groups .................
Productivity and Performance Results for
Control and Experimental Group Employees ,.
T-Tests Difference in Means Between
Pretest and Posttest Scores .................
Response Rate of Employee Surveys
Applicable to the NPS .......................
Theme Responses to the Employee Questionnaire
Regarding Work Schedule Questions ...........
T-tests, Compressed AWS Productivity Pretest
and Posttest Means...........................
Age Productivity Correlation, MARO Human
Resources Flexitime Experiment...............
T-test, MARO Flexitime AWS Productivity
Pretest and Posttest Means...................
Use of Annual and Sick Leave.................
NPS Employee Terminationsf 1990 and 1991 * *
T-Tests, Performance Evaluations Pretest
and Posttest Means ..........................
T-Tests and ANOVA Independent
Group Analysis Summary ......................
x


CHAPTER I
INTRODUCTION
The history of civilization is closely joined with
those remarkable efforts of society over the years to
improve upon the methods by which human suirvival is
directed and organized. Although it may not be possible
to precisely distinguish all of the characteristics of
humans from other animals, it is generally accepted that
work, as a distinct activity, is more closely associated
with man than any other related mental or physical
endeavor of the animal species. Contrary to the more
common opinion held regarding the "primitive" human,
i.e., that where natural conditions permit, they work as
little as possible, the facts seem to suggest quite the
opposite (Herskovits 1962). Even under the most idyllic
of settings, mankind has not rested nor seemingly been
content with what nature has provided. Consequently,
organization of work quite possibly began during the very
earliest stages of human development in an effort to
augment the food supply through more regulated and
controlled methods of foraging, hunting, and later
agriculture.
The emergence of labor -labor as a quantity of
effort detached from an individuals life and bought on
the market in fixed amounts -- had a profound effect upon
what might be considered as a pure work concept or a work


ethic. The so-called "industrial revolution" did inspire
a demand for material and economic well-being, but it
also flourished at the cost of immediate human suffering
with the promise and potential for an eventual better
life for all people (Gladden 1912, 258).
In the human society, work has assumed a very
pervasive and powerful role. Once mankind worked only in
order to ensure survival, then work seemingly was cloaked
in a moral and righteous fabric; all who could work did,
because idleness was presumed to be bad; a person without
a job was a misfit (Weber 1958>, and association by
profession contributed to the establishment of craft and
trade unions. So intertwined has the work role become
with an individualpersonality, that it has been
observed that we attempt to place people by asking what
it is that they do rather than who his or her ancestors
were (Biesanz 1964). Organizational theory therefore
attempted to address concepts that expressed the idea
that (1)employees are people with a variety of needs
that must be satisfied if they are to lead full and
healthy lives and to perform effectively in the
workplace, and (2) the needs of individuals and
organizations should be integrated.
Max Weber credited the origin of the work ethic to
Martin Lutherfs interpretation of "calling. According
2


to Luther, by laboring in your calling you expressed
brotherly love through the services and products you
produced for society, Weber therefore concluded that
_T. the pursuit of wealth as an end in itself was
reprehensible; but the attainment of it as a fruit of
labour in a calling was a sign of God^s blessing....11
(Weber 1958,172). Literature now suggests that the
American work ethic is changing significantly. In Work,
Mobility and Participation (Cole 1979) postulates that
the American work ethic has been weakened by rising
levels of affluence and union control. David J.
Cherrington (1980), in reporting on a survey conducted in
1976, concludes that members of today's workforce,
especially young workers, do not have the same attitudes
as previous generations towards the importance of work,
pride, and craftsmanship.1 Another survey, conducted by *
Conducted under the Erteszek Working Paper Series,
Graduate School of Management, Brigham Young
University,191 questions were asfced about workers'
attitudes towards their jobs, their company, their
community, and work in general* Over 3,000
responses were received. The most desirable work
-related outcome was "feeling pride and
craftsmanship in your work." The second most
desirable was "getting more money or a larger pay
increase.11 Compared with older workers, younger
workers felt that pride in craftsmanship and joy in
being of service were less desirable, having leisure
and free time were more desirable, and doing a poor
job more acceptable. Leisure and free time were
less important to women and that females endorsed
pride and diligence slightly more than males.
3


the Gallup organization during September, 1979, seemed to
indicate that if an individual were given a choice to
work or not to work nearly 50 percent of those responding
would prefer not to work*2 Therefore, although this
paper does not deal with attitudes of people or
specifically address the American work ethic, it will be
necessary to consider these values in arriving at
conclusions concerning the application of alternative
work schedules.
Social research in the 1930s placed emphasis upon
the human relations approach to organization and
management, but another body of research begun in the I
2 Leonard A. Wood, Executive Vice President, the
Gallup Organization, Inc,, refers to this survey in
an article entitled "Changing Attitudes and the Work
Ethic The Work Ethic in Business, ed* W* Michael
Hoffman and Thomas J. Wiley. Cambridge:
Oelgeschlager, Gunn and Hain,1981 The question
was asked "If you had your choice, and money and/or
child care were not a problem, which of the
following would you prefer to do: *1 wouldn^ work,
I;d work part time, I^d work full time?11 Only one
man in 10, now currently employed^ chose not to work
at all Three in ten would switch to part-time
positions, while a majority (61 percent) would
continue working full-time if given the choice, A
somewhat different pattern emerged among women. For
women working full-timef part-time employment is
seen as the ideal-40 percent would choose working
on a part-time basis. Fourteen percent would prefer
not to work, and 42 percent would continue working
full-time it given a choice*
4


1940s and 1950s contributed a somewhat different
perspective of the workers place in the work situation.
This approach, known as organizational humanism/ formed a
bridge between the human relations approach and what is
now referred to as modern organizational theory. It was
postulated that, in addition to previous assumptions of
the human relations approach management can, and should,
promote positive motivation through delegation of
responsibility, permitting discretion and creativity on
the job*3 The report of a special task force to the
Secretary of Health, Education, and Welfare also
addresses the changing social character of today's
generation and concludes: ".*the very high personal and
social costs of unsatisfying work should be avoided
through the redesign of v/ork. {W.E. Upjohn Institute
1981,94).
Redesigning work to more closely fit the needs of
the worker significantly differs from job engineering, as
presented by Frederick W. Taylor, The job engineering
3 Much of this has been attributed to the work of Frederick
W, Taylor and his concepts regarding "scientific
management,11 i.e,, the one best way of accomplishing
work. Taylor's philosophy was, however, much deeper than
that. It was Taylor^s thought that the maximum good for
all society can only come about through the positive
cooperation of labor and management. In other words,
there must be harmonious collaboration in order to
achieve a common objective.


approach concentrates on maximizing the efficiency of
workers and quite often leads to the creation of
fractionalized, boring jobs- Redesign of work to provide
for job enrichment began in the 1940s at IBM (Walker
1950). Progress was initially rather slow, but more
attention and activity was eventually directed toward the
development of motivational factors in job assignments
when the intuitively appealing theories of Herzberg, et
al.{1959) and Maslow (1954) were published and began to
be generally accepted.
The average workweek for non-agriculture wage and
salary workers decreased from 65.7 hours per week in 1850
to 387 hours per week in 19874 There are many
reasons, that need not be discussed here, for this
substantive change, but it is quite clear that the
conmitments and obligations that evolve from social,
family, and recreational interests have tended to
increase while the formal working week was
correspondingly being reduced. Somewhat similar to work
redesign, in respect to achievement of employee morale
benefits, is work rescheduling. Use of flexible schedules
4 1850-1930 workweek statistics obtained from J. Frederick
Dewhurst and associates, America^ Needs and Resources;
A New Survey. New York: Twentieth Century Fund, 1955 -
More recent data was published in the Statistical
Abstract of the United States,1989. U.S. Department of
Coiomerce, Bureau of the Census.
6


permits employees to cope better with nonwork
impingements on work (e*g child-care, schooling needs,
commuting, personal fitness schedules, etc.). Job
sharing, flexitime, and currently, telecommuting, are
examples of such programs.
German economist Christel Kraemerer originated the
concept of flexitime in the 1960s as an inducement to
bring housewives and mothers into the work force to
alleviate the labor shortage in Germany. In 1967, the
West German aerospace firm of Messerschmidt-Bokow-Blohn,
introduced the concept of "gleitzert" or gliding time as
a means of reducing traffic congestion in the area near
their operations. It was based upon the assumption that,
in many work situations, rigid starting and stopping
times are unnecessary and could be adapted into a more
flexible system. Later the concept was modified the idea
of a core time was added, and the system was adopted by
other European countries. Although accurate data has not
been supplied, it has been estimated that about a third
of the Swiss labor force had been under flexitime, as
well as substantial numbers of white-collar and other
workers in Germany, England, France and Scandinavia* By
1972, flexitime was being used in the United States at
Control Data Corporation, and at other firms on a trial
basis.
7


Information with respect to the proclaimed
advantages of imaginative job design proposals such as
staggered hoursf flexitour, flexitime and group
flexibility (generically referred to as flexitime,
compressed work week and its variations) quickly spread
and numerous experiments and adaptations of alternative
work schedules occured. Articles in professional
journals and other publications described the basic
principles of the system and announced in enthusiastic
terras the varied results, but more often the successes,
of the alternative work schedule experiments* Many of
the early flexitime examinations were primarily
descriptive in nature, or lacked the scientific rigor
desired. Four studies did investigate flexitime in an
experimental fashion:(1} Evans and Partridge (1973), who
examined the impact of flexitime on non-supervisory
employees in a large insurance company over a five month
period, (2) Golembiewski, et al.(1975 who examined,
over an extended period, pretest and posttest data
results from an experiment at a research organization,
(3) Schein, et al. <1977), who focused on the impact of
flexitime on productivity, and Harrick, et al.(1986),
who conducted field studies on the effect of alternate
work schedules on productivityy leave usage, and employee
attitudes. Control groups were not utilized, however.
8


The Congress of the United States found that new
trends in the usage of the four-day workweeks, flexible
work hours, and other variations in workday and workweek
schedules in the private sector seemed to show sufficient
promise to warrant carefully designed, controlled,
and evaluated experimentation by Federal agencies over a
3-year period to determine whether and in what situations
such varied work schedules can be successfully used by
Federal agencies on a permanent basis_"(1978 92 Stat.
755) The Public Law that authorized this experiment was
to last until September 29,1981. The Law was later
amended to extend it 1111til July 4,1982, and it was again
extended for three more years, until July 23,1985.
Permanent authority was subsequently granted on December
23,1985 (99 Stat.1350).
Based upon responses to a General Accounting Office
questionnaire (U* S. Comptroller General B-2160211985),
an estimated 489,000 permanent Federal employees were
working under an alternate work schedule arrangement in
1985. The General Accounting Office now estimates that
37 percent of all federal employees are currently working
under some kind of alternate work schedule/ or
approximately 1,145,000 employees* Although initial
reaction to the adoption of alternate work schedules was
positive with respect to employee morale, mixed reactions
9


were received by the General Accounting Office regarding:
(1)service to the public, < 2) efficiency of government
operations, (3) mass transit, (4) energy consumption, ¢5)
employment opportunities, and (6) work productivity.
The National Park Service (NPS) is a bureau within
the Department of Interior. In 1989 it had 12,737
permanent employees, which made it Interior's largest
bureau from the standpoint of employment. The NPS
benefits from a large and active constituency; its
supporters have pressed for independent agency status,
and the NPS is consistently viewed more favorably by more
Americans than any other Federal agency. 5 This
popularity is further supported by the growth of public
visitation to the National Parks. According to the
National Park Service, total visitation grew from 451,000
in 1918 (Albright 1985,103), to 351,911,180 in 1989.6
Because of its long history of providing quality
public service, its highly seasonal operation, and
variety of occupational specialties, the application and
5 The Roper Poll, interpreted by Adweek#s Marketing Week in
its January 8,1990 issue, shows the National Park
Service with an 82 percent favorable rating, ahead of
second place FBI at 77 percent/ and NASA at 76 percent-
For at least the 10th straight year, this poll has placed
the National Park Service as the top agency in
Government.
6 U.S. Department of the Interior, National Park Service,
Division of Statistics*
10


approval of alternate work schedules in the NPS was
approached rather cautiously. Legislation permitting
Federal agencies to experiment with alternate work
schedules was enacted on September 29,1978, but it was
not until October 23,1979, that the NPS distributed any
information regarding the availability of these
experimental work schedule programs. Initial instructions
to supervisors suggested that only a very limited number
of 11 experimental11 programs should be initiated, and that
those employees whose work involved program management
and direct public contact work could not be included.
Each Regional Director was delegated the authority to
establish alternate schedules for employees, but they
were to be terminated if there were any evidence of
reduced productivity, a lowered level of public service,
or an increase in agency costs.
Informal surveys, conducted approximately six months
after implementation, seemed to indicate that those
employees who were under an experimental work schedule
were well satisfied with the arrangement; but those
employees who were working under a traditional schedule
expressed considerable reservations over the use of the
alternate work schedules. Adverse comments, primarily
from managers and supervisors, contradicted favorable
comments, i,e., energy consumption has increased,
11


productivity has declined; carpools have been disrupted,
etc. Since the earlier informal survey, no further
review of this program has taken place.
Another alternate work schedule, or arrangement, is
that of the flexible workplace. Referred to as
"telecommuting or "flexiplace, this prograin officially
began implementation on February 12,1990, with the first
workshop for the agency coordinators, Telecoiaiauting
creates an "office without walls,11 allowing people to
work at home or at satellite offices, to commute" using
computers and telephone lines instead of cars. One of
the obvious challenges of work at home is merging work
and home life successfully.
At this time 31 agencies and bureaus have taken the
first steps toward establishing work-at-home programs by
designating coordinators to direct flexiplace
arrangements within their organizations. Guidelines have
been produced, and it is anticipated that further
procedural refinement will occur as the program evolves-
The Department of Interior has indicated that they will
support only 50 experimental flexiplace arrangements, and
these will be only in the headquarters office (with a
limit of five employees from each bureau). One of the
more pervasive concerns is the issue of productivity, or
the potential lack thereof, that is making agency
12


managers quite skeptical- A lesson can be learned from
earlier experiences in adopting flexitime; flexiplace
(telecommuting) should not be implemented without careful
thought about its impact on operations and/or agency
mission*
Although other factors may influence the success or
failure of an alternate workplace arrangement, there is
no question that public sector productivity is a
continuing strategic issue and concern. It would
therefore be appropriate to question whether any employee
benefit program should be introduced if its subsequent
implementation would have a deleterious effect upon
employee productivity, effectiveness, or efficiency.
Alternate work schedules are being increasingly accepted
and adopted almost as if they were an extension of an
employee rights, and they currently are part of union
contract negotiations as other agreed upon conditions and
circumstances of employment. In July 1985, President
Reagan asked Congress to pass a joint resolution
establishing productivity as a national goal and to enact
management legislation to improve government management
tools. Can we identify the flexible work schedule as
being one of those tools?
13


CHAPTER II
THE PROBLEM
Statement of the Problem
The purpose of this study is to investigate and
evaluate the effect and impact of the use of alternate
work schedules upon employee productivity in the National
Park Service, a bureau of the Department of the Interior.
At the present time, alternate work schedules within
the National Park Service are being approved without
sufficient research or investigation as to the effect
that approval of such schedules would have upon the
specific work activity involved. Based upon certain
encouragement from the Director's office on October 12,
1984/ and as reinforced in the Alternative Work Schedules
Handbook of 1985, use of alternative work schedules is
supported within the National Park Service as long as "
approved schedules do not interfere with the mission
of the organization...,1,1 In addition to its broad
mission and purpose 11... to conserve the scenery and the
Memorandum from the Director of the National Park Service
to the Directorate, Field Directorate, WASO Office and
Division Chiefs, and all Superintendents, October 12,
1984. Draft of Alternate Work Schedules Handbook, from
Assistant Director, Personnel and Administrative
Services, March 18,1985. p*8 *


natural and historic objects therein,, *t,2 the National
Park Service, along with all other Federal agenciesf is
subject to Executive Order 12552, February 25f 1986,
which seeks a 20 percent productivity increase through
higher quality, improved timeliness, and lower
cost
Most National Park Service employees prefer
alternative work schedules (Johnson et al.1984; Johnson
and Salvi 1985). 2 3 4 An alternative work schedule is any
work schedule which differs from what would be considered
to be the normal, accepted, customary, work schedule for
the organization. If the normal work schedule or tour
required an employee to work onsite 8:00 AM to 5:00 PM,
Monday through Friday, any variance of this work schedule
would be considered an alternative work schedule. Thus,
compressed work weeksf flexitime, part-time employment,
shift work, or any other work arrangement that differs
from that of the normal schedule work be considered an
2 An Act To establish a National Park Service, and for
other purposes approved August 25,1916. 39 Stat. 535 *
U.S. Code, Title 16.
3 Memorandum from President Reagan to heads of executive
departments and agencies, July 31,1985.
4 In these two studies, 71% of all permanent, and 58% of
all temporary employees indicated that they worked a
traditional 40 hourr five day week, but, only 36% of the
permanent and 43% of the temporary employees actually
preferred the traditional schedule.
15


alternative work: schedule- A variety of flexible and
coinpressed schedules are found in all National Park
Service management offices and most park areas; no record
of the number of such schedules approved are maintained,
and, in most instances, no effort to record work time
performed outside of the normal work hours is attempted.
Although supervisors have the authority to disapprove
requests for alternative work schedules if it appears
that such approval would adversely impact public service
or productivityf it is questionable if they would do so
because of employee pressure and lack of management
support. No effort is made in the National Park Service
to routinely measure productivity, and therefore, the
subsequent effect of alternate work schedules upon
productivity is largely unknown.
Significant differences exist between regions,
oftices, and park areas, on the approach to approval of
compressed work schedules. Although nearly all
organizational units permit some form of flexible work
schedules, i*e*, varying the beginning and ending work
hours and adhering to accepted "core11 work hours,
coinpressed work hour schedules have not been universally
accepted within the National Park Service. The principal
organizational units within the National Park Service are
the individual park areas themselvesf the 10 regional
16


offices under which all park areas are located, the
Denver Service Center {construction, design, and planning
office), the Harpers Ferry Interpretive Planning Center,
and the bureau headquarters office in Washington D.C,.
From a total payroll of 17,923 National Park Service
employees on April 25,1990,1,725 were working under a
compressed work schedule, and this varied by geographic
regions from a low of 10 (Mid-Atlantic Region,1,498
employees), to a high of 608 (Western Region, 2,777
employees). Thus, it would appear that management
application and/or acceptance of a compressed work
schedule is not consistent from bureau to bureau.
Interest in a flexiplace arrangement has been
expressed by quite a few employees, but National Park
Service management has not as yet addressed this issue.
Hypotheses, or Research Questions
A number of hypotheses can be formulated with
respect to research design and the problem statement.
Hypotheses will be tested through field study, but it is
perhaps important to state at this point that although
much of the research will be ex post factor certain
elements will be subject to direct intervention and/or
manipulation because of the organizational position and
status of the researcher
17


The basic thrust of the first hypothesis will be to
determine whether or not compressed work schedules have
any effect upon productivity in the National Park
Service. Previous research findings and theories suggest
that compressed work schedules have but little effect on
productivity. Pretest-posttest comparisons involving
experimental and control groups will be structured to
respond to the following theorems:
Null: There is no significant difference

,05)
in the levels of productivity between
experimental (compressed work schedules)
and control groups for the pretest-posttest
time periods.
Productivity measurements for experimental
{compressed work hour employees) and
control groups indicate that employees not
under a compressed work schedule (control
groups) demonstrate, on an average, higher
productive levels during pretestposttest
time periods.
A second hypothesis is associated with the use of
alternative eight-hour work schedules. These are fixed
schedules which do not vary from day to day. The arrival
and departure tiroes are according to a set schedule known
to both the employer and the supervisor in advance. The
schedule includes ten (10) work days in each pay period*
Each work day is eight (8) hours in length excluding the
scheduled lunch period. This schedule differs from the
normal eight-hour schedule in that the scheduled arrival
Research;
HI
18


and departure times need not coincide with the
traditional eight-hour schedule and the employee may
schedule a lunch period longer than the customary or
minimum lunch period.
Null: There is no significant difference (p >.05)
in the levels of productivity between
experimental (alternative eight-hour work
schedules) and control groups for the
pretest-posttest time periods.
Productivity measurements for experimental
(flexible work schedule employees) and
control groups indicate that employees not
under a compressed work schedule (control
groups) demonstrate, on an average, higher
productive levels during pretest-posttest
time periods.
Each of the above hypotheses would be tested for
both immediate and longterm comparison and effect- The
logic of separate testing is to determine whether or not
productivity enhancements, or losses, have been
influenced by employee temporal response to change
(Hawthorne)*
It has been frequently suqqested that application of
flexible schedules reduces absenteeism, which would
adversely affect productivity (Rosow 1981; Craddock,
Lewis and Rose 1981; Coltrin and Barendse 1981).In the
Federal Government this could be equally applicable to
unscheduled and brief use of annual leave as well as
employee use of sick leave. With the exception of the
work of Harrick, et al(1986), involving an omce of a
Research:
H2
19


federal government agency, this supposition has not been
fully tested as yet with respect to federal employees,
and, more specifically, with National Park Service
employees;therefore, a third and a fourth hypothesis
would involve the use of annual leave for brief but
unscheduled periods, and the use of sick leave*
Null: There is no significant difference (p >.05)
in the use of annual leave by employees for
brief unscheduled periods between
experimental and control groups for the
pretest-posttest time period.
Research: Use of annual leave for brief unscheduled
periods is, on an average, less for
employees in experimental groups than for
those in control groups during the pretest
-posttest time period.
There is no significant difference (p >.05)
in the use of sick leave by employees
between experimental and control groups for
the pretest-posttest time period.
Research: Use of sick leave for brief unscheduled
periods is, on an average, less for
H4 employees in experimental groups than for
those in control groups during the pretest-
posttest time period.
As was the case with the testing of the first two
hypotheses, a long posttest comparison will also be
implemented to determine more precisely if there is a
relationship between the use of leave by employees and
alternative work schedules. Since it can be argued that
compressed work schedules may have a more deleterious
effect upon absenteeism, alternative eight-hour work
H3
Null:
20


schedules and compressed schedules will be tested
separately*
Another human factor which can contribute to
productivity decline is high employee turnover. It has
been generally claimed that since flexitime generally
increases employee satisfaction, turnover is typically
lower (Coltrin and Barendse 1981; Welch and Gordon 1980;
Rubin 1979; U. S, Office of Management and Budget 1973).
In response to a specific question in The National Park
Service Employee Survey of 1983, regarding their leaving
the National Park Service, 937 employees (n=3452)
indicated that they would look for a new job outside of
the National Park Service during the next year. Lack of
career advancement opportunities and training were the
most frequently mentioned reasons. No measurable answers
in excess of 1 percent of the total responding were
attributable to work schedules. This would seem to
suggest that work schedules are not a significant reason
that would necessarily influence employee turnover* A
fifth hypothesis would therefore relate to actual
turnover, and whether or not the ability to utilize a
flexible work schedule reduced the turnover rate.
Comparisons would be made between regions and/or offices
and parks that maintained either conservative or liberal
21


policies with respect to the approval of flexible work
schedules.
Null: There is no significant difference (p >,05)
with respect to the rate of employee
turnover between offices and/or parks who
offer liberal work schedule opportunities
(experimental),and those who offer
primarily only traditional work schedules
(control group)*
Research: Employment turnover is lower with respect
to offices and/or parks who offer liberal
H5 work schedule opportunities (experimental),
than it is for offices and/or parks who
offer primarily only traditional work
schedules (control group).
A sixth and final hypothesis that would be tested is
related to employee morale and performance. Literature
dealing with alternative work schedules and productivity
is near-unanimous in its association of morale
improvement and the adoption of alternative work
schedules. A questionnaire using a matrix question
format and Likert response categories would be
distributed to a random sampling (sufficient to achieve
95 percent accuracy) of all employees previously
identified as working under an alternative work schedule*
The questionnaire would be also randomly distributed to
other employees not working under an alternate schedule,
in addition, a sampling of the annual performance
evaluation rating provided to those employees working
under an alternate work schedule would be compared to a
22


proportionate sampling of those employees not working
under an alternate work schedule. The same level of
certainty (95 percent) would dictate the sample size.
Null: There is no significant difference (p >.05)
between the level of employee work
satisfaction (morale) and his/her
individual performance evaluation for those
employees working under an alternative work
schedule (experimental group), and those
who work under more traditional work
schedules (control group).
Employees who work under an alternative
work schedule {experimental group) tend to
have higher morale, and, on an average,
higher performance evaluations than those
employees who work under more traditional
work schedules (control group).
Delimitations and Limitations of the Study
Much of the information that will be assembled is
subject to those limitations inherent in ex post facto
field studies. The independent variable, flexitime, was
not assigned by the researcher to an individual in a
random fashion, nor were the selection of work hours* In
some instances the individual may have requested a
specific compressed work schedule and was instead
assigned by his/her organization to another but more
limited work schedule. The same circumstance exists with
respect to all control group work schedules. These were
also selected by the respective organizational management
officials, as were those individuals who were considered
Research:
H6
23


by management to be not eligible, by virtue of their
position or status, for a flexible work schedule* In the
course of the research effort, certain delimiting
factors, such as sample size or regional selection, may
also exist and require fuller examination and discussion
in the context of their potential effect upon the study
and eventual research conclusions.
Limitations
Each National Park Service organization unit has its
own distinct culture and personality* Geographic and
ecological differences can explain some of the
variations, but the probable principal causes are
staffing requirements and needs that are influenced by
organizational purpose and mission, functional favoritism
(i*e,, management preference and priority to ranger
activities and requirements over those related to
maintenance, research, or administration), and
leadership. Because of these cultural differences,
effective leaders in one park area do not necessarily
become effective leaders in another park area and/or
office. This cultural difference is partially identified
in Table 1 below where approval and use of other than
flexible or traditional work schedules (compressed 4/10
24


8.43
.07
10.54
6.75
7.16
26.14
5.42
3.22
4.79
21.89
17.19
1.20
15.52
Alaska 415
Mid-Atlantic 1,498
Midwest 1,158
North-Atlantic 1,407
National Capital 2,554
Pacific Northwest 704
Rocky Mountain 2,065
Southeast 2,421
Southwest 1,210
Western 2,777
Denver SC 576
Harpers Ferry 249
Washington Office 889
and 5/4/9) are dependent upon established regional and
park policies.
Table 2.1
Number of National Park Service Employees
and Authorized Compressed Work Schedules
As of April 25,1990
Region/Office No. of Authorized
Employees Compressed W/S
Percent
Authorized
17,923 1,725 9.62
Another limiting factor will be the amount of
reliance that will of necessity be placed upon secondary
data sources. For example, the information supplied in
Table 2.1, was acquired from Department of Interior
personne1/payro11 records, which are the result of
reporting by a variety of National Park Service personnel
offices. In the event that an office elects not to
report the approval of an alternative work schedule, then
35)02295a3fi4t27a58>08993Js
6 £
25


that particular schedule would not be statistically
represented. It is proposed that selective testing be
applied to assure validity and reliability to bureauwide
reported figures.
Because the author does not have access to all of
the NPS areas and offices, the sampling frame for this
study is limited and therefore a simple random sample of
the target population cannot be drawn. This would
seemingly also be true with respect to the personal
characteristics of target groups, as differences in race
and sex become more apparent between rural and
metropolitan park areas and offices. The possibility of
stratified sampling, where and when available, will be
used to reduce the aspect of sampling error. Since no
previous contacts with National Park Service employees
have occurred, with the exception of the National Park
Service Employee Survey of 1983, it is felt that no other
known factors which may affect generalization and/or
external validity should bear upon, or significantly
influence, this research-
Limitations on theoretical relationships between
worker productivity, sick leave usage, unprogrammed
absences, turnover, and performance evaluations also
exist Although there are limited opportunities to
measure productivity from a quantitative perspective,
26


most productivity measurements are qualitative in nature
and/ thus subject to a certain amount of ambiguity. Of
necessity, considerable reliance has been placed upon
bureau operational efficiency and quality of public
service. Evaluation of either organizational
characteristic is, however, rather subjective. Previous
literature does suggest that relationships do exist
between flexible work schedulesr reductions in
absenteeism, turnover, improved performance and employee
morale, as well as with other variables, but the extent
of relationship and/or the possibility of another
unrelated factor coexisting, and perhaps influencing a
positive/negative relationship has not been conclusively
determined.
Delimitations
The researcher is the Associate Regional Director
for Administration in one of the major National Park
Service regional offices- Quite obviously, considerable
information will be readily available with respect to
this region and the parks within this region* In
addition, opportunities for intervention are inimitably
better if implemented within the home region, which may
potentially present a selection bias.
It has been earlier suggested that National Park
Service organizational culture has a decided influence
27


over the approval, establishment of policies applicable
to, and effective encouragement and perhaps ultimate
success of, alternative work schedules. Although many of
the more traditional National Park Service managers
profess their support of the AWS program, and also
acknowledge the existence of forceful guidance by the
Department and OPM in this regard, an unspoken but
implied message is received by the lower supervisory
levels in the organization that they are expected to take
a conservative approach in their implementation of this
particular program. However, even without this tacit
message, the Scime bias appears quite evident within most,
if not all, supervisory levels with respect to the use of
compressed work schedules. The sole exception would be
where direct program advantage or benefit obviously
occurs. An example of this would be in the case of field
trail maintenance people who spend eight 10-hour days in
a row in the backcountry of a park area and then have six
days off duty. This results in reduced travel time, less
associated perdiem costs, and improved work programming.
Definition of Terms
A variety of words or expressions is being used in a
precise sense in the proposalf since the proposed
research depends upon specific definition of terms, the
28


following definitions, and definitional sources if
available, are being provided:
alternative work schedule, an alternative to the standard
5-day 40 hour workweek, a rearrangement of workhours.
Janice Neipert Hedges- Special Flexitime Reports,1977.
A substitute to the standard five-day, 40-hour workweek.
May be either a change in the number of days worked, a
change in the number of hours worked^ or a change in both
days and hours John W* Newstom and Jon L* Pierce,
Alternative Work Schedules: The State of the Art, 1979,
bandwidth, total number of hours in the interval between
the earliest possible workhour starting time and the
latest finishing time. (All other things being equal, the
larger the bandwidth, the greater the possible
flexibility to the employee) Robert T. Golembiewski and
Carl W, Proehl,Jr., A Survey of the Empirical
Literature on Flexible Workhours: Character and
Consequences of a Major Innovation. 1978.
basic workweek, number of hours in a normal workweek, as
established by collective bargaining agreements or
statutory law. Premium payments must usually be paid for
time worked in excess of the basic workweek. Jay M.
Shafritz, Dictionary of Personnel Management and Labor
Relations, 1980. Section 6101 of title 5, United States
Code requires each agency head to establish a basic
administrative workweek of 40 hours.
climate, a set of properties in the work environment
specific to a particular organization, that may be
assessed by the way the organization deals with its
employees and its societal and task environments. Andrew
D, Szilagyi and Marc J. Wallace, Organizational Behavior
and Performance,1987*
compressed schedule/ in the case of a full-time employee,
an 80-hour biweekly basic work requirement which is
scheduled for less than 10 workdays; and in the case of a
part-time employee, a biweekly basic work requirement of
less than 80 hours which is scheduled for less than 10
workdays. P,L. 95-390, 9/29/78.
core time, designated hours during which an employee must
be present for work- P*L. 97-221, 7/23/82.
29


credit hours, any hours within a flexible schedule which
are in excess of the number of hours in an employee's
basic work requirement and which the employee elects to
work so as to vary the length of a workweek or workday.
P.L, 97-221, 7/23/82.
culture, a pattern of basic assumptions invented,
discovered, or developed by a given group as it learns to
cope with its problems of external adaption and internal
integration that has worked well enough to be
considered valid and, therefore, to be taught to new
members as the correct way to perceive, think, and feel
in relation to those problems. Edgar H. Schein,
Organizational Culture and Leadership, 1985.
effectiveness, the extent to which an organization
accomplishes some predetermined goal or objective; the
overall performance of an organization from the viewpoint
of some strategic constituency. Jay M. Shafritz, Facts
on File Dictionary, 1986.
efficiency, the promotion of administrative methods that
will produce the largest store of results for a given
objective at the least cost; the reduction of material
and personnel costs while maximizing precision, speed,
and simplicity in administration Jay M. Shafritz,
Facts on File Dictionary, 1986.
5/4/9, term given to a compressed work schedule which
calls for nine work days, rather than ten, within a given
two-week work period. The five work days of the first
work week are nine hours each, as are three days of the
second work weekr and the fourth day is of eight hours
duration*
flexible work hours, designated hours during which an
employee may elect the time of arrival at and departure
from work, for the purpose of varying arrival and
departure times, or if and to the extent permitted, for
the purposes of accumulating credit hours. P.L. 97-221,
7/23/82*
flexiplace, an arrangement which permits employees to
work at home or at other approved sites away from the
office for all or part of the work week. Flexiplace,
flexible work place, work-at-home, telecommuting, and
teleworking all refer to paid employment away from the
traditional office. Presidents Council on Management
30


Iiaprovement. Guidelines for Pilot Flexible Workplace
Arrangements,1990.
flexitime/ a way of scheduling work so that an employee
must be present during certain "core" hours of the day,
but can begin and end work any time before and after
these core hours (some organizations may establish limits
as to when flexitime hours may begin or must end; see
bandwidth). Don Hellriegel and John W. Slocum, Jr.9
Organizational Behavior, 1979.
flexitour, a flexitime work schedule which requires
employees to choose, for a relatively long period of
time, starting and quitting times for a standard eight-
hour shift, Jerome M. Rosow and Robert Zager, New Work
Schedules for a Changing Society, 1981. Employee will
preselect starting time; may modify schedule with prior
notification and approval of supervisor* P.L. 99-196f
12/23/85.
four/ten and/or four^day workweek, a regular full-time
workweek of 40 hours is worked in four 10-hour days
instead of five eight-hour days. Stanley D. Nollen,
What is happening to flexitime, flexitour, gliding tine,
the variable day? And permanent part-time employraent?
And the four-day week? 1980 *
gliding rime, a flexitime work schedule which allows
daily variations in starting and quitting times but keeps
a rigid eight-hour shift. Jerome M. Rosow and Robert
Zager, New Work Schedules for a Changing Society.1981.
job design, also called job redesign. Organizing work in
a manner consistent with societal goals, demonstrating
that the social and psychological basis of work is as
significant to long-term productivity and efficiency as
are the traditional physiological factors. Jay M,
Shafritz, Dictionary of Personnel Management and Labor
Relations,1980. A systematic process of determining, by
observations or study, the major work-related behaviorsf
responsibilities, skills, and experiences necessary to
satisfactorily perform a task. Gilbert B. Siegel and
Robert C, Myrtle, Public Personnel Administration> 1985
maxiflex, a flexitime work schedule which allows daily
variations with no core hours required, making it easy to
compress the workweek into three or four days. Jerome M.
Rosow and Robert Zager, Hew Work Schedules for a Changing
Society, 1981* For Federal Government employees, an
31


employee may vary the length of the work week and workday
as long as he or she is present for core time which is
scheduled on less than all five weekdays; must work or
account for the basic work requirement, e.g* 80 hours in
a bi-weekly pay period;credit hour carry-over is limited
to a maximum of 10 hours. P.L. 99-196f 12/23/85.
moraler moral or mental condition with respect to
courage, discipline, confidence, enthusiasm, willingness
to endure hardship, etc., within a group, in relation to
a group, or within an individual. Webster^s New World
Dictionary. Third College Edition, 1988. Collective
attitude of the workforce toward their work environment
and a crude measure of the organizational climate. Jay
M, Shafritz, Dictionary of Personnel Management and Labor
Relations/ 1980.
motivation, an emotional stimulus that causes a person to
act, an amalgam of all the factors in one's working
environment that foster (positively or negatively)
productive efforts. Jay M. Shafritz, Dictionary of
Personnel Management and Labor Relations, 1980. A
predisposition to act in a specific goal-directed manner.
Don Hellriegel and John W, Slocum, Jr., Organizational
Behavior, 1979.
organizationr any unified, consolidated group of
elements; systematized whole; especially, a) a body of
persons organized for a specific purpose, as a club,
union, or society, b) the administrative personnel or
executive structure of a business, c) all the
functionaries, committees, etc of a political party.
Webster's New World Dictionary, Third College Edition,
1988. Any structure and process of allocating jobs so
that common objectives may be achieved. Jay M, Shafritz,
Facts on File Dictionary, 1985. A system of consciously
coordinated personal activities or forces of two or more
persons. Chester I. Barnard, The Functions of the
Executive. 1938.
organizational humanism, a set of organizational theories
stressing that work held intrinsic interest for the
worker, that workers sought satisfaction in their work,
that they wanted to work rather than avoid it, and that
they could be motivated through systems of positive
incentives (such as participation in decision making).
George J. Gordon, Public Administration in America.1982,
32


performaoce, the act of performing; execution,
accomplishment, fulfillment/ etc,, operation or
functioning, usually with regard to effectiveness, as of
a machine, something done or performed; deed or feat.
Webster^ New World Dictionary, Third College Edition.
1988. Demonstration of a skill or competence. Jay M,
Shafritz, Dictionary of Personnel Management and Labor
Relations,1980
pretest-posttestt an evaluation of circumstances,
capabilities or conditions (dependent variable) prior to,
and after an event (independent variable), for the
purposes of validating the intervention and measuring
effect on the dependent variable. Earl Babbie, The
Practice of Social Research, 1983*
productivity9 the efficiency with which resources are
used to produce a government service or product at
specified levels of quality and timeliness. E.O.12552,
2/25/86.
qualitative, research data framed in the form of words
rather than numbers, collected in a variety of ways
(observation, interviews, extracts from documents, tape
recordings) and are usually "processed11 somewhat before
they are ready for use. Matthew B. Miles and A. Michael
Huberman, Quality Data Analysis^ 1985. Research
techniques linked to participant observation, content
analysis, formal and informal interviewing, clinical case
studies, life-history construction, videotaping
behavioral displays, archival data surveys, historical
analysis. The invention and use of varied unobtrusive
measures, and various formal schools of thought and
procedure in social science such as dramaturgic analysis,
semiotics, frame analysis, ethno-methodology, and
conventional analysis, John Van Maanen, Quality
Methodology. 1983.
quality-of-work-life f a series of organizational
interventions, activities, designed to improve the work
place for employees. Andrew D. Szilagle and Marc J.
Wallace, Organizational Behavior and Performance,1987.
quantitative, having to do with quantity, capable of
being measured, WebsterNew World Dictionary, Third
College Edition, 1988* The numerical representation and
manipulation of observations for the purpose of
describing and explaining the phenomena that those
33


observations reflect, Earl Babbie, The Practice of
Social Research, 1983.
relationship, the quality or state of being related;
connection Websterfs Hew World Dictionary, Third
College Edition, 1988. Interrelationships among several
persons considered simultaneously. Earl Babbie, The
Practice of Social Research, 1983 *
reliability, that quality of measurement method that
suggests that the same data would have been collected
each time in repeated observations of the same
phenomenon. Earl Babbie, The Practice of Social
Research, 1983.
research, a careful,systematic, patient study and
investigation in some field of knowledgef undertaken to
discover or establish facts or principles. Websterfs New
World Dictionary, Third College Edition, 1988. A
systematic and intensive study for a fuller knowledge of
the subject studied. Basic research attempts to uncover
new scientific knowledge and understanding with little
regard as to when, or specifically how, the new facts
will be used. Applied research is conducted with a
special purpose in mind; it is directed toward a specific
problem, or toward a series of problems, that stand in
the way of progress in a particular area. Encyclopedic
Dictionary of Systems and Procedures,1966,
satisfaction, anything that brings gratification,
pleasuref or contentment. Websterfs Wew World
Dictionary, Third College Edition, 1988. People's
feelings about the rewards they have received, a
consequence of past events, the willingness of
individuals to continue as employees of an organization
and to show up for work on a regular basis, not
necessarily associated with motivation or increases in
work productivity, Edward E, Lawler, III, High
involvement Management.1986_ Worker contentment brought
about by the presence of conditions that meet certain
physical and psychological needs. Abraham H. Maslow,
Motivation and Personality,1954
scientific management, closed model theory, improvements
to organizational efficiency and economy for the sake of
increased production; time-motion studies; man-as-machine
conception, primarily associated with the work of
34


Frederick W* Taylor. Nicholas Henry, Public
Administration and Public Affairs, 1980*
secondary data sourcesr information collected by others
and archived in some form. The term secondary
information is frequently used to refer to secondary data
(the raw data obtained in various studies) and secondary
sources (the published srnnmaries of these data). David
W, Stewart, Secondary Research, 1984.
significant difference, through statistical testing it
can be determined that the size of differences found
between groups are not due to chance. The probability of
rejecting a null hypothesis when it is true. By
establishing a significance level of 01,only outcomes
that have a probability equal to or less than that will
allow a rejection of the null hypothesis. Susan Welch
and John C. Comer, Quantitative Methods for Public
Administration, 1983 In the context of tests of
statistical significance, the degree of likelihood than
an observed, empirical relationship could be attributable
to sampling error. A relationship is significant at the
05 level if the likelihood of its being only a function
of sampling error is no greater than 5 out of 100. Earl
Babbie, The Practice of Social Research, 1983.
telecommuting, implied the use of high-tech
telecommunications and computers to perform work from
remote locations. Work-at^home, flexiplace, covers work
regardless of high-tech or low-tech applications.
Presidents Council on Management Improvement*
Guidelines for Pilot Flexible Workplace Arrangements,
1990.
turnover, movement of individuals into, through, and out
of an organization. The total number (or percentage) of
separations that occurs over a given time period- Jay M.
Shafritz, Dictionary of Personnel Management and Labor
Relations,1980.
variable day, a flexitime work schedule which allows an
employee to work a longer or shorter time on any given
day as long as the total hours worked adds up correctly
at the end of a week, bi-weekly, or monthly period
dependent on agreed upon arrangements* Jerome M. Rosow
and Robert Zager, New Work Schedules for a Changing
Society, 1981. Employee may vary the length of the
workday as long as he or she is present for daily core
time with limits established by the organization; must
35


work or account for the basic work requirement, e.g. 40
hours per week for a full time employee; credit hour
carry-over between pay periods is limited to a maximum of
10 hours. P*L, 99-196, 12/23/85.
work scheduler a work arrangement which identifies the
hours of work that an employee will perform within a
given pay period, subject to not exceeding 80 hours in
the designated two week pay period. Section 6101,5 USC.
Summary
Research on the effect of alternative work schedules
on individual performance or productivity in the Federal
Government has, to this point, been somewhat limited.5
It is also noted that much of this research is based upon
soft information and/or is restricted in certain
scientific approaches, including pretest and posttest
evaluations- Those studies that did offer suitable
D, Hauser, The Impact of Flexitime on Organizational
Functioning: A Field Study. Washington, D.C.: U.S.
Office of Personnel Managementf June, 1980- C, Orpen,
f,Effects of flexible working hours on employee
satisfaction and performance: A field experiment,11
Journal of Applied Psychology, 66,(1981):113-115 R.T,
Golembiewski and C. W. Proehl,Jr "Public Sector
Applications of Flexible Workhours: A Review of Available
Experience1 Public Administration Review, 40, n.1,
(January/February 1900):72-85. Oscar Mueller and Muriel
Cole1Concept wins converts at Federal agency," Monthly
Labor Review,100, n* 2, (February, 1977):71-74. John
J. Macut, Measuring Productivity Under a 4-Day Week,n
Monthly Labor Review, 97, n. 4, (April, 1974): 55-56.
Edward J, Harrick, Gene R. Vanek, and Joseph F,
Michlitsch, uAlternate Work Schedules, Productivity,
Leave Usage, and Employee Attitudes: A Field Study,"
Public Personnel Management.15, n. 2, (Summer, 1986):
159-169.
36


levels of result validity, with the possible exception of
the Harrick, Vanelc, and Michlitsch <1986) study, suffered
from small sample size. A variety of alternative work
schedules are now being used in the National Park
Service; however, alternative schedules are not uniformly
considered by parks or offices within the National Park:
Service, and their use is primarily a matter of local
discretion and/or managerial support* After approval
has been granted, it would appear that no formal process
now exists for review or continued monitoring of these
approved schedules.
Available literature suggests that alternative/
flexible work schedules can lead to improved employee
work satisfaction, higher morale, decreased absenteeism,
and reduction in employee turnover. Claims were
initially made that alternative work schedules also
brought about increased productivity, but the literature
now suggests that no change in productivity will occur as
a result of the adoption of alternative work schedules.
This would seem to be consistent with the work of a
number of behavioralists, researchers, and academicians*
Edward Lawler^ for example, rather forcefully points out
that nthe simplistic belief that satisfaction leads to
productivity is wrong. Thus, even though improvements in
the workplace increase satisfaction, increases in
37


productivity and organizational effectiveness may not
occur11 (Lawler 1986, 135). Except for the collection of
limited survey data, no research with respect to the
above conclusions, or that applicable to the effect of
alternative work schedules on leave use or personal
performance, has ever been conducted within the National
Park Service.
There have been some very significant changes in the
practice of management, organizational behavior and
theory, and general personnel processes, in the past 50
years. Although many of these changes and concepts
originated in the private sector as a result of
competitive practices, their adoption in public sector
has been quite rapid, and the public sector has also been
responsible for the introduction of several innovative
pay plans and employee benefit programs* Tom Burns
maintained that organizations were the simultaneous
workings of at least three social systems:(1)formal
authority, (2) cooperative, and (3) political (Burns and
Stalker 1961)_ It was his emphasis on the organization
as a social system that is the most useful basis for
further discussion of this problem* With this in mind,
it is suggested that although satisfaction and motivation
are not necessarily mutually exclusive, they do have some
very discernible differences*
38


When considering the organization as a social
system, and each individual as a part of this system, the
individual's needs therefore become a matter of
organizational and managerial importance. Expectancy
theory {Vroom 1964) suggests that all employees do not
have the same need level, but rather that policies must
of necessity consider the needs of all employees and
permit individual choices wherever possible. Flexible
work hours are rapidly becoming, if they have not already
become, a employee benefit and/or discretionary option.
Quite frequently the workplace is used to satisfy needs
other than those related to the individual^ economic
condition, self-esteem and/or creative fulfillment. In
this manner the workplace occasionally serves as an
outlet for, and an extension of, individual expression.
Accordingly, employees motivated in this manner will
often seek from the workplace social accommodation for
many of their personal needs. Project ACHIEVE, for
example, was introduced by the Department of Interior
initially to better emphasize the needs of the
handicapped employee, and secondly to encourage and
accelerate federal government measures productivity
improvement; f,.. .carrying out programs with the miniimim
number of employees and getting a dollar's worth for a
dollar spent.,""
39


CHAPTER III
REVIEW OF THE LITERATURE
Theoretical Review
Since the advent of alternate work schedules in
private and public institutions, a variety of experiments
have taken place with an equally diverse number of
outcomes. Since our emphasis here is on the effect of
these schedules on productivity, literature applicable to
the study and/or effect of such work schedules was
reviewed* Although numerous articles, of necessityf
include observations of the possible or potential causal
effect of alternate work schedules upon productivity,
only six (not including those studies conducted by the
U.S. Office of Personnel Management)/ directly assess
this phenomenon.1 Not surprisingly, the literature
tends to identify that there is a definite weakness in
the data supplied since so much is "soft" or largely
based upon opinion rather than more reliable "hard data.
Even the reports prepared by OPM to the President and
Congress (U.S. Office of Personnel Management, Interim
and Final Reports 1981 and 1982) were suspect by the
General Accounting Office (U,S, Comptroller General 1980)
because of weaknesses in testing, GAO questioned whether
the prevailing methods of analysis used would provide
l
Harrick, Vanek, and Michlitsch (1986), Kim and Campagna
(1981),Narayanan and Nath (1982), Orpen (1981)f Rainey
and Wolf (1981),and Schein, Maurer, and Novak (1977).


Congress the information it needed to decide the extent
to which variable work hours should be allowed in the
Federal work force.
A study that was conducted by the American
Management Association, through Goodneasure, Inc, {1985 >,
attempted to identify the similarities and differences
among all of the important work innovations that have
been introduced in America since the 1960s. Goodmeasure
selected 21 of these innovations that they considered to
be major work alternatives, and they further categorized
and placed them into four groups, or clusters, that
seemed to share common features;(1)job-related work
alternatives, (2) quality of working life alternatives,
(3) alternative organizational structures, and, (4)
employee participation and control. The study utilized
data obtained from survey responses from a random sample
of 10,000 American Management Association Members
(n=l/618). The results, by industry distribution,
indicated that from those alternative work arrangements
applicable to flexible schedules or to a flexible
workplace, government was the major user. The public
sector also used more work innovations, on an average,
than did the private sector.
The concept of encouraging worker participation was
popular in the 1930s, on the theory that an involved
41


employee is subsequently more satisfied with his/her
position and thus is inclined to be more productive-
Within the past two decades, this approach has been
broadened and given greater emphasis with the development
of a variety of employment policies and practices
covering work scheduling ("Flexitime, Dunfs Review
1977)- There is extensive literature which covers the
quality of work life movement and the use of alternative
work schedules, but that which specifically addresses the
effect of alternative work schedules upon productivity,
particularly within the Federal Government, is either
rather limited or only superficially deals with the
matter of productivity impact*
One particular study, however, by Harrick/ Vanek,
and Michlitsch (1986), did use a pretest/posttest design
and a longitudinal frame covering 18 months of
productivity and leave usage time, and provided for 16
months between measures of worker satisfaction. This
study took place in an office of a federal government
agency, and reported study results indicating that there
were some significant reductions in leave usage, no
change overall in employee satisfaction, and mixed
results with respect to productivity. Federal
legislation that authorized the three-year Federal
Government experimental program with compressed and
42


flexible work schedules also required that each
experiment be closely monitored and provide for
evaluative reporting.
This literature review begins with an historical
summary of the major events and Federal legislation which
led to the establishment and growth of the alternative
work schedule. Some definitions of the principal
schedules adopted by industry and government are then
reviewed with emphasis upon announced advantages and/or
disadvantages Although this review highlights in
general basic literature devoted to the effect of
alternative work schedules; the findings of some of the
more significant studies of the impact of alternative
work schedules on productivity, employee morale, and
other performance factors, are illustrated by table 3.1.
Table 31
Findings of Studies/Research Experiments
Effects of Alternative Work Schedules
Study Effect upon Productivity Effect Upon Morale Other Effects &
Inproved NC Decreased Ionroved NC Decreased Favorable UnFavorable
Craddock,
Lewis &
Rose (1981) 42.8% 52.4% 4.8X
Dunham, Pierce Studies indicated that
& Castaneda effectivenessproducti vity
(1987) changed very 1ittle.
Evans & Not measured.
Partridge
(1973)
Use of
69. 30.3% 0.0% Leave
Anticipated reactions were
confirmed. No signincant
enhancements noted.
Some improvement,
but 1ittle noticeable
change over time.
43


Table 3*1.(contd,)
Study Im Effect upon Productivity oroved NC Decreased Effect Upon Morale Zmroved HC Decreased Other Effects Favorable UnFavorable
Finkle (1979) Not measured. Generally positive. Reduction Increases in leave, in cost/ absenteeism, overhead, and tardiness.
Golembfewski, Hi lies & Kagno (1974) Initial increase, then maintain previous levels. Significant increases were noted. Some supervisory concerns, however. Conmuting, quality of life.
Golembtewski, Yeager & HHles* (1975) Productivity was reported as being increased. Improved. But very smal)sample (n=ll).
Harrick t Vartek & Mi chiitsch (1986) Mi wed results. Of six tasks observed, productivity increased in three and decreased in three. Except for satisfaction with an AWS employee work satisfaction and/or morale did not change. Both sick and annual leave use went down.
Hausser, D. (1980) No significant effect. No significant effect. Convenience for employees.
Ivancevich & Lyon (1977) Increased productivity for 1st 12 months, but had no effect after ZA months. Improved, but not to a measurable level Job satisfaction satisfaction and stress.
Kim & Campagna (1981) Mixed reactions on performance, but generally positive. No significant effect. Reduced short term absenteeism.
Narayanan & Nath (1982) No change in reported productivity. No significant effect.
National Center for Productivity and Quality of Working Life (1977) Six activities were measured, all initially showed improvementr and then three declined. Improved. Attitudes of employees were enhanced but not supervisors. Absenteei sm increased.
Orpen (1961) Negligible effect on performance or productivity. Improved job satisfaction.
Pierce & Newstrom No reported improvement in productivity but performance was improved. No significant relationship with job satisfaction. Reduced absenteeism.
Rainey & Wolf (1981) Experimental and control groups experienced reduction in quantity and improvenent in quality. Employee satisfaction with job and working conditions showed improvement. Short term gain versus long term alienation.
Ralston, Anthony & Gustafson (1985) 24% productivity improvement when shared resources were a factor. No change when shared resources were not a pre-existing factor. Not measured. Productivity can be expanded without increasing the capital outlay.
44


Table 3.1,1 [contd.)
Study Effect upon Productivity Effect Upon Morale Other Effects
Iroraved ______Decreased Itnroved NC Decreased Favorable Unfavorable
Schein, Maurer
& Novak
(1977)
U.S. Office of
Personnel
Management
(1981)
U.S. Office of
Personnel
Management
(1982)
No clear-cut conclusions
with regard to the impact
of AUS on productivity.
Suggested positive effect.
46% 46% 8% Not specifically measured,
Reported an overall increase but overall responses were
of 43% in quality, 51% in generally positive,
quantity, and 44% in product
timeliness.
30% 60% 10%
From approximately 325,000
participating employees, it
was determined that a more
favorable rather than an
unfavorable productivity
impact occurred.
Morale of participating
employees Improved, but
morale of those who had
to "take up the slack,"
decreased. HanagerUl
and supervisory moral
unfavorably impacted.
Savings in energy,
travel.Reduction
in leave usage.
Experiments
in 65 units
dropped.
Enployee travel
costs reduced
slightly.
Numerous
problems for
management,
Historical Analysis
In an attempt to seek a better way to make use of
existing, but antiquated, public transportation system,
an experiment took place in Metz, France in 1955, Public
schools and all employees adopted a citywide staggered
start system. Because of the success that was achieved
from these alternating starting times, the concept of
stagaering starting times was adopted by a number of
European cities. During the 1960s a number of German
industrial firms initiated employee programs that
permitted the use of 'flexible*1 work schedules and
"gliding time" as an inducement to expand the existing
45


workforce and reduce traffic congestion* Although these
proved to be very popular and effective programs, the
needs of management to respond to periods of critical
work load led to the establishment by these firms of
certain "core hours when all employees were required to
be present. By 1972, it was estimated that about a third
of the Swiss labor force, as well as substantial numbers
of white collar and other workers in Germany, England,
France and Scandinavia were working under flexible work
schedules
Flexitime appeared in the United States in the early
1970s, introduced by a number of firms that had
international affiliations By 1973 there were
approximately 30 large U.S, firms and 200 smaller
organizations permitting certain employees the
opportunity to select, within agreed upon limits, their
working hours (Hedges 1973; Newstrom and Pierce 1979;
Zawacki and Johnson 1976) A variety of reports and
informal accounts offered enthusiastic encouragement, and
a number of early examinations and experiments involving
flexitime also attested to the advantages of nearly all
imaginative job design programs that were undertaken,
However, many of these early flexitime examinations were
descriptive and/or anecdotal, and lacked the degree of
scientific rigor that is normally desired (Bolton 1971;
46


Wade 1973; Narayanan and Nath 1982)). There were three
early studies which did investigate flexitime in an
experimental fashion (Evans and Partridge 1973;
Golembiewski, Yeager, and Hilles 1975; Schein, Maurer,
and Novak 1977) / and each these of studies noted
improvement in employee satisfaction and detected only
minor changes in productivity.
In 1978, the Congress of the United States found
that these new trends in the usage of four-day workweeks,
flexible work hours, and other variations in workday and
workweek schedules in the private sector demonstrated
sufficient promise, and that controlled experimentation
by Federal Agencies over a three-year period was
therefore warranted. With a number of extensions,
experimentation continued until 1985, at which time
authority for use of alternative work schedules was
permanently granted. The background statement contained
in the legislative history of P.L. 97221, indicated that
considerable weight was given to the data contained in
the Interim and Final Report to the President and the
Congress regarding the alternate work schedulesf
experimental program-
The Interim (1981) and Final(1982) reports prepared
by the Office of Personnel Management asserted that the
47


program, in most cases, was beneficial to the public and
to the employees themselves. However, the reports also
indicated that there were some notable examples of
failures.
Experiments with alternative work schedules in the
non-federal public sector are now receiving only limited
attention and, thereforer most of the literature that is
available is primarily descriptive, dated, or skewed in
favor of the alternative schedules. Christensen(1989)
survey of senior human resource executives, for example,
was reported to be not representative of U,S. businesses
or even of very large corporations, and was considered by
Christensen to be possibly skewed in favor of firms that
provide flexible scheduling and staffing alternatives -
since it was more probable that firms providing flexible
schedules would have participated in the survey (surveys
were mailed to 2,775 executives, but only 521 responded)
Another survey, involving 308 of a total of 500
participating managers, conducted by the American
Management Society Foundation (1989), reported that the
major advantage of flextime was 11 improved employee
morale11 and about half of these respondents also stated
that flextime also decreased tardiness and absenteeism.
This same AMSF survey noted that there are problems. One
of the respondent managers stated on page 2 of this
46


survey that: We feel flextime, rather than improving
productivity, has reduced it because of lack of
supervision during all hours of work. Also, there are
distractions caused by co-workers coming to and leaving
work at different times.M
In his analysis of flexitime in the public sector,
Rubin (1979 280) noted that flexitime has been
implemented in the city government of Inglewood,
California, with positive results.,..11 A March 1990
telephone call to the city administrator of Inqlewood
indicated that flexitime was discontinued, and since it
happened so long ago, only vague impressions of the
reasons for this could be offered* This situation seems
to have occurred with respect to experiments in both the
public sector and the non-public sector, and, although by
itself it is not remarkable, it does suggest that most of
the literature now available, particularly as it involves
the non-federal public sector, may suffer from lack of
currentness* City government and counties have become
involved in experiments in areawide flexitime to either
reduce pollution or improve rush hour traffic problems of
other such associated reasons. Rosow and Zager (1983)
have identified efforts in several large cities, and in
the combined area of Seattle and King County, Washington,
to persuade employers to adopt flexitime for their
49


employees State government experiments were highlighted
by McKann {1973) with respect to certain studies on
Florida experiments, and there is a fairly comprehensive
review of available experiences in a Golembiewski and
Proehl article (1980) which summarizes public sector
applications of flexible workhours. Golembiewski and
Proehl note that although public sector flexitime
applications have the same pattern of positive vs.
negative effects as business applications (32 public
sector studies were reviewed), they have some concern
about the niraber and quality of the public sector
studies. A field study which took place in an office of
a federal government agency (Harrick, Vanek, and
Michlitsch (1986), used a pretest/posttest design and a
longitudinal frame covering 18 months of productivity
time, had not as yet been conducted.
Of particular interest is the paper prepared by Kim
and Campagna (1981) on the effects of flexitime on
employee attendance and performance at a county welfare
oftice. This study attempts to justify the argument that
flexitime has a positive impact on attendance and
performance among public sector employees, but it also
notes the potential for the ,fHawthorne11 (Mayo 1933)
50


and/or the Rosenthal effect," 2 since the employees
were certainly aware of the experiment and were not
blind" to what was going on- In June 1977, the Kentucky
Department of Personnel began a ten-week trial of
flexitime (Craddock, Lewis, and Rose, 1981).This study
dealt mainly with attitudes and the satisfaction of the
employees themselves with the experiment. The effects of
flexitime upon productivity were measured only through
the comments of the employees involved and how they felt
about their performance; if it improved, if they only
"held their own,11 or they perhaps fell behind- Since
these data are highly subjective, the study is
significant for review purposes solely as a measure of
employee attitude and interests in improving family life.
The literature does not offer much insight with
respect to flexitime experiences in city or municipal
goverment one study commented upon the indirect effect
of recruitment for a large general hospital upon a small
private hospital, and its subsequent adoption of
flexitime to resolve a staffing problem with respect to
nurses, was published by M. Ami, Director of Nurses at
2 (Rosenthal 1976) Rosenthal suggests that if experimenters
can affect the behavior of those with whom they
interact, than others also may affect behaviors based
upon their expectations of subiect behavior.
51


Lakeland Manor Hospital(1983) This article provided
some penetration as to how flexible schedules of work can
be used to attract and retain skilled employees.
Union involvement, with respect to alternative work
schedules, in the available literature was surprisingly
light. For Federal employees, this could possibly be
because the Federal Services Impasses Panel had
previously held that federal agencies had to meet a
statutory burden of proof in rejecting union requests for
flexitime workweeks on grounds of decreased productivity,
diminished service to the public or increased costs,3
However, by passage of P. 99-196 (99 Stat.1350 1985),
collective bargaining rights to flexitime with respect to
union representation have now been fully ensured*
Alternative Work Schedules Definitions and Usage
Since the classic Western Electric-Hawthorne study
in 1927, management has become increasingly aware of the
important effects of group attitude and climate on work
performance, and of the capacity of every person for
work at whatever level of proficiency they are capable of
achieving. Regardless of what the acceptance may be of
these known behavioral characteristics, one of the
3 Federal Services Impasses Panel, Case No. 83FSIP64,
5 August,1983.
52


traditional assumptions about how organizations are best
run is that everyone should come to work at the same time
and leave at the same time. This assumption is
consistent with the idea that standardization is
important and that everyone should be treated in the same
way. It is inconsistent, however, with the reality that
people have different preferences about when they want to
come to work and when they want to leave work. It also
ignores the fact that people find themselves in different
family situations, that transportation is not equally
available for everyone, and that there are a nmnber of
disadvantages in having everyone arrive simultaneously,
namelyf overcrowding in elevators, etc., transportation
problems, and so on. There are, of course, certain
undeniable advantages to regular work hours from a
supervisory and managerial perspective, and the
literature produced to data not only describes the varied
alternative work schedules available or being currently
used, but also identifies the advantages and
disadvantages of each.
Although there are many variations of alternative
work schedules discussed in available literature, two
principal alternatives to the predominant 5-day, 9 to 5,
40-hour week have been essayed and/or adopted, the
compressed schedule of a four-day work week and
53


11 flexitime.11 Flexible workplace arrangements will be
discussed later. There are several kinds of flexitime
programs that are and have been followed. The most
limited of these is flexitour, under which employees
choose a starting and quitting hour that they follow for
a specified period of time, such as a week or a month.
Another model is gliding time, in which a standard eight-
hour day is worked, but each employee can vary his or her
own starting and quitting times daily. The variable day
system is the most flexible where more or less eight
hours are worked in a day, but debit and credit hours can
be accumulated as long as they balance at the end of the
pay period. In other instances, a more restrictive
flexitime schedule is adopted which provides for an
employee to select his own tour, but must stay on this
schedule unless approval to change has been received.
This could mean a selected starting and departure time
for a eight hour dayf 5-day week.
Flexitime
Since Messerschmidt-Bokow-Blohn introduced the
concept of gliding time as a means of reducing traffic
congestion in 1967, the system has come to be called
flexitime. The Dictionary of Personnel Management and
Labor Relations (Shafritz 1980) provides the following
definition on pages 113-114:
54


-flexitime, flexible work schedule in which workers can,
within a prescribed band of time in the morning and again
in the afternoon, start and finish work at their
discretion as long as they complete the total number of
hours required for a given period, usually a month. That
is, the workday can vary from day to day in its length as
well as in the time that it begins and ends. The morning
and evening bands of time often are designated as 11 quiet
time.n Telephone calls and staff meetings are confined to
"core time which generally runs from midmorning to mid-
afternoon. Time clocks or other mechanical controls for
keeping track of the hours worked usually are a part of
flexitime systems.
Business Week in a 1972 article which commented
briefly on flexitime, stated that 11...Flexitime is
shorthand for flexible working hours..Elbing, et al.
{1974f 29), covered all Alternative Work Schedules as
"flexible workinghour plansHedges (1977 62) noted
that T, *.. Schedules that compressed a full 40-hour
workweek into 4, or even 3f days dominated the early
innovations* But before the mid-1970s, a different type
of schedule flexitime gained prominence - # 11 The
ability to arrange schedules to fit individual and joint
needs was addressed by Swart <1974), Evans and Partridge
(1973), and Baum and Young (1973 and the term
"flexitime" appeared in public administration and
personnel text books shortly thereafter.
Buchele (1977) describes MFlexitime and the Four-Day
Week as an arrangement that allows workers a couple
55


of hours latitude in starting and stopping work each day
without reducing the total time worked per week....11
Shafritz, Hyde, and Rosenbloom (1981)introduces
flexitime in a chapter on job design, quoting Donahue^
(1975) definition of flex time as it is applicable to New
York office employees and how it differs from the federal
designation of T,flexitime.11 This clearly points out how
pure flex time permits employee selection of the hours of
reporting and departure as opposed to the federal process
of mutually agreeing to a starting and stopping time that
may differ from others in the work unit but cannot be
changed without approval.
The advantages of flexitime, specifically its effect
on the quality of both work and family life, have been
enthusiastically described by a number of individuals.
Bernardrs (1979) rather positive article, as well as the
work of Newstrom and Pierce <1979), describe the
potentials for the system but caution that careful
planning is important prior to any implementation, Nollen
(1980) suggested that flexitime will extend to 25 to 30
percent of all workers by the end of this decade since
(1)firms will receive economic gains that endure at a
cheap price, (2) flexitime is broad-gauged; it can and is
being used in many different work settings, and (3)
flexitime's track record in Europe has already been
56


established. More balanced reports are provided by Rubin
(1979), Owen ¢1977), and Wheat (1982)* Rubin and Owen
both point out the critical need for planning,
communications and supervision with respect to "core"
time. Core time, the hours during which all employees
must be present (Finkle 1979), along with "bandwidth"
(the period within which all hours must be worked) may
possibly present a scheduling problem for some
individuals if management is too restrictive in its
designations. Wheat recounts past successes, including
recruitment enhancements, but decries the control
mechanisms and the abandonment of some otherwise useful
irregular tours.
A representative cross section of the literature
suggests that alternative work schedules can potentially
offer a wide array of advantages, but the AWS system also
has disadvantages. Reported results from a variety of
scientific experiments and operational experiences range
from no change to some degree of either a positive or
negative effect, with the exception of job attitudes and
leisure-time satisfaction, which reportedly are improved
by alternate work schedulesf the predicted consequences
of work schedule experiments, with respect to
productivity are that of no significant change.
Based upon existing theory and the observations from
57


a number of experiments, a variety of positive effects
and outcomes can be obtained if organizational
implementation of an alternate work schedule program has
been appropriately designed and managed. It has been
claimed that flexible work schedules can reduce tardiness
and absenteeism, reduce overtime, increase employee
morale, increase productivity, reduce turnover, improve
planning and communication, reduce commuter congestion,
increase customer service, broaden work opportunities,
and achieve better utilization of recreation and service
facilities. Although it may be possible to achieve some
of these organizational attributes, it would be highly
unlikely that they could all occur through any single
experiment. The themes of the articles by Coltrin
(1981), Hedges (197), and Rosow (1983), bring this out
very clearly. In the Coltrin article it is mentioned
that while most research seems to demonstrate the
existence of favorable results, few, if any, system
failures have been studied. This same theme is
expressed by Hedges, who cautions that not every
environment offers the same prospect for success? in
fact, she notes that a few establishments have abandoned
flexitime as unworkable.
58


Rosow reports that flexitime fails about five
percent of the time, compressed workweeks about 25
percent* Consequently, although there are many positive
claims that would assuredly seem to be significant
recommendations for adoption of alternative work
schedules, these also have been somewhat balanced by
certain noted disadvantages.
The principal disadvantage that has been expressed
is that supervisors and managers are not available or
present at all times lends to the fear that employees
will take advantage of the situation. This problem was
noted within the literature in most instances, and was
probably the largest concern of management cited in
research material.4 * * * This conclusion was rather vividly
4 Simcha Ronen, ,fArrival and Departure Patterns of Public
Sector Employees Before and After Implementation of
Flexitime, Personnel Psychology, 34, n. 4,(1981)*
G. W. Rainey, Jr., L. Wolf, "Flex-Time: Short-Term
Benefits; Long-Term...?M Public Administration Review,
41,n.1,{Jan/Feb, 1981). G.W, Rainey, Jr., L. Wolf,
f,The Organizationally Dysfunctional Consequences of
Flexible Work Hours: A General Overview," Public
Personnel Management,11,n. 2,(1982). J. N, Hedges,
"Flexible schedules:problems and issues,,f Monthly Labor
Review,100, n. 2, (February, 1977). S. A. Coltrin, B.
D. Barendse, nIs Your Organization a Good Candidate for
Flexitime?11 Personnel Journal,60, n. 9 (September,
1981).J. S. Kim, A. F. Campagna, ^Effects of Flexitime
on Employee Attendance and Performance: A Field
Experiment," Academy of Management Journal.24, n. 4 ,
(1981).J. Welch, D. Gordon, HAssessing The Impact of
Flexitime on Productivity,n Business Horizons> 23, n. 6,
(December, 1980). R. Rubin, "Flexitime: Its
Implementation in the Public Sector," Public
59


supported in the review of productivity, leave usage and
employee satisfaction in the National Park Serviced
Rocky Mountain Regional Office, One divisional unit that
had but limited supervision was permitted almost infinite
flexibility in its work schedule and eventually the
employees and supervisors became almost contemptuous of
their responsibilities* As a subsequent result,
productivity and morale in the division significantly
deteriorated in less than three years.
It was suggested that flexitime creates extra costs
for lighting, heating and air conditioning. A study of
300 European countries conducted by Management Centre
Europe (1976) indicated that cost was their largest
complaint with respect to flexible tour experiments,
followed by the unavailability of key personnel and loss
of organizational control. Limited applicability
sometimes exists and, consequently, flexitime may not
lend itself to every situation (Coltrin and Barendse
1981; Rainey and Wolf 1981).The possibility that
overtime iaay be reduced under a flexible work schedule
may be seen as a disadvantage by certain employees.
Union concerns about this have been noted by Hedges
{1977) f Rubin (1979), and Owen (1977). Disadvantages
Administration Review. 39, n. 3, (May/June, 1979).
60


associated with flexitime, from an organizational
perspective, seem to revolve around control and
coordination as well as the possibility of additional
cost, whereas employee concerns seem to focus on possible
increased work-related transportation and loss of
overtime. Narayanan and Nath (1982) report on a research
experiment which suggests that contrary to previous
research (Evans and Partridge 1973), and other
theoretical positions which view flexitime as job
enrichment (Gibson, Ivancevich, and Donnelly,1979
employees reported no change in attitude with respect to
satisfaction with work. Nollen and Martin (1978)/
however, suggest that job satisfaction may more closely
follow experiments with flexitime when sound planning and
efforts to restructure positions were initiated first-
Compressed Schedule
For purposes of this literature review, a compressed
work schedule will mean the compression of a full work
schedule, whether it is based upon a 40-hour week or a
80-hour bi-weekly, into fewer days than accustomed or
normal.An example of this would be four 10-hour days.
Anything less than the customary nuxaber of work hours in
the work period would be part-time employment- There is
a tendency to equate flexitime with such systems as the
compressed work week, such as the 5/4^9 or the four 10-
61


hour days. But they are distinctly different. Rainey
and Wolf (1982) in discussing unanticipated consequences
of experiments with alternative work schedules, note that
the tendency to confuse the two different systems leads
to certain negative impressions regarding the 'pure11
flexitime, whereby an employee selects his/her own hours
of work (exclusive of core time) and compressed time
schedules, including the 5/4-9, which retain fixed work
shifts, removing any semblance of personal autonomy in
reporting and attendance. The Dictionary of Personnel
Management and Labor Relations (Shafritz, 1980) provides
the following definition of a four-day workweek:
reallocation of the standard 40-hour workweek over
four days instead of five. By lengthening the workday,
employees get a 3-day weekend every week with no loss of
pay. This concept differs from the 4-day/ 32-hour
workweek that some union leaders advocate
Unfortunately, there is no attempt at a
comprehensive definition for compressed workweek or
compressed tour in the Dictionary. Newstrom and Pierce
(1979) refer to the "4/40" plan which dated back to the
1940s with both the Gulf and Mobil Oil Companies and
achieved some popularity during the early 1970s, This
plan required employees to work four 10-hour days^ with a
three day weekend. The "eight-day week11 is also mentioned
as an innovation, but unlike the 4/40, the worker then
has a four-day break rather than a three day before
62


starting the cycle again. Atwood (1979) provides several
examples of both techniques along with use of part time
workers in assessing work schedules and their
contributions to productivity.
An excellent definition of compressed work schedules
can be found in the U.S. Office of Personnel Management's
Interim Report to the President and the Congress (1981):
...Any work schedule which enables a full-time
employee to complete the basic work requirement of
80 hours, in less than 10 full workdays, in each
biweekly pay period.*..
In Appendix C of this report the initial use of
compressed schedules by the private sector in the United
States is noted:
...Unlike the flexible schedule, major activity with
the compressed schedule has been in the United
States where it also originated. During the 196Os
some business leaders in the American Management
Association studied the feasibility of reducing the
number of days rather than hours worked per week.
Out of this effort came the compressed schedule or
4-day workweek.*.
Use of the compressed schedule had expanded
significantly between 1970 and 1973. In the previously
cited interim Report to the President and the Congress,
it was stated that the number of companies offering a
four-day compressed schedule rose from approximately 70
companies in 1970 to nearly 000 in 1973 This report
also provided some additional support to attest to the
remarkable growth of compressed schedules: 11.. By 1978
63


an estimated 10,000 private sector companies with over a
million workers were using compressed schedules, while
approximately 7 million additional private sector
employees were reported using flexible schedules--,"
(p. C-l). The general acceptance and use of this
schedule was also noted by Elbing, Gadon, and Gordon
(1974) who indicated that, according to the April 30,
1973, Wall Street Journal and other sources, some 3,000
U.S* Companies have adopted the four-day week, compared
to an estimated 40 in 1970. Hedges (1981)provided a
graphic presentation which suggested that in 1973
approximately 1,75 percent of all nonfarm workers were
working under a compressed workweek* The 1989 AMS
Flexible Work Survey, conducted by the American
Management Society Foundation, reflected only a 1 percent
scheduling acceptance for compressed tours from 1985
through 1988. This was increased to 2 percent in 1989,
but it was stated that: "The four-day workweek shows
little acceptance Nollen (1980) in referring to the
four-day workweek commented " Four-day and other
compressed workweeks started in this country about 1970
(they were never much used in Europe) and spurted to 1.3
million workers, or 2.2 percent of the work force, by
1975... *11 These figures would somewhat tend to support
Hedges figures for 1973, but then Nollen goes on to
64


state M...Most of the Federal government work force in
Denver went to a four-day workweek in July,1979"""
However, in commenting upon the "Denver Experiment," in
its Interim Report to the President and the Congress, the
Office of Personnel Management states that. Those on
compressed represent approximately 20 percent of Denver's
Federal workforce....11 This would therefore suggest that
of those agencies electing to participate in the
experiment,12,560 employees were involved and of this
total, aDDroximately 44 percent (5528) were on a
compressed schedule. This differs substantially from the
level of interest reported by Nollen and is reflective of
the frequent disparity in statistical inferences supplied
in the literature. The same situation can be detected
with respect to data supplied regarding flexible work
schedules, Schein {1977 464) states that 11.. *The
increasing popularity of flexible working hours and the
paucity of research on organizational outcomes heightens
the need for further study. .11 Golembiewski and Proehl
(1980) also indicated that flexitime research may be
criticized on technical grounds because of some rather
obvious shortcomingsf and he goes on to identify a
variety of design deficiencies, i,e., use of post-test
only, insufficient comparison groups, faulty statistical
treatment, etc, Nollen (1979 18) identifies three
65


specific deficiencies:(1)only successful flexitime
programs are reported in the literature; (2) reports from
organizations may be biased upward as positive results
were expected and to fulfill these expectations only
positive results were found; (3) the scientific quality
of productive measurements is questionable. Bernard
(1979 56) in his article also pointed out the vagueness
of the statistical and research evidence supplied and
they raise questions as to the quality and number of
studies particularly those involving the public sector.
A particularly informative summary of the use of
flexitime and compressed work schedules was produced by
the Work in America Institute (1981). The summary was
directed by Jerome M- Rosow and Robert Zager (1983), and
it noted that flexitime options were split almost evenly
among flexitour, variable day, and gliding time. Total
growth in the use of flexitime had increased from 4.7
percent of all employees in 1974 to 11.9 percent in 1980.
Use of compressed schedules, however, had only slightly
increased, from 1.9 percent to 2, percent.
Most literature on compressed work schedules tends
to support the contention that longer daily work hours
and shorter workweeks increase the time people have for
fajnily activities, recreation or educational purposes,
reduce commuting costs, and give some employees the
66


opportunity to take on a second job* Howeverf the longer
work days required under the compressed workweek could
create the possibility of fatigue and reduced employee
effectiveness toward the end of the day and thereby
reduce the time available to the employees to engage in
personal and domestic pursuits during most of the days in
each week (Rainey and Wolf 1982). Ronald Lang, who was
with the Canadian Labor Congress, was quoted in Business
Week as rejecting compressed work weeks because they "go
beyond the bounds of health and safety standardsby
exceeding the eight-hour limit that unions fought to
establish (MEurope Likes Flexi-Time Work11 1972). Olson
and Brief (1978), however, point out that proponents of
the four-day workweek suggest that such a work schedule
not only reduces employee absenteeism, it increases job
satisfaction and productivity, gives employees more time
to "unwind" from their job as well as decreasing
employees# requests for time off for personal matters.
Nollen (1980) cautions that although the compressed
workweek is not applicable in every setting, it can help
businesses that have costly start-ups and shutdowns,
inadequately utilized capital equipment, and travel time
to work sites-
The experiment by the Federal Government which took
67


place as a result of P. 95-390, more commonly referred
to as the Federal Employees Flexible and Compressed Work
Schedules Act of 1978, attempted to provide a summary of
results whxch were taken from a sample of 70 work units
selected by OPM to be a part of a longitudinal cross-
sectional study. Work units were not required to submit
the narrative evaluation report to 0PM, and since tms
narrative non-technical assessment represented the work
unites total evaluation of the program, hard data and/or
statistical evidence was not available. In addition,
much of the data used to form the statistical profile was
later provided by the work units themselves. Since these
work units actively sought and preferred alternative work
schedules, the findings of the study are probably biased.
OPM provided on page 23 of this report that: According
to the report, The purpose of the profile was to enable
OPM to collect a standardized set of quantitative
information summarizing experimenting organizationsr
experiences with AWS.,.11 (p- 23). It would appear,
however, that much of the data was collected in a post-
test fashion without expected scientific rigor. It is
also noted that OPM had taken a rather positive approach
to this national experiment, and quite possibly elected
not to devote much review or comment to unsuccessful
experiments Thus, OPM's analysis of results of
68


experiments with compressed tours seem only to reveal
positive comments as was the case with flexible work
schedules. On the other handr individual agency reports
(p,125) offer a conflicting picture. Agencies reported
that under the 4-10 plan 55 percent of their
participating work units experienced success, 35 percent
of the work units experienced only partial success, and
10 percent experienced total failure* Under the 5/4/9
plan, 62 percent experience success, 28 percent
experienced partial success and 10 percent experienced
failure.
Flexible Workplace Arrangements
Although flexible workplace arrangements are not a
part of this study, recognition for their potential
influence on work schedules has been noted in this study
since the actual location of the work performed would be
shifted away from what is now the primary, and the
traditional, work site. Thus, telecommuting could
eventually impact residual work schedules at the
traditional work site.
The idea of "telecommuting" reportedly began nearly
20 years ago with Jack Nilles, an engineer who wanted to
reduce auto pollution in Los Angeles. He proposed
69


setting up satellite offices in the suburbs near the
homes of clerical workers Instead of driving downtown^
clerks would use computer terminals from these satellite
offices In a six-month experiment in 1973, Nilles
claimed that turnover dropped to zero from 33 percent,
office rental costs were reduced, and productivity rose
18 percent (Nilles, et al1976), Although telecommuting
is still rather new, an estimated 13 to 15 million people
now work at home at least part of the time (Atkinson
1985). This accelerated interest in teleconunuting stems
not only from its being a popular option from an employee
standpoint, but also from its mitigating effect on peak-
hour traffic, work travel, increased energy consumption,
additional office space, and environmental concerns.
Basically, telecommuting allows people to work part
or full time at home-based or satellite offices, instead
of commuting every day to work. It can also somewhat be
compared to an updated version of the age-old Mcottage
industry" concept (Edwards and Edwards 1985). Like other
alternative work options, the process offers a wide array
of advantages and disadvantages. One major problem that
telecommuting does present, which is not found in other
alternative work options, is the difficulty that the
worker will face in attempting to successfully merge work
and home life. When one is working at home/ there will
70


be a number of distractions and diversions normal to a
domestic environment that is not found in the typical
oftice setting. How the worker effectively deals with
this problem is unquestionably the major issue impacting
telecommuting. Gil E. Gordon and Marcia M. Kelly, who
are regular contributors to the Telecommuting Review (The
Gordon Report), rather thoroughly covered most advantages
and disadvantages in Telecommuting: How to Make it Work
for You and Your Company (1986), while also stressing
that telecommuting is basically a social phenomenon.
Thus, most of those human/emotional responses and
circumstances related to change become equally applicable
with respect to telecommuting implementation.
The term telecommuting additionally refers to the
process of electronically transferring the results of
job-related work from a home, or satellite site/ to the
principal office It is questionable that such extreme
decentralization would have been possible without the
availability of the current vast array of sophisticated
new communications and computerized equipment. It
therefore becomes possible for firms, such as the J.C.
Penney Company, to relocate from large cities to small
towns like Plano, Texas (Lewis 1988). Thomas B. Cross
(1986) has suggested that telecommuting is now in the
same position today that data processing was in 25 years
71


ago, i*e., executives do not grasp the significance of
the new technology and thereby cost their organizations a
competitive edge by losing time in adopting it. There is
also, in all probability, more managerial resistance to
the basic concept of telecommuting than there had been
earlier during the initial consideration processes of
flexitime and compressed time. Trust certainly is
important, but many managers equate employee performance
with such non-merit factors as punctuality, sociability,
and appearance rather than with quality and quantity of
work produced {Downing-Faircloth 1982).
Unlike flexible work schedules^ flexiplace would
appear to have aroused in the government a more active
and earlier interest than did flexible work schedules.
In addition to the work of Jack Nilles for the Department
of General Services, State of California (Martin 1909), a
telecommuting pilot project was initiated by the South
Coast Air Quality Management District in El Monte,
California on October 7,1988, involving 30 employees;
and the County of Los Angles is currently moving forward
with a pilot project that could involve up to 2,000 of
the countyfs 17,000 employees who work in downtown Los
Angeles (Gordon 1989) Other similar projects have been
initiated, or the telecommuting process is being
carefully considered, by other county and city
72


governments. The city of Ft. Collins, Colorado, for
example, has recently adopted a policy regarding
telecommuting which very clearly points out that the
principal goal of the program is improvement in
productivity. This aspect of proposed flexiplace
application is unique to the degree that most proposals
heretofore have stressed either cost reduction or
employee retention,
A governmentwide project is now underway which will
permit Federal employees to participate in flexible
workplace arrangements, Sponsored by The President's
Council on Management Improvement (PCMI), this project
will test the feasibility and utility of flexible
workplace arrangements, on a limited basis, through pilot
tests scheduled to begin in early 1990. The project co-
directors are Thomas Cowley7 General Services
Administration, and Dr. Wendell Joice, Office of
Personnel Management. Dr. Joice had earlier been
responsible for the Office of Personnel Management's
evaluation of the flexible and compressed work schedules
project in the Federal Government. Guidelines have been
issued (Presidents Council on Management Improvement
1990), and most Federal agencies are participating in the
project- It will be up to individual agencies to decide
how many and the type of employees that may participate.
73


The Office of Personnel Management is publishing a bi-
monthly newsletter, entitled Flexiplace Focus for the
benefit of participants in the Federal flexible workplace
pilot study, and Federal agencies are also associated
with certain pilot projects being conducted by state
agencies *5
Productivity
The overall objective of behavioral science is to
find out how to increase the productivity of
organizations through optimal use of their human
resources. The amount of literature available which
touches upon human productivity is enormous, and it is
not the intention of this review to attempt to cover all
of the material which has been published with respect to
productivity, but rather to focus upon that which
addresses the effect of alternate work schedules upon
productivity*
In comparing techniques of applying Human Resource
Management, Teasley {1983 363) measures and defines
5 In addition to very visible support in the State of
Colorado, the Federal Government is actively involved
in telecoramunications programs in other states, such as
Washington, and Colorado. The recent Colorado Symposium
on Telecommuting, was presented in conjunction with the
Governor's Office of Policy and Research and the U*S*
Department of Energy.
74


productivity as .the ratio of an organizational output
to an input (P=0/I." Hatry (1982) suggests that
although that may be the most common definition for
productivity, to apply it to any particular government
service is a complex task and subject to controversy. He
further offers on page 427:
*.-Productivity measurement generally has been
defined in the public sector as encompassing both
efficiency and effectiveness. Efficiency indicates
the extent to which the government produces a given
output with the least possible use of resources.
Effectiveness indicates the amount of the end
product, the real service to the public, that the
government is providing....
The Office of Personnel Management (1978), defines
productivity as:
11 the siim of the efficiency, effectivenessf quality
and responsiveness with which products and services
are delivered.... (p.4)
In a special study conducted by the U.S. Congress
(Congressional Budget Office, 1987)t labor productivity
of the sampled federal civilian workforce was reported to
have increased at an annual rate of 1.4 percent from 1977
through 1986. It should be noted here that in the
federal government, productivity is measured in two
different ways. The first method measures determinate
labor quantities against the amount of services provided.
The second method, which was initiated in 1986, uses a
more comprehensive measure that considers labor along
75


with capital and other resources involved in production*
An analysis of productivity measurement in the federal
government (U.S. Office of Management and Budget 1964),
provided a more simplistic explanation of how the federal
government measures productivity improvement;
tfcarrying out programs with the minimum number of
employees and getting a dollar's worth for a dollar
spent"
The definition which I feel more suitably fits the
situation and circiunstances that will be presented in
this paper is that provided by Sutermeister (1976, 5):
.output per employee-hour, quality considered.ft One
productivity measurement alone, howeverf is of little
use. It is also important to consider the effect that
individual productivity may have upon group productivity
and/or organizational productivity. For this reason,
productivity consideration will include the worker, the
work process and the managerial consequences.
Since the advent of alternate work schedules in
private and public institutions, a variety of experiments
have taken place with an equally diverse number of
outcomes* Since our emphasis here is on the effect of
these schedules on productivity, the literature
applicable to the study and/or effect of such work
76


schedules were reviewed. Although numerous authors
suggest that AWS may have an effect upon productivity, as
noted earlier, there were only six attempts to assess
this phenomenon and of these, only three are empirical
studies. The literature does identify this weakness, and
notes that the data supplied have been obtained largely
from employee opinions rather than as a result of
extended term pretesting and posttesting of alternative
work schedule experiments. Harrick, Vanek, and
Michlitsch ¢1986), Kim and Campagna (1981)and Orpen
(1981) did conduct empirical studies, but only Harrick,
et al-,and Orpen employed a pretest/posttest design.
Even the Interim report (OPM 1981)was suspect by the
General Accounting Office (U.S. Comptroller General,
1980) because of weaknesses in testing, and the General
Accounting Office also questioned whether the methods of
analysis used would provide Congress the information it
needed to decide the extent to which variable work hours
should be allowed in the Federal work force.
The Interim Report (OPM 1981)indicated that
although there were some negative revelations from a
"small1 number of experimenting organizations, these
revelations should not obscure the fact that information
supplied through the overall study n..overwhelmingly
suggest[s] that the benefits derived from reducing
77


unproductive staff time and enhancing productivity among
the large majority of experimenting organizations far
outweigh the few problems encountered in administering
AWS programs" (p.43), The Interim Report tied
productivity to 'efficiency of operations, H and relied
primarily upon narrative statements since statistical
measurement of productivity was not otherwise available.
No pre- or posttesting was taken of the sample work units
selected, nor was the selection of these units random
(willingness to participate was a deciding factor). In
the 1982 OPM Alternative Work Schedules Experiment -
Final Report, deficiencies in the interim report study
were also noted and qualifying information was
subsequently provided on page 2:
The experimental program demonstrates that AWS can
not be mandated for all organizations. As the
General Accounting Office (GAO) pointed out in its
evaluation of the AWS experimental design, the
experiment selection was not a random sample of
government organizations. Instead, organizations
that thought the experiments would be useful and
effective in their operations formally submitted
requests to OPM to experiment* Such self selection
is the opposite of the random sajnple concept. The
1500 organizations and 325,000 employees
participating nationwide is still well less than 20
percent of the total Federal workforce. Because of
the self selection feature and the limited share of
the universe, it is not possible statistically to
generalize results to the entire Federal
Government
78


The Final Report also noted that:
"some organizations did experience adverse
effects in the key areas of productivity and service
to the public and all alternative work schedule
options were not equally successful in all
organizational settings- (p.4)
Literature considering alternative work schedules
and their effect upon productivity or performance
primarily dwells upon impressionable generalities
associated with new approaches to human resource
management. With respect to the concept of
telecommuting, Jack M. Nilles, president of JALA
Associates, a Los Angeles consulting firm in management
uses of information technology, notes with respect to
telecommuting, "Even the most hard-nosed skeptical
managers say, 'At least productivity hasn^t gone down,H
(McGee 1980 60). Job design is another avenue where
opportunities for job enrichment are frequently explored.
Time management, in this context, offers the job holder
through alternative work schedules the greatest possible
freedom to decide when to start work, when to stop work,
when to take lunch or a break andf if taken to the
extreme degreef how to assign priorities. Principal
emphasis in the literature on alternative work schedules
is how such schedules improve the quality of life.
Morale is often mentioned as the beneficiary of the
adoption of alternative schedules, Peter F. Drucker ties
79


morale to performance in the following manner:
11 , Morale9 in an organization does not mean that
#people get along together;f the test is performance, not
conformance*..n (1974, 455)*
The Dictionary of Personnel Management and Labor
Relations (Shafritz 1980) takes a similar approach in
defining morale: 11 .collective attitude of the workforce
toward their work environment and a crude measure of the
organizational climate. ...11
With this in mind, most literature that discusses
the application of alternative work schedules touches
only lightly upon specific and hard" examples of
productivity improvements but instead attests to
improvement in job satisfaction by the employee; less
gathering, talking and gossiping, more work concentration
and improved "morale.H Others point out the "biological
clock" process whereby any one individual will probably
not be equally productive throughout various periods of
the day. This concept is explained in more depth by
Pierce and Newstrom (1980,121)who posited ", a
flexible working hours arrangement can provide the
context for a more efficient utilization of the human 24-
hour clock (circadian or diurnal rhythms) and can
decrease the amount of stress experienced by some
employees *. .11 A more recent article (McKendrick 1989)
80


which discussed a recent AMS Foundation Survey (1989),
did not even touch upon productivity, but instead
indicated that respondents were more or less evenly split
between citing morale or decreased tardiness/absenteeism
as being the main advantages of flexitime.
In compliance with the requirements of P*L. 95-390,
the Office of Personnel Management analyzed what they
felt was a unique use of alternative work schedules which
they called the "Denver Experiment" In reporting on the
results of this experiment {0PM, 1981), productivity is
only lightly touched upon in summationf but these limited
comments in Appendix D appear to be most interesting:
.-.the 11 Denver Experiment11 has: *. Improved or had a
neutral effect on efficiency in government
operations, including productivity.*..
"have been four terminations to date, with two
more planned prior to the March, 1982 cut-off. Two
of these were the result of uncontrollable factors
unrelated to AWS. The other four were the result of
difficulties encountered with AWS. For the group,
these included productivity declines, increased
supervisory workload, and problems in ensuring
adequate employee and supervisory coverage,,**
"Overall, participating installations in Denver
perceived AWS as generally having positive or
neutral effects on the productivity of their
organizations, Forty^one percent reported no change
in the quality, or timeliness of work as a result of
AWS experimentation* Another 41% reported increases
in one or more aspects (no declines in others).
Fourteen percent indicated declines - 4% reported
offsetting effects....
Other observations on this very perplexing effect
81


attest to the lack of specificity with which the
relationship between alternate work schedules and
productivity is measured, Rubin (1979) refers to the
biological clock effect, but offers only a general
statement that "productivity has generally increased when
flexitime is used p.278). Examples of productive uses
of time are limited to observations that employees
normally start to work on time and come to work intent on
doing the job rather than just putting in hours,1 Owen
(1977,155) notes that The new systems of flexible hours
have not been in operation long enough and are not
sufficiently widespread to provide dennite measures of
their social potential.11 But then, he also warns of
potential problems with productivity by adding,
n...productivity must be maintained in each case without
undermining the earnings, hours, and working conditions
of the work force
There is a dennite change in the tenor of the
literature from the earlier to the later publications,
which is only reasonable since with increased
experimentation more research occurred, and the research
was more objective and scientifically obtained. This is
not to say that the early literature was not
appropriately scholarly^ but it was presented more often
in the context of what was then known about alternative
82


work schedules and their possible effects upon the work
force and productivity. This is particularly true of the
enormous amount of material which was developed in the
pre1980 period. The work of Bernard (1979, 54> in
examining flexitime potential notes: "As of this writing,
no analytical study exists which clearly indicates the
net value of FWH to a firm...." Yet, in commenting upon
productivity, Bernard states: "Flexitime primarily
affects productivity through one or more of three
channels: improvement of morale, reduction in stress, or
reductions of frictions such as turnover or absenteeism.f,
This finding seems to agree with later observations of
the potential or real effects. The same theme is
mentioned by Golembiewski and Proehl(1978), Newstrom and
Pierce (1979), 01son and Brief(1978>, and Nollen (1979),
who report upon the information as supplied, but conclude
in a manner similar to that of Schein (197, 465)
earlier, ,rGiven the differing outcomes of the
experimental groups, as well as experimental design
limitations, no clear cut conclusions with regard to the
impact of flexible working hours on productivity can be
made....M
A review of post 1980 literature reveals that
although there are still some obvious carryovers of the
earlier enthusiasm and credulity, later articles provide
83


a more balanced approach, still positive, but somewhat
more cautious. Coltrin and Barendse (1981,713) f for
example, point out that .The apparent success record
of flexitime may be simply the result of lack of
reporting of failures.* Managers should keep in mind
that while most research seems to demonstrate favorable
results, few failures of the system have been
studied.*.." Ronen (1981,818) in his observations with
respect to the public sector, notes that tardiness is
virtually eliminated under flexitime, but cautions,
Despite the advantages associated with flexible working
hours, many supervisors have expressed anxiety about the
degree of variation in an employee's arrival and depart-
ure times... With the implementation of a flexible
working hours program, employees tailor their schedule to
suit their personal needs and .,* these needs dictate the
characteristics of arrival and departure patterns.
Rainey and Wolf (1982, 56) also observe that Federal
employees M... will occasionally admit that fnot much
gets done around here between 4:30 and 6:30# .11
The specific impact upon productivity appears to
still be an unknown but presumedly neutral factor.
Numerous studies have now taken place which attempt to
measure the effect, but they tend to vary in their
results. In commenting upon this observation, Narayanan
84


and Nath (1982, 228) report as follows on the impact of
flexitime on productivity: The study found no
change in reported productivity as a result of the
introduction of flexitime a finding that was true for
all levels. (The evidence here corroborates the 1977
findings of Schein, Maurer, and Novak.) Narayanan and
Nath go on to say that "Convergence of these two studies
regarding the impact on productivity leads us to suggest
that the introduction of flexitime by itself does not
result in any significant productivity gains.
In another study, Rainey and Wolf (1981,60)
conducted an experiment on productivity at a bureau of
the Social Security Administration, Their findings
indicate a substantial and approximately equal increase
in accuracy of output for both experimental and control
sections, which may be attributed to the new processing
procedure. Quantity of output, however, fell
significantly.*. in the control unit and not in the
experimental unit...H There is some reason to believe
that the improvement reflects improved attitudinal
factors - commitment, enthusiasm, or sense of
responsibility rather than mechanical factors such as
reduced tardiness. ...,f Narayanan and Nath (1982 226) in
testing attitudinal factors in a flexitime experiment
conducted at a multinational corporation in the northeast
85


note the absence of research and state that more needs to
be conducted " before any generalization can be made
with respect to the impact of flexitime on quality of
supervision, job satisfaction and productivity..
It would therefore appear that attitude and personal
perceptions have to a great extent influenced what may or
may not be the impact of alternate work schedules on
productivity Management and/or supervisory employees
may view the alternate work schedules as primarily an
accommodation for the employees, and probable cause of
future operational difficulties and breakdown in control
functions_ The employees, on the other hand^ may view
this as a challenge, an opportunity, or in some instances
a device by the supervisory hierarchy to reduce costs at
their expense. If the latter occurs, it is possible,
even probable that productivity could suffer as the
employee attempts to beat ,fthe system,11 may malinger or
even consider or encourage sabotage. Flexitime is
popular among employees and some managers, but what about
the impact on first-line supervisors, and how they
perceive the effect of the alternative work schedules on
productivity? Only one article fully addressed this
issue, but the thoroughness of the work of Golembiewski,
Fox, and Proehl <1980 46-47 would seem to adequately
fill this void in the literature available. After
86


recognizing the problems, and there were manyr
Golembiewski reports in his summation:
..Despite all the problems, however, most
supervisors saw definite upbeat effects of flexitime
on productivity and morale. Although only a few of
the supervisors could make objective measurements of
productivity, almost all found it easy to make a
confident estimate of trends in output. Of the 43
supervisors who participated in the sensing groups,
more than 70 percent believed that productivity had
increased under flexitime, about 20 percent saw no
change, and only about 7 percent felt that a
noticeable deterioration in productivity had
occurred....
Golembiewski, Fox, and Proehl, also note that
flexitime will not 11. . transform ineffective supervisors
into managerial paragons. On the contrary, flexitime is
likely to exaggerate some managerial flaws: lack of
confidence in judging performance/ the need to
overcontrol, lack of trust, and so on.
Earlier it was mentioned that only six articles in
the literature available relate solely to the impact of
alternative work schedules on productivity, although
others in their analysis of alternative work schedules
touch upon the productivity outcome. The Schein, Maurer,
and Novak: article (1977) f describes the sampling of 246
clerical-level employees over a four month period during
an experimental flexible working hours program. It was
reported that the results were mixed but, overall,the
introduction of flexible working hours did not have an
87


adverse impact on productivity. Stanley D. Nollen (1979),
in evaluating the effects, worked with completed surveys
and studies, and the results deprived of the original raw
data may reflect, in part, bias or inaccurate
reporting/sampling processes but, in reviewing these 14
case studies and eight surveys, Nollen rationalizes that
the answer to the question of whether or not increased
productivity will result from alternate work schedules
lies with how the alternate work schedules are managed.
Welch and Gordon (1980) analyzed the impact of
flexitime implementation on productivity in the claims
processing department of a medium-sized insurance company
in the southwestern United States- The study covered a
period which began in November, 1977, and ran through
December, 1980. The study indicated that there were
improvements in productivity, which were applicable to
claims processing, the primary problems appeared to be
centered around personnel:
"It is probable that firms which use such a system
will require supervisors to work longer hours.
During this study, one of the three supervisors
resisted the system and attempted to negatively
influence the attitude of other supervisors. In
order to overcome these problems with lower
management, it is recommended that supervisors also
be offered a flexible work schedule.... (page 65)
The fourth and last comprehensive study solely
related to productivity and/or performance considered all
88


previous work in this field and benefitted from the
consultive arrangement with other source matter experts.
Kim and Campagna (1981) conducted a study over a four-
month period on a flexible work schedule experiment at a
county welfare agency involving 353 employees within 4
divisions. After noting the possibility of "Hawthorne"
and/or f,Rosenthal11 effects, and that the employees were
also aware that the results of the study could
potentially affect the entire agency, they still find
that:
the results of this study appear to justify the
argument that the flexitime program has a positive
impact on attendance and performance among public
sector employees. It should be noted that this
study investigated only one of several forms of the
flexitime program. Additional research on various
other forms ot the flexitime program in different
organizational contexts is clearly warranted.
(page 740)
With this last study, it can be established that it
is possible, from literature available, to determine that
the use of alternative work schedules can have a
beneficial effect upon productivityf but only when
approved for appropriate work units. Approval of an
alternate work schedule without considerable research and
planning could lead to dysfunctional consequences. A few
studies did utilize some levels of scientific rigor, and
did deal with alternative work pattern applications in an
experimental manner It is useful to consider the work of
89


Dr_ Ben Burdetsky and Dr. Marvin Katzman, who examined
through survey the experiences of organizations who had
previously implemented alternative work schedules. The
results of this survey was published in the Journal of
Systems Management, December, 1981, under the title of
"Alternative Work Pattern Applications," and the results
indicated that those responding felt good about the work
schedules, reported improved morale, but little tangible
improvement in productivity occurredf although it might
otherwise have been expected. The survey information
only, quite naturally, included data from those who
elected to respond, so it could be argued that opinions
supporting improved productivity or faltering morale may
have not been obtained from the sample available. The
work of Golembiewski in measuring professed workplace
enhancements, such as productivity, morale, etc., from
adoption of flexitime is significant. It is interesting
to compare his article (1982) f,Do Flexible Workhour
Effects Decay Over Time?,11 with some of his earlier
observations* He looks at what he calls flexitime at T=5
years, and from survey data obtained from a large firm
who had earlier experimented with flexitime, he suggests
that although a 11... substantial proportion of
respondents believe the F-T program is abused, .11 (p. 39),
he still is convinced from the survey data supplied that
90


Full Text

PAGE 1

ALTERNATE WORK SCHEDULES AND THEIR EFFECT UPON PRODUCTIVITY IN THE NATIONAL PARK SERVICE by Harold P. Danz M.A., University of Northern Colorado, 1981 A thesis submitted to the Faculty of the Graduate School of the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Doctor of Public Administration 1992

PAGE 2

This thesis for the Doctor of Public Administration degree by Harold P. Danz has been approved for the Graduate School of Public Affairs by 1klr;.. 'oTte

PAGE 3

Danz, Harold P. (D.P.A.) Alternate Work Schedules and Their Effect Upon Productivity in the National Park Service Thesis directed by Associate Professor Robert W. Gage The National Park Service is the largest bureau office within the Department of Interior. When Congress permanently consented to the authority to approve flexible and compressed schedules in 1982, each of the 10 Regional Offices in the National Park Service developed a formal or, in some instances, an informal policy regarding use of alternate work schedules. These policies ranged from very liberal to ultra conservative. Although there have been several efforts to observe and experiment with alternate work schedules to determine organizational consequences, much of the subject literature is anecdotal in nature. Those studies that did make use of experimental and control groups primarily reported mixed results or suggested that no measurable effect resulted from the experiments. Six experiments were conducted in the National Park Service, and, in addition, a questionnaire was mailed to 433 randomly selected permanent employees. The objectives of the experiments and questionnaire were

PAGE 4

to measure employee productivity, performance, use of leave, and level of satisfaction, and to determine if there were significant differences between employees working under a traditional work schedule and those working under alternate work schedule(s). Six scientific hypotheses were formulated around the above research objectives, and through a series of tests and analyses, it was determined through the use of control and experimental groups that (1) employees working under a compressed work schedule demonstrated somewhat lower productivity levels during the pretestposttest time periods, and (2) employees working under a flexible work schedule demonstrated marginally higher productivity levels. However, in each of the above circumstances the difference was not statistically significant. Use of unscheduled annual leave was greater for both compressed and flexible work groups than for those groups working under traditional work schedules. Use of sick leave was marginally greater but did not reach statistical significance for these groups. Employee turnover in the National Park Service is extremely low, and no measurable termination/hire effect was associated with the availability of alternate work schedules. Performance evaluation ratings and employee iv

PAGE 5

satisfaction levels were higher for both alternate work schedule employees and traditional work schedule employees. No significant statistical difference was detected. The form and content of this abstract are approved. I recommend its publication. Signed Robert W. Gage v

PAGE 6

TABLE OF CONTENTS Figures ix Tables X CHAPTER I. INTRODUCTION 1 II. THE PROBLEM 14 Statement of the Problem 14 Elements, Hypothesis, or Research Questions 17 Delimitations and Limitations of the Study 24 Limitations . 24 Delimitations 27 Definition of Terms 28 Summary 36 III. REVIEW OF THE LITERATURE 40 Theoretical Review 40 Historical Analysis .. 45 Alternate Work Schedules -Definitions and Usage . . . . . . 52 Flexitime 54 Compressed Schedule 61 Flexible Workplace Arrangements 69 Productivity 74 Summary 93

PAGE 7

IV. RESEARCH DESIGN . . . . . . 97 General Structure and Methodology Framing 97 Specific Procedures 103 Research Population 112 Instrumentation 115 Data Collection 119 Treatment of the Data 122 Summary 124 V. ANALYSIS AND RESULTS 126 Study Variables 127 The Antecent Variables 127 The Intervening Variables 129 The Independent Variables 135 The Dependent Variables 137 Initial Analysis 142 Results of Preliminary Analysis 142 Comparability of Demographic Variables for Experimental and Control Groups . 142 The Effect of Alternative Work Schedules on Productivity . . . . . . 148 Questionnaire/Survey Responses 157 Hypothesis Testing . . 163 Hypothesis Testing, for Hypotheses 1, 2, 3, and 4 . . 163 Hypothesis Testing, for Hypotheses 5 and 6. 176 Summary of the Findings 184 vii

PAGE 8

VI. CONCLUSIONS AND SUMMARY . 187 Work Schedule Productivity Results 189 Usage of Sick and Unscheduled Annual Leave 190 Employee Turnover in Employment 192 Employee Performance Evaluation and Morale 192 APPENDIX A. Reviewer's Guide to Questionnaire 200 B. Questionnaire 205 C. Questionnaire Significance Response Testing 212 D. Comparisons of Equality of Means Experimental Groups . . . BIBLIOGRAPHY viii 216 221

PAGE 9

FIGURES Figure 4.1 Variable Elaboration Model .... 105 5.1 Age Distribution-Human Resource Groups. 144 5.2 Age Distribution Human Resource Groups Compressed AWS Experiments. . . . 145 5.3 Age Distribution MARO Human Resource Groups Flexitime Work Schedule Experiments . 145 5.4 Age Distribution -RMNP Maintenance Group Compressed AWS Experiment . . . . 146 5.5 Age and Productivity Correlation, Human 5.6 5.7 Resource Control Groups . . . . 147 Age -Productivity Correlation, Human Resource Experimental Groups . . Productivity Human Resource Groups 148 154 5.8 Productivity -Compressed AWS 5.9 5.10 5.11 5.12 Human Resource Experiments . . . . 155 Productivity Experiment MARO Flexitime MARO -Productivity Pretest MARO -Productivity Posttest MARO -Productivity Increase/Decreases ix 156 170 171 171

PAGE 10

TABLES Table 2.1. Number of National Park Service Employees and Authorized Compressed Work Schedules 25 3.1. Findings of Studies/Research Experiments. 43 4.1. AWS Human Resource Distribution by Regions. 107 5.1. 5.2. Compressed AWS by Regions and Offices Age Comparisions for Experimental and Control Groups -F-Ratio Analysis 136 144 5.3. Mean Productivity by Organization and Experimental/Control Groups . 149 5.4. Productivity and Performance Results for 5.5. 5.6. Control and Experimental Group Employees 150 T-Tests -Difference in Means Between Pretest and Posttest Scores Response Rate of Employee Surveys Applicable to the NPS . . . 152 159 5.7. Theme Responses to the Employee Questionnaire Regarding Work Schedule Questions . . . 162 5.8. T-tests, Compressed AWS Productivity Pretest and Posttest Means. . . . . . . 167 5.9. Age -Productivity Correlation, MARO Human Resources Flexitime Experiment. . . . 170 5.10. T-test, MARO Flexitime AWS Productivity Pretest and Posttest Means. 172 5.11. Use of Annual and Sick Leave .. 175 5.12. NPS Employee Terminations, 1990 and 1991 177 5.13. T-Tests, Performance Evaluations -Pretest and Posttest Means . . . . . . 180 5.14. T-Tests and ANOVA Independent Group Analysis Summary . X 181

PAGE 11

CHAPTER I INTRODUCTION The history of civilization is closely joined with those remarkable efforts of society over the years to improve upon the methods by which human survival is directed and organized. Although it may not be possible to precisely distinguish all of the characteristics of humans from other animals, it is generally accepted that work, as a distinct activity, is more closely associated with man than any other related mental or physical endeavor of the animal species. Contrary to the more common opinion held regarding the "primitive" human, i.e., that where natural conditions permit, they work as little as possible, the facts seem to suggest quite the opposite (Herskovits 1962). Even under the most idyllic of settings, mankind has not rested nor seemingly been content with what nature has provided. Consequently, organization of work quite possibly began during the very earliest stages of human development in an effort to augment the food supply through more regulated and controlled methods of foraging, hunting, and later agriculture. The emergence of labor --labor as a quantity of effort detached from an individual's life and bought on the market in fixed amounts --had a profound effect upon what might be considered as a pure work concept or a work

PAGE 12

ethic. The so-called "industrial revolution" did inspire a demand for material and economic well-being, but it also flourished at the cost of immediate human suffering with the promise and potential for an eventual better life for all people (Gladden 1972, 258). In the human society, work has assumed a very pervasive and powerful role. Once mankind worked only in order to ensure survival, then work seemingly was cloaked in a moral and righteous fabric; all who could work did, because idleness was presumed to be bad; a person without a job was a misfit (Weber 1958), and association by profession contributed to the establishment of craft and trade unions. So intertwined has the work role become with an individual's personality, that it has been observed that we attempt to place people by asking what it is that they do rather than who his or her ancestors were (Biesanz 1964). Organizational theory therefore attempted to address concepts that expressed the idea that (1) employees are people with a variety of needs that must be satisfied if they are to lead full and healthy lives and to perform effectively in the workplace, and (2) the needs of individuals and organizations should be integrated. Max Weber credited the origin of the work ethic to Martin Luther's interpretation of "calling." According 2

PAGE 13

to Luther, by laboring in your calling you expressed brotherly love through the services and products you produced for society. Weber therefore concluded that the pursuit of wealth as an end in itself was reprehensible; but the attainment of it as a fruit of labour in a calling was a sign of God's blessing .... (Weber 1958, 172). Literature now suggests that the American work ethic is changing significantly. In Work, Mobility and Participation (Cole 1979) postulates that the American work ethic has been weakened by rising levels of affluence and union control. David J. Cherrington (1980), in reporting on a survey conducted in 1976, concludes that members of today's workforce, especially young workers, do not have the same attitudes as previous generations towards the importance of work, pride, and craftsmanship.1 Another survey, conducted by 1 Conducted under the Erteszek Working Paper Series, Graduate School of Management, Brigham Young University, 191 questions were asked about workers' attitudes towards their jobs, their company, their community, and work in general. Over 3,000 responses were received. The most desirable work -related outcome was "feeling pride and craftsmanship in your work." The second most desirable was "getting more money or a larger pay increase." Compared with older workers, younger workers felt that pride in craftsmanship and joy in being of service were less desirable, having leisure and free time were more desirable, and doing a poor job more acceptable. Leisure and free time were less important to women and that females endorsed pride and diligence slightly more than males. 3

PAGE 14

the Gallup organization during September, 1979, seemed to indicate that if an individual were given a choice to work or not to work nearly 50 percent of those responding would prefer not to work.2 Therefore, although this paper does not deal with attitudes of people or specifically address the American work ethic, it will be necessary to consider these values in arriving at conclusions concerning the application of alternative work schedules. Social research in the 1930s placed emphasis upon the human relations approach to organization and management, but another body of research begun in the 2 Leonard A. Wood, Executive Vice President, the Gallup Organization, Inc., refers to this survey in an article entitled "Changing Attitudes and the Work Ethic," The Work Ethic in Business, ed. W. Michael Hoffman and Thomas J. Wiley. Cambridge: Oelgeschlager, Gunn and Hain, 1981. The question was asked "If you had your choice, and money andjor child care were not a problem, which of the following would you prefer to do: 'I wouldn't work, I'd work part time, I'd work full time?" Only one man in 10, now currently employed, chose not to work at all. Three in ten would switch to part-time positions, while a majority (61 percent) would continue working full-time if given the choice. A somewhat different pattern emerged among women. For women working full-time, part-time employment is seen as the ideal 40 percent would choose working on a part-time basis. Fourteen percent would prefer not to work, and 42 percent would continue working full-time if given a choice. 4

PAGE 15

1940s and 1950s contributed a somewhat different perspective of the worker's place in the work situation. This approach, known as organizational humanism, formed a bridge between the human relations approach and what is now referred to as modern organizational theory. It was postulated that, in addition to previous assumptions of the human relations approach, management can, and should, promote positive motivation through delegation of responsibility, permitting discretion and creativity on the job.3 The report of a special task force to the Secretary of Health, Education, and Welfare also addresses the changing social character of today's generation and concludes: ... the very high personal and social costs of unsatisfying work should be avoided through the redesign of work .... (W.E. Upjohn Institute 1981, 94). Redesigning work to more closely fit the needs of the worker significantly differs from job engineering, as presented by Frederick W. Taylor. The job engineering 3 Much of this has been attributed to the work of Frederick w. Taylor and his concepts regarding "scientific management," i.e., the one best way of accomplishing work. Taylor's philosophy was, however, much deeper than that. It was Taylor's thought that the maximum good for all society can only come about through the positive cooperation of labor and management. In other words, there must be harmonious collaboration in order to achieve a common objective. 5

PAGE 16

approach concentrates on maximizing the efficiency of workers and quite often leads to the creation of fractionalized, boring jobs. Redesign of work to provide for job enrichment began in the 1940s at IBM (Walker 1950). Progress was initially rather slow, but more attention and activity was eventually directed toward the development of motivational factors in job assignments when the intuitively appealing theories of Herzberg, et al. (1959) and Maslow (1954) were published and began to be generally accepted. The average workweek for non-agriculture wage and salary workers decreased from 65.7 hours per week in 1850 to 38.7 hours per week in 1987.4 There are many reasons, that need not be discussed here, for this substantive change, but it is quite clear that the commitments and obligations that evolve from social, family, and recreational interests have tended to increase while the formal working week was correspondingly being reduced. Somewhat similar to work redesign, in respect to achievement of employee morale benefits, is work rescheduling. Use of flexible schedules 4 1850-1930 workweek statistics obtained from J. Frederick Dewhurst and associates, America's Needs and Resources: A New Survey. New York: Twentieth Century Fund, 1955. More recent data was published in the Statistical Abstract of the United States, 1989. u.s. Department of Commerce, Bureau of the Census. 6

PAGE 17

permits employees to cope better with nonwork impingements on work (e.g., child-care, schooling needs, commuting, personal fitness schedules, etc.). Job sharing, flexitime, and currently, telecommuting, are examples of such programs. German economist Christel Kraemerer originated the concept of flexitime in the 1960s as an inducement to bring housewives and mothers into the work force to alleviate the labor shortage in Germany. In 1967, the West German aerospace firm of Messerschrnidt-Bokow-Blohn, introduced the concept of "gleitzert" or gliding time as a means of reducing traffic congestion in the area near their operations. It was based upon the assumption that, in many work situations, rigid starting and stopping times are unnecessary and could be adapted into a more flexible system. Later the concept was modified, the idea of a core time was added, and the system was adopted by other European countries. Although accurate data has not been supplied, it has been estimated that about a third of the Swiss labor force had been under flexitime, as well as substantial numbers of white-collar and other workers in Germany, England, France and Scandinavia. By 1972, flexitime was being used in the United States at Control Data Corporation, and at other firms on a trial basis. 7

PAGE 18

Information with respect to the proclaimed advantages of imaginative job design proposals such as staggered hours, flexitour, flexitime and group flexibility (generically referred to as flexitime, compressed work week and its variations) quickly spread and numerous experiments-and adaptations of alternative work schedules occured. Articles in professional journals and other publications described the basic principles of the system and announced in enthusiastic terms the varied results, but more often the successes, of the alternative work schedule experiments. Many of the early flexitime examinations were primarily descriptive in nature, or lacked the scientific rigor desired. Four studies did investigate flexitime in an experimental fashion: (1) Evans and Partridge (1973), who examined the impact of flexitime on non-supervisory employees in a large insurance company over a five month period, (2) Golembiewski, et al. (1975), who examined, over an extended period, pretest and posttest data results from an experiment at a research organization, (3) Schein, et al. (1977), who focused on the impact of flexitime on productivity, and Harrick, et al. (1986), who conducted field studies on the effect of alternate work schedules on productivity, leave usage, and employee attitudes. Control groups were not utilized, however. 8

PAGE 19

The Congress of the United States found that new trends in the usage of the four-day workweeks, flexible work hours, and other variations in workday and workweek schedules in the private sector seemed to show sufficient promise to warrant" carefully designed, controlled, and evaluated experimentation by Federal agencies over a 3-year period to determine whether and in what situations such varied work schedules can be successfully used by Federal agencies on a permanent basis .... (1978 92 Stat. 755) The Public Law that authorized this experiment was to last until September 29, 1981. The Law was later amended to extend it until July 4, 1982, and it was again extended for three more years, until July 23, 1985. Permanent authority was subsequently granted on December 23, 1985 (99 Stat. 1350). Based upon responses to a General Accounting Office questionnaire (U.S. Comptroller General B-216021 1985), an estimated 489,000 permanent Federal employees were working under an alternate work schedule arrangement in 1985. The General Accounting Office now estimates that 37 percent of all federal employees are currently working under some kind of alternate work schedule, or approximately 1,145,000 employees. Although initial reaction to the adoption of alternate work schedules was positive with respect to employee morale, mixed reactions 9

PAGE 20

were received by the General Accounting Office regarding: (1) service to the public, (2) efficiency of government operations, (3) mass transit, (4) energy consumption, (5) employment opportunities, and (6) work productivity. The National Park Service (NPS) is a bureau within the Department of Interior. In 1989 it had 12,737 permanent employees, which made it Interior's largest bureau from the standpoint of employment. The NPS benefits from a large and active constituency; its supporters have pressed for independent agency status, and the NPS is consistently viewed more favorably by more Americans than any other Federal agency. 5 This popularity is further supported by the growth of public visitation to the National Parks. According to the National Park Service, total visitation grew from 451,000 in 1918 (Albright 1985, 103), to 351,911,180 in 1989.6 Because of its long history of providing quality public service, its highly seasonal operation, and variety of occupational specialties, the application and 5 The Roper Poll, interpreted by Adweek's Marketing Week in its January 8, 1990 issue, shows the National Park Service with an 82 percent favorable rating, ahead of second place FBI at 77 percent, and NASA at 76 percent. For at least the lOth straight year, this poll has placed the National Park Service as the top agency in Government. 6 U.S. Department of the Interior, National Park Service, Division of Statistics. 10

PAGE 21

approval of alternate work schedules in the NPS was approached rather cautiously. Legislation permitting Federal agencies to experiment with alternate work schedules was enacted on September 29, 1978, but it was not until October 23, 1979, that the NPS distributed any information regarding the availability of these experimental work schedule programs. Initial instructions to supervisors suggested that only a very limited number of "experimental" programs should be initiated, and that those employees whose work involved program management and direct public contact work could not be included. Each Regional Director was delegated the authority to establish alternate schedules for employees, but they were to be terminated if there were any evidence of reduced productivity, a lowered level of public service, or an increase in agency costs. Informal surveys, conducted approximately six months after implementation, seemed to indicate that those employees who were under an experimental work schedule were well satisfied with the arrangement; but those employees who were working under a traditional schedule expressed considerable reservations over the use of the alternate work schedules. Adverse comments, primarily from managers and supervisors, contradicted favorable comments, i.e., energy consumption has increased, 11

PAGE 22

productivity has declined; carpools have been disrupted, etc. Since the earlier informal survey, no further review of this program has taken place. Another alternate work schedule, or arrangement, is that of the flexible workplace. Referred to as "telecommuting" or "flexiplace," this program officially began implementation on February 12, 1990, with the first workshop for the agency coordinators. Telecommuting creates an "office without walls," allowing people to work at home or at satellite offices, to "commute" using computers and telephone lines instead of cars. One of the obvious challenges of work at home is merging work and home life successfully. At this time 31 agencies and bureaus have taken the first steps toward establishing work-at-home programs by designating coordinators to direct flexiplace arrangements within their organizations. Guidelines have been produced, and it is anticipated that further procedural refinement will occur as the program evolves. The Department of Interior has indicated that they will support only 50 experimental flexiplace arrangements, and these will be only in the headquarters office (with a limit of five employees from each bureau). One of the more pervasive concerns is the issue of productivity, or the potential lack thereof, that is making agency 12

PAGE 23

managers quite skeptical. A lesson can be learned from earlier experiences in adopting flexitime; flexiplace (telecommuting) should not be implemented without careful thought about its impact on operations andjor agency mission. Although other factors may influence the success or failure of an alternate workplace arrangement, there is no question that public sector productivity is a continuing strategic issue and concern. It would therefore be appropriate to question whether any employee benefit program should be introduced if its subsequent implementation would have a deleterious effect upon employee productivity, effectiveness, or efficiency. Alternate work schedules are being increasingly accepted and adopted almost as if they were an extension of an employee rights, and they currently are part of union contract negotiations as other agreed upon conditions and circumstances of employment. In July 1985, President Reagan asked Congress to pass a joint resolution establishing productivity as a national goal and to enact management legislation to improve government management tools. Can we identify the flexible work schedule as being one of those tools? 13

PAGE 24

CHAPTER II THE PROBLEM Statement of the Problem The purpose of this study is to investigate and evaluate the effect and impact of the use of alternate work schedules upon employee productivity in the National Park Service, a bureau of the Department of the Interior. At the present time, alternate work schedules within the National Park Service are being approved without sufficient research or investigation as to the effect that approval of such schedules would have upon the specific work activity involved. Based upon certain encouragement from the Director's office on October 12, 1984, and as reinforced in the Alternative Work Schedules Handbook of 1985, use of alternative work schedules is supported within the National Park Service as long as ... approved schedules do not interfere with the mission of the organization .... "1 In addition to its broad mission and purpose ... to conserve the scenery and the 1 Memorandum from the Director of the National Park Service to the Directorate, Field Directorate, WASO Office and Division Chiefs, and all Superintendents, October 12, 1984. Draft of Alternate Work Schedules Handbook, from Assistant Director, Personnel and Administrative Services, March 18, 1985. p.8.

PAGE 25

natural and historic objects therein .... "2 the National Park Service, along with all other Federal agencies, is subject to Executive Order 12552, February 25, 1986, which seeks a 20 percent productivity increase through ... higher quality, improved timeliness, and lower cost .... "3 Most National Park Service employees prefer alternative work schedules (Johnson et al. 1984; Johnson and Salvi 1985). 4 An alternative work schedule is any work schedule which differs from what would be considered to be the normal, accepted, customary, work schedule for the organization. If the normal work schedule or tour required an employee to work onsite 8:00 AM to 5:00 PM, Monday through Friday, any variance of this work schedule would be considered an alternative work schedule. Thus, compressed work weeks, flexitime, part-time employment, shift work, or any other work arrangement that differs from that of the normal schedule work be considered an 2 An Act To establish a National Park Service, and for other purposes approved August 25, 1916. 39 Stat. 535. U.S. Code, Title 16. 3 Memorandum from President Reagan to heads of executive departments and agencies, July 31, 1985. 4 In these two studies, 71% of all permanent, and 58% of all temporary employees indicated that they worked a traditional 40 hour, five day week, but, only 36% of the permanent and 43% of the temporary employees actually preferred the traditional schedule. 15

PAGE 26

alternative work schedule. A variety of flexible and compressed schedules are found in all National Park Service management offices and most park areas; no record of the number of such schedules approved are maintained, and, in most instances, no effort to record work time performed outside of the normal work hours is attempted. Although supervisors have the authority to disapprove requests for alternative work schedules if it appears that such approval would adversely impact public service or productivity, it is questionable if they would do so because of employee pressure and lack of management support. No effort is made in the National Park Service to routinely measure productivity, and therefore, the subsequent effect of alternate work schedules upon productivity is largely unknown. Significant differences exist between regions, offices, and park areas, on the approach to approval of compressed work schedules. Although nearly all organizational units permit some form of flexible work schedules, i.e., varying the beginning and ending work hours and adhering to accepted "core" work hours, compressed work hour schedules have not been universally accepted within the National Park Service. The principal organizational units within the National Park Service are the individual park areas themselves, the 10 regional 16

PAGE 27

offices under which all park areas are located, the Denver Service Center (construction, design, and planning office), the Harpers Ferry Interpretive Planning Center, and the bureau headquarters office in Washington D.C .. From a total payroll of 17,923 National Park Service employees on April 25, 1990, 1,725 were working under a compressed work schedule, and this varied by geographic regions from a low of 10 (Mid-Atlantic Region, 1,498 employees), to a high of 608 (Western Region, 2,777 employees). Thus, it would appear that management application andjor acceptance of a compressed work schedule is not consistent from bureau to bureau. Interest in a flexiplace arrangement has been expressed by quite a few employees, but National Park Service management has not as yet addressed this issue. Hypotheses, or Research Questions A number of hypotheses can be formulated with respect to research design and the problem statement. Hypotheses will be tested through field study, but it is perhaps important to state at this point that although much of the research will be ex post facto, certain elements will be subject to direct intervention andjor manipulation because of the organizational position and status of the researcher. 17

PAGE 28

The basic thrust of the first hypothesis will be to determine whether or not compressed work schedules have any effect upon productivity in the National Park Service. Previous research findings and theories suggest that compressed work schedules have but little effect on productivity. Pretest-posttest comparisons involving experimental and control groups will be structured to respond to the following theorems: Null: There is no significant difference (p >.05) in the levels of productivity between experimental (compressed work schedules) and control groups for the pretest-posttest time periods. Research: Productivity measurements for experimental (compressed work hour employees) and H1 control groups indicate that employees not under a compressed work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. A second hypothesis is associated with the use of alternative eight-hour work schedules. These are fixed schedules which do not vary from day to day. The arrival and departure times are according to a set schedule known to both the employer and the supervisor in advance. The schedule includes ten (10) work days in each pay period. Each work day is eight (8) hours in length excluding the scheduled lunch period. This schedule differs from the normal eight-hour schedule in that the scheduled arrival 18

PAGE 29

and departure times need not coincide with the traditional eight-hour schedule and the employee may schedule a lunch period longer than the customary or minimum lunch period. Null: There is no significant difference (p >.05) in the levels of productivity between experimental (alternative eight-hour work schedules) and control groups for the pretest-posttest time periods. Research: Productivity measurements for experimental (flexible work schedule employees) and H2 control groups indicate that employees not under a compressed work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. Each of the above hypotheses would be tested for both immediate and longterm comparison and effect. The logic of separate testing is to determine whether or not productivity enhancements, or losses, have been influenced by employee temporal response to change (Hawthorne). It has been frequently suggested that application of flexible schedules reduces absenteeism, which would adversely affect productivity (Rosow 1981; Craddock, Lewis and Rose 1981; Coltrin and Barendse 1981). In the Federal Government this could be equally applicable to unscheduled and brief use of annual leave as well as employee use of sick leave. With the exception of the work of Harrick, et al. (1986), involving an office of a 19

PAGE 30

federal government agency, this supposition has not been fully tested as yet with respect to federal employees, and, more specifically, with National Park Service employees;therefore, a third and a fourth hypothesis would involve the use of annual leave for brief but unscheduled periods, and the use of sick leave. Null: There is no significant difference (p >.05) in the use of annual leave by employees for brief unscheduled periods between experimental and control groups for the pretest-posttest time period. Research: Use of annual leave for brief unscheduled periods is, on an average, less for H3 employees in experimental groups than for those in control groups during the pretest -posttest time period. Null: There is no significant difference (p >.05) in the use of sick leave by employees between experimental and control groups for the pretest-posttest time period. Research: Use of sick leave for brief unscheduled periods is, on an average, less for H4 employees in experimental groups than for those in control groups during the pretestposttest time period. As was the case with the testing of the first two hypotheses, a long posttest comparison will also be implemented to determine more precisely if there is a relationship between the use of leave by employees and alternative work schedules. Since it can be argued that compressed work schedules may have a more deleterious effect upon absenteeism, alternative eight-hour work 20

PAGE 31

schedules and compressed schedules will be tested separately. Another human factor which can contribute to productivity decline is high employee turnover. It has been generally claimed that since flexitime generally increases employee satisfaction, turnover is typically lower (Coltrin and Barendse 1981; Welch and Gordon 1980; Rubin 1979; U.S. Office of Management and Budget 1973). In response to a specific question in The National Park Service Employee Survey of 1983, regarding their leaving the National Park Service, 937 employees (n=3452) indicated that they would look for a new job outside of the National Park Service during the next year. Lack of career advancement opportunities and training were the most frequently mentioned reasons. No measurable answers in excess of 1 percent of the total responding were attributable to work schedules. This would seem to suggest that work schedules are not a significant reason that would necessarily influence employee turnover. A fifth hypothesis would therefore relate to actual turnover, and whether or not the ability to utilize a flexible work schedule reduced the turnover rate. Comparisons would be made between regions andjor offices and parks that maintained either conservative or liberal 21

PAGE 32

policies with respect to the approval of flexible work schedules. Null: There is no significant difference (p >.05) with respect to the rate of employee turnover between offices andjor parks who offer liberal work schedule opportunities (experimental), and those who offer primarily only traditional work schedules (control group). Research: Employment turnover is lower with respect to offices andjor parks who offer liberal H5 work schedule opportunities (experimental), than it is for offices andjor parks who offer primarily only traditional work schedules (control group). A sixth and final hypothesis that would be tested is related to employee morale and performance. Literature dealing with alternative work schedules and productivity is near-unanimous in its association of morale improvement and the adoption of alternative work schedules. A questionnaire using a matrix question format and Likert response categories would be distributed to a random sampling (sufficient to achieve 95 percent accuracy) of all employees previously identified as working under an alternative work schedule. The questionnaire would be also randomly distributed to other employees not working under an alternate schedule. In addition, a sampling of the annual performance evaluation rating provided to those employees working under an alternate work schedule would be compared to a 22

PAGE 33

proportionate sampling of those employees not working under an alternate work schedule. The same level of certainty (95 percent) would dictate the sample size. Null: There is no significant difference (p >.05) between the level of employee work satisfaction (morale) and hisjher individual performance evaluation for those employees working under an alternative work schedule (experimental group), and those who work under more traditional work schedules (control group). Research: Employees who work under an alternative work schedule (experimental group) tend to H6 have higher morale, and, on an average, higher performance evaluations than those employees who work under more traditional work schedules (control group). Delimitations and Limitations of the Study Much of the information that will be assembled is subject to those limitations inherent in ex post facto field studies. The independent variable, flexitime, was not assigned by the researcher to an individual in a random fashion, nor were the selection of work hours. In some instances the individual may have requested a specific compressed work schedule and was instead assigned by hisjher organization to another but more limited work schedule. The same circumstance exists with respect to all control group work schedules. These were also selected by the respective organizational management officials, as were those individuals who were considered 23

PAGE 34

by management to be not eligible, by virtue of their position or status, for a flexible work schedule. In the course of the research effort, certain delimiting factors, such as sample size or regional selection, may also exist and require fuller examination and discussion in the context of their potential effect upon the study and eventual research conclusions. Limitations Each National Park Service organization unit has its own distinct culture and personality. Geographic and ecological differences can explain some of the variations, but the probable principal causes are staffing requirements and needs that are influenced by organizational purpose and mission, functional favoritism (i.e., management preference and priority to ranger activities and requirements over those related to maintenance, research, or administration), and leadership. Because of these cultural differences, effective leaders in one park area do not necessarily become effective leaders in another park area andjor office. This cultural difference is partially identified in Table 1 below where approval and use of other than flexible or traditional work schedules (compressed 4/10 24

PAGE 35

and 5/4/9) are dependent upon established regional and park policies. Table 2.1 Number of National Park Service Employees and Authorized Compressed Work Schedules As of April 25, 1990 Region/Office No. of Authorized Percent Employees Compressed w /S Authorized Alaska 415 35 8.43 Mid-Atlantic 1,498 10 .07 Midwest 1,158 122 10.54 North-Atlantic 1,407 95 6.75 National Capital 2,554 183 7.16 Pacific Northwest 704 184 26.14 Rocky Mountain 2,065 112 5.42 Southeast 2,421 78 3.22 Southwest 1,210 58 4.79 Western 2,777 608 21.89 DenverSC 576 99 17.19 Harpers Ferry 249 3 1.20 Washington Office 889 138 15.52 17,923 1,725 9.62 Another limiting factor will be the amount of reliance that will of necessity be placed upon secondary data sources. For example, the information supplied in Table 2.1, was acquired from Department of Interior personnel/payroll records, which are the result of reporting by a variety of National Park Service personnel offices. In the event that an office elects not to report the approval of an alternative work schedule, then 25

PAGE 36

that particular schedule would not be statistically represented. It is proposed that selective testing be applied to assure validity and reliability to bureauwide reported figures. Because the author does not have access to all of the NPS areas and offices, the sampling frame for this study is limited and therefore a simple random sample of the target population cannot be drawn. This would seemingly also be true with respect to the personal characteristics of target groups, as differences in race and sex become more apparent between rural and metropolitan park areas and offices. The possibility of stratified sampling, where and when available, will be used to reduce the aspect of sampling error. Since no previous contacts with National Park Service employees have occurred, with the exception of the National Park Service Employee Survey of 1983, it is felt that no other known factors which may affect generalization andjor external validity should bear upon, or significantly influence, this research. Limitations on theoretical relationships between worker productivity, sick leave usage, unprogrammed absences, turnover, and performance evaluations also exist. Although there are limited opportunities to measure productivity from a quantitative perspective, 26

PAGE 37

most productivity measurements are qualitative in nature and, thus subject to a certain amount of ambiguity. Of necessity, considerable reliance has been placed upon bureau operational efficiency and quality of public service. Evaluation of either organizational characteristic is, however, rather subjective. Previous literature does suggest that relationships do exist between flexible work schedules, reductions in absenteeism, turnover, improved performance and employee morale, as well as with other variables, but the extent of relationship andjor the possibility of another unrelated factor coexisting, and perhaps influencing a positivejnegative relationship has not been conclusively determined. Delimitations The researcher is the Associate Regional Director for Administration in one of the major National Park Service regional offices. Quite obviously, considerable information will be readily available with respect to this region and the parks within this region. In addition, opportunities for intervention are inimitably better if implemented within the home region, which may potentially present a selection bias. It has been earlier suggested that National Park Service organizational culture has a decided influence 27

PAGE 38

over the approval, establishment of policies applicable to, and effective encouragement and perhaps ultimate success of, alternative work schedules. Although many of the more traditional National Park Service managers profess their support of the AWS program, and also acknowledge the existence of forceful guidance by the Department and OPM in this regard, an unspoken but implied message is received by the lower supervisory levels in the organization that they are expected to take a conservative approach in their implementation of this particular program. However, even without this tacit message, the same bias appears quite evident within most, if not all, supervisory levels with respect to the use of compressed work schedules. The sole exception would be where direct program advantage or benefit obviously occurs. An example of this would be in the case of field trail maintenance people who spend eight 10-hour days in a row in the backcountry of a park area and then have six days off duty. This results in reduced travel time, less associated perdiem costs, and improved work programming. Definition of Terms A variety of words or expressions is being used in a precise sense in the proposal, since the proposed research depends upon specific definition of terms, the 28

PAGE 39

following definitions, and definitional sources if available, are being provided: alternative work schedule, an alternative to the standard 5-day 40 hour workweek, a rearrangement of workhours. Janice Neipert Hedges. Special Flexitime Reports, 1977. A substitute to the standard five-day, 40-hour workweek. May be either a change in the number of days worked, a change in the number of hours worked, or a change in both days and hours. John W. Newstom and Jon L. Pierce, Alternative Work Schedules: The State of the Art, 1979. bandwidth, total number of hours in the interval between the earliest possible workhour starting time and the latest finishing time. (All other things being equal, the larger the bandwidth, the greater the possible flexibility to the employee) Robert T. Golembiewski and Carl w. Proehl, Jr., A Survey of the Empirical Literature on Flexible Workhours: Character and Consequences of a Major Innovation, 1978. basic workweek, number of hours in a normal workweek, as established by collective bargaining agreements or statutory law. Premium payments must usually be paid for time worked in excess of the basic workweek. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. Section 6101 of title 5, United States Code requires each agency head to establish a basic administrative workweek of 40 hours. climate, a set of properties in the work environment specific to a particular organization, that may be assessed by the way the organization deals with its employees and its societal and task environments. Andrew D. Szilagyi and Marc J. Wallace, Organizational Behavior and Performance, 1987. compressed schedule, in the case of a full-time employee, an 80-hour biweekly basic work requirement which is scheduled for less than 10 workdays; and in the case of a part-time employee, a biweekly basic work requirement of less than 80 hours which is scheduled for less than 10 workdays. P.L. 95-390, 9/29/78. core time, designated hours during which an employee must be present for work. P.L. 97-221, 7/23/82. 29

PAGE 40

credit hours, any hours within a flexible schedule which are in excess of the number of hours in an employee's basic work requirement and which the employee elects to work so as to vary the length of a workweek or workday. P.L. 97-221, 7/23/82. culture, a pattern of basic assumptions -invented, discovered, or developed by a given group as it learns to cope with its problems of external adaption and internal integration -that has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems. Edgar H. Schein, Organizational Culture and Leadership, 1985. effectiveness, the extent to which an organization accomplishes some predetermined goal or objective; the overall performance of an organization from the viewpoint of some strategic constituency. Jay M. Shafritz, Facts on File Dictionary, 1986. efficiency, the promotion of administrative methods that will produce the largest store of results for a given objective at the least cost; the reduction of material and personnel costs while maximizing precision, speed, and simplicity in administration. Jay M. Shafritz, Facts on File Dictionary, 1986. 5/4/9, term given to a compressed work schedule which calls for nine work days, rather than ten, within a given two-week work period. The five work days of the first work week are nine hours each, as are three days of the second work week, and the fourth day is of eight hours duration. flexible work hours, designated hours during which an employee may elect the time of arrival at and departure from work, for the purpose of varying arrival and departure times, or if and to the extent permitted, for the purposes of accumulating credit hours. P.L. 97-221, 7/23/82. flexiplace, an arrangement which permits employees to work at home or at other approved sites away from the office for all or part of the work week. Flexiplace, flexible work place, work-at-home, telecommuting, and teleworking all refer to paid employment away from the traditional office. President's Council on Management 30

PAGE 41

Improvement. Guidelines for Pilot Flexible Workplace Arrangements, 1990. flexitime, a way of scheduling work so that an employee must be present during certain "core" hours of the day, but can begin and end work any time before and after these core hours (some organizations may establish limits as to when flexitime hours may begin or must end; see bandwidth). Don Hellriegel and John W. Slocum, Jr., Organizational Behavior, 1979. flexitour, a flexitime work schedule which requires employees to choose, for a relatively long period of time, starting and quitting times for a standard eighthour shift. Jerome M. Rosow and Robert Zager, New Work Schedules for a Changing Society, 1981. Employee will preselect starting time; may modify schedule with prior notification and approval of supervisor. P.L. 99-196, 12/23/85. four/ten and/or four-day workweek, a regular full-time workweek of 40 hours is worked in four 10-hour days instead of five eight-hour days. Stanley D. Nallen, What is happening to flexitime, flexitour, gliding time, the variable day? And permanent part-time employment? And the four-day week? 1980. gliding time, a flexitime work schedule which allows daily variations in starting and quitting times but keeps a rigid eight-hour shift. Jerome M. Rosow and Robert Zager, New Work Schedules for a Changing Society, 1981. job design, also called job redesign. Organizing work in a manner consistent with societal goals, demonstrating that the social and psychological basis of work is as significant to long-term productivity and efficiency as are the traditional physiological factors. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. A systematic process of determining, by observations or study, the major work-related behaviors, responsibilities, skills, and experiences necessary to satisfactorily perform a task. Gilbert B. Siegel and Robert C. Myrtle, Public Personnel Administration, 1985. maxiflex, a flexitime work schedule which allows daily variations with no core hours required, making it easy to compress the workweek into three or four days. Jerome M. Rosow and Robert Zager, New Work Schedules for a Changing Society, 1981. For Federal Government employees, an 31

PAGE 42

employee may vary the length of the work week and workday as long as he or she is present for core time which is scheduled on less than all five weekdays; must work or account for the basic work requirement, e.g. 80 hours in a bi-weekly pay period;credit hour carry-over is limited to a maximum of 10 hours. P.L. 99-196, 12/23/85. morale, moral or mental condition with respect to courage, discipline, confidence, enthusiasm, willingness to endure hardship, etc., within a group, in relation to a group, or within an individual. Webster's New World Dictionary, Third College Edition, 1988. Collective attitude of the workforce toward their work environment and a crude measure of the organizational climate. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. motivation, an emotional stimulus that causes a person to act, an amalgam of all the factors in one's working environment that foster (positively or negatively) productive efforts. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. A predisposition to act in a specific goal-directed manner. Don Hellriegel and John W. Slocum, Jr., Organizational Behavior, 1979. organization, any unified, consolidated group of elements; systematized whole; especially, a) a body of persons organized for a specific purpose, as a club, union, or society, b) the administrative personnel or executive structure of a business, c) all the functionaries, committees, etc. of a political party. Webster's New World Dictionary, Third College Edition, 1988. Any structure and process of allocating jobs so that common objectives may be achieved. Jay M. Shafritz, Facts on File Dictionary, 1985. A system of consciously coordinated personal activities or forces of two or more persons. Chester I. Barnard, The Functions of the Executive, 1938. organizational humanism, a set of organizational theories stressing that work held intrinsic interest for the worker, that workers sought satisfaction in their work, that they wanted to work rather than avoid it, and that they could be motivated through systems of positive incentives (such as participation in decision making). George J. Gordon, Public Administration in America, 1982. 32

PAGE 43

performance, the act of performing; execution, accomplishment, fulfillment, etc., operation or functioning, usually with regard to effectiveness, as of a machine, something done or performed; deed or feat. Webster's New World Dictionary, Third College Edition, 1988. Demonstration of a skill or competence. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. pretest-posttest, an evaluation of circumstances, capabilities or conditions (dependent variable) prior to, and after an event (independent variable), for the purposes of validating the intervention and measuring effect on the dependent variable. Earl Babbie, The Practice of Social Research, 1983. productivity, the efficiency with which resources are used to produce a government service or product at specified levels of quality and timeliness. E.O. 12552, 2/25/86. qualitative, research data framed in the form of words rather than numbers, collected in a variety of ways (observation, interviews, extracts from documents, tape recordings) and are usually "processed" somewhat before they are ready for use. Matthew B. Miles and A. Michael Huberman, Quality Data Analysis, 1985. Research techniques linked to participant observation, content analysis, formal and informal interviewing, clinical case studies, life-history construction, videotaping behavioral displays, archival data surveys, historical analysis. The invention and use of varied unobtrusive measures, and various formal schools of thought and procedure in social science such as dramaturgic analysis, semiotics, frame analysis, ethno-methodology, and conventional analysis. John Van Maanen, Quality Methodology, 1983. quality-of-work-life, a series of organizational interventions, activities, designed to improve the work place for employees. Andrew D. Szilagle and Marc J. Wallace, Organizational Behavior and Performance, 1987. quantitative, having to do with quantity, capable of being measured. Webster's New World Dictionary, Third College Edition, 1988. The numerical representation and manipulation of observations for the purpose of describing and explaining the phenomena that those 33

PAGE 44

observations reflect. Earl Babbie, The Practice of Social Research, 1983. relationship, the quality or state of being related; connection. Webster's New World Dictionary, Third College Edition, 1988. Interrelationships among several persons considered simultaneously. Earl Babbie, The Practice of Social Research, 1983. reliability, that quality of measurement method that suggests that the same data would have been collected each time in repeated observations of the same phenomenon. Earl Babbie, The Practice of Social Research, 1983. research, a careful, systematic, patient study and investigation in some field of knowledge, undertaken to discover or establish facts or principles. Webster's New World Dictionary, Third College Edition, 1988. A systematic and intensive study for a fuller knowledge of the subject studied. Basic research attempts to uncover new scientific knowledge and understanding with little regard as to when, or specifically how, the new facts will be used. Applied research is conducted with a special purpose in mind; it is directed toward a specific problem, or toward a series of problems, that stand in the way of progress in a particular area. Encyclopedic Dictionary of Systems and Procedures, 1966. satisfaction, anything that brings gratification, pleasure, or contentment. Webster's New World Dictionary, Third College Edition, 1988. People's feelings about the rewards they have received, a consequence of past events. the willingness of individuals to continue as employees of an organization and to show up for work on a regular basis, not necessarily associated with motivation or increases in work productivity. Edward E. Lawler, III, High Involvement Management, 1986. Worker contentment brought about by the presence of conditions that meet certain physical and psychological needs. Abraham H. Maslow, Motivation and Personality, 1954. scientific management, closed model theory, improvements to organizational efficiency and economy for the sake of increased production; time-motion studies; roan-as-machine conception, primarily associated with the work of 34

PAGE 45

Frederick W. Taylor. Nicholas Henry, Public Administration and Public Affairs, 1980. secondary data sources, information collected by others and archived in some form. The term secondary information is frequently used to refer to secondary data (the raw data obtained in various studies) and secondary sources (the published summaries of these data). David W. Stewart, Secondary Research, 1984. significant difference, through statistical testing it can be determined that the size of differences found between groups are not due to chance. The probability of rejecting a null hypothesis when it is true. By establishing a significance level of .01, only outcomes that have a probability equal to or less than that will allow a rejection of the null hypothesis. Susan Welch and John C. Comer, Quantitative Methods for Public Administration, 1983. In the context of tests of statistical significance, the degree of likelihood than an observed, empirical relationship could be attributable to sampling error. A relationship is significant at the .05 level if the likelihood of its being only a function of sampling error is no greater than 5 out of 100. Earl Babbie, The Practice of Social Research, 1983. telecommuting, implied the use of high-tech telecommunications and computers to perform work from remote locations. Work-at-home, flexiplace, covers work regardless of high-tech or low-tech applications. President's Council on Management Improvement. Guidelines for Pilot Flexible Workplace Arrangements, 1990. turnover, movement of individuals into, through, and out of an organization. The total number (or percentage) of separations that occurs over a given time period. Jay M. Shafritz, Dictionary of Personnel Management and Labor Relations, 1980. variable day, a flexitime work schedule which allows an employee to work a longer or shorter time on any given day as long as the total hours worked adds up correctly at the end of a week, bi-weekly, or monthly period dependent on agreed upon arrangements. Jerome M. Rosow and Robert Zager, New Work Schedules for a Changing Society, 1981. Employee may vary the length of the workday as long as he or she is present for daily core time with limits established by the organization; must 35

PAGE 46

work or account for the basic work requirement, e.g. 40 hours per week for a full time employee; credit hour carry-over between pay periods is limited to a maximum of 10 hours. P.L. 99-196, 12/23/85. work schedule, a work arrangement which identifies the hours of work that an employee will perform within a given pay period, subject to not exceeding 80 hours in the designated two week pay period. Section 6101, 5 USC. Summary Research on the effect of alternative work schedules on individual performance or productivity in the Federal Government has, to this point, been somewhat limited.5 It is also noted that much of this research is based upon soft information andjor is restricted in certain scientific approaches, including pretest and posttest evaluations. Those studies that did offer suitable 5 D. Hauser, The Impact of Flexitime on Organizational Functioning: A Field Study. Washington, D.C.: U.S. Office of Personnel Management, June, 1980. C. Orpen, "Effects of flexible working hours on employee satisfaction and performance: A field experiment," Journal of Applied Psychology, 66, (1981):113-115. R.T. Golembiewski and C. W. Proehl, Jr., "Public Sector Applications of Flexible Workhours: A Review of Available Experience," Public Administration Review, 40, n. 1, (January /February 1980): 72-85. Oscar Mueller and Muriel Cole,"Concept wins converts at Federal agency," Monthly Labor Review, 100, n. 2, (February, 1977):71-74. John J. Macut, "Measuring Productivity Under a 4-Day Week," Monthly Labor Review, 97, n. 4, (April, 1974): 55-56. Edward J. Harrick, Gene R. Vanek, and Joseph F. Michlitsch, "Alternate Work Schedules, Productivity, Leave Usage, and Employee Attitudes: A Field Study," Public Personnel Management, 15, n. 2, (Summer, 1986): 159-169. 36

PAGE 47

levels of result validity, with the possible exception of the Harrick, Vanek, and Michlitsch (1986) study, suffered from small sample size. A variety of alternative work schedules are now being used in the National Park Service; however, alternative schedules are not uniformly considered by parks or offices within the National Park Service, and their use is primarily a matter of local discretion andjor managerial support. After approval has been granted, it would appear that no formal process now exists for review or continued monitoring of these approved schedules. Available literature suggests that alternative; flexible work schedules can lead to improved employee work satisfaction, higher morale, decreased absenteeism, and reduction in employee turnover. Claims were initially made that alternative work schedules also brought about increased productivity, but the literature now suggests that no change in productivity will occur as a result of the adoption of alternative work schedules. This would seem to be consistent with the work of a number of behavioralists, researchers, and academicians. Edward Lawler, for example, rather forcefully points out that "the simplistic belief that satisfaction leads to productivity is wrong. Thus, even though improvements in the workplace increase satisfaction, increases in 37

PAGE 48

productivity and organizational effectiveness may not occur" (Lawler 1986, 135). Except for the collection of limited survey data, no research with respect to the above conclusions, or that applicable to the effect of alternative work schedules on leave use or personal performance, has ever been conducted within the National Park Service. There have been some very significant changes in the practice of management, organizational behavior and theory, and general personnel processes, in the past 50 years. Although many of these changes and concepts originated in the private sector as a result of competitive practices, their adoption in public sector has been quite rapid, and the public sector has also been responsible for the introduction of several innovative pay plans and employee benefit programs. Tom Burns maintained that organizations were the simultaneous workings of at least three social systems: (1) formal authority, (2) cooperative, and (3) political (Burns and Stalker 1961). It was his emphasis on the organization as a social system that is the most useful basis for further discussion of this problem. With this in mind, it is suggested that although satisfaction and motivation are not necessarily mutually exclusive, they do have some very discernible differences. 38

PAGE 49

When considering the organization as a social system, and each individual as a part of this system, the individual's needs therefore become a matter of organizational and managerial importance. Expectancy theory (Vroom 1964) suggests that all employees do not have the same need level, but rather that policies must of necessity consider the needs of all employees and permit individual choices wherever possible. Flexible work hours are rapidly becoming, if they have not already become, a employee benefit andjor discretionary option. Quite frequently the workplace is used to satisfy needs other than those related to the individual's economic condition, self-esteem andjor creative fulfillment. In this manner the workplace occasionally serves as an outlet for, and an extension of, individual expression. Accordingly, employees motivated in this manner will often seek from the workplace social accommodation for many of their personal needs. Project ACHIEVE, for example, was introduced by the Department of Interior initially to better emphasize the needs of the handicapped employee, and secondly to encourage and accelerate federal government measures productivity improvement; ... carrying out programs with the minimum number of employees and getting a dollar's worth for a dollar spent .... 39

PAGE 50

CHAPTER III REVIEW OF THE LITERATURE Theoretical Review Since the advent of alternate work schedules in private and public institutions, a variety of experiments have taken place with an equally diverse number of outcomes. Since our emphasis here is on the effect of these schedules on productivity, literature applicable to the study andjor effect of such work schedules was reviewed. Although numerous articles, of necessity, include observations of the possible or potential causal effect of alternate work schedules upon productivity, only six (not including those studies conducted by the u.s. Office of Personnel Management), directly assess this phenomenon.1 Not surprisingly, the literature tends to identify that there is a definite weakness in the data supplied since so much is "soft" or largely based upon opinion rather than more reliable "hard" data. Even the reports prepared by OPM to the President and Congress (U.S. Office of Personnel Management, Interim and Final Reports 1981 and 1982) were suspect by the General Accounting Office (U.S. Comptroller General 1980) because of weaknesses in testing, GAO questioned whether the prevailing methods of analysis used would provide 1 Harrick, Vanek, and Michlitsch (1986), Kim and Campagna (1981), Narayanan and Nath (1982), Orpen (1981), Rainey and Wolf (1981), and Schein, Maurer, and Novak (1977).

PAGE 51

Congress the information it needed to decide the extent to which variable work hours should be allowed in the Federal work force. A study that was conducted by the American Management Association, through Goodmeasure, Inc. (1985), attempted to identify the similarities and differences among all of the important work innovations that have been introduced in America since the 1960s. Goodmeasure selected 21 of these innovations that they considered to be major work alternatives, and they further categorized and placed them into four groups, or clusters, that seemed to share common features; (1) job-related work alternatives, (2) quality of working life alternatives, (3) alternative organizational structures, and, (4) employee participation and control. The study utilized data obtained from survey responses from a random sample of 10,000 American Management Association Members (n=1,618). The results, by industry distribution, indicated that from those alternative work arrangements applicable to flexible schedules or to a flexible workplace, government was the major user. The public sector also used more work innovations, on an average, than did the private sector. The concept of encouraging worker participation was popular in the 1930s, on the theory that an involved 41

PAGE 52

employee is subsequently more satisfied with hisjher position and thus is inclined to be more productive. Within the past two decades, this approach has been broadened and given greater emphasis with the development of a variety of employment policies and practices covering work scheduling ("Flexitime," Dun's Review 1977). There is extensive literature which covers the quality of work life movement and the use of alternative work schedules, but that which specifically addresses the effect of alternative work schedules upon productivity, particularly within the Federal Government, is either rather limited or only superficially deals with the matter of productivity impact. One particular study, however, by Harrick, Vanek, and Michlitsch (1986), did use a pretestjposttest design and a longitudinal frame covering 18 months of productivity and leave usage time, and provided for 16 months between measures of worker satisfaction. This study took place in an office of a federal government agency, and reported study results indicating that there were some significant reductions in leave usage, no change overall in employee satisfaction, and mixed results with respect to productivity. Federal legislation that authorized the three-year Federal Government experimental program with compressed and 42

PAGE 53

flexible work schedules also required that each experiment be closely monitored and provide for evaluative reporting. This literature review begins with an historical summary of the major events and Federal legislation which led to the establishment and growth of the alternative work schedule. Some definitions of the principal schedules adopted by industry and government are then reviewed with emphasis upon announced advantages andjor disadvantages. Although this review highlights in general basic literature devoted to the effect of alternative work schedules; the findings of some of the more significant studies of the impact of alternative work schedules on productivity, employee morale, and other performance factors, are illustrated by table 3.1. Table 3.1 Findings of Studies/Research Experiments Effects of Alternative Work Schedules Study Effect upon Productivity Effect Upon Morale Other Effects Improved NC Decreased Improved NC Decreased Favorable UnFavorable Craddock, Lewis & Rose (1981) Dunham, Pierce, & Castaneda (1987) Evans & Partridge (1973) 42.8% 52.4% 4.8% Studies indicated that effectiveness/productivity changed very little. Not measured. 69.7% 30.3% 0.0% Anticipated reactions were confirmed. No significant enhancements noted. Some improvement, but little noticeable change over time. 43 Use of Leave

PAGE 54

Tab 1 e 3. 1. ( contd. ) Study Effect upon Productivity lnproved NC Decreased Finkle (1979) Golembiewski, Hilles & Kagno (1974) Golembiewski, Yeager & Hilles. (1975) Harrick, Vanek & Michlitsch (1986) Hausser, D. (1980) Ivancevich & Lyon (1977) Kim & Campagna (1981) Narayanan & Nath (1982) National Center for Productivity and Quality of Working Life (1977) Or pen (1981) Not measured. Initial increase, then maintain previous levels. Productivity was reported as being increased. Mixed results. Of six tasks observed, productivity increased in three and decreased in three. No significant effect. Increased productivity for 1st 12 months, but had no effect after 24 months. Mixed reactions on performance, but generally positive. No change in reported productivity. Six activities were measured, all initially showed improvement, and then three declined. Negligible effect on performance or productivity. Effect Upon Morale lnproved NC Decreased Generally positive. Significant increases were noted. Some supervisory concerns, however. Improved. But very small sample (n=11). Except for satisfaction with an AWS, employee work satisfaction and/or morale did not change. No significant effect. Improved, but not to a measurable level No significant effect. No significant effect. Improved. Attitudes of employees were enhanced but not supervisors. Improved job satisfaction. Other Effects Favorable UnFavorable Reduction Increases in leave, in cost/ absenteeism, overhead. and tardiness. Commuting, quality of 1 ife. Both sick and annual leave use went down. Convenience for employees. Job satisfaction satisfaction and stress. Reduced short term absenteeism. Absenteeism increased. Pierce & Newstrom (1982) No reported improvement in No significant relationship Reduced Rainey & Wolf (1981) Ralston, Anthony & Gustafson (1985) productivity but performance with job satisfaction. absenteeism. was improved. Experimental and control groups experienced reduction in quantity and improvement in quality. 24% productivity improvement when shared resources were a factor. No change when shared resources were not a pre-existing factor. Employee satisfaction with job and working conditions showed improvement. Not measured. 44 Short term gain versus long term alienation. Productivity can be expanded without increasing the capital outlay.

PAGE 55

Table 3.1.1 (contd.) Study Effect upon Productivity IJ!I)roved NC Decreased Schein, Maurer No clear-cut conclusions & Novak with regard to the impact (1977) of AWS on productivity. U.S. Office of Personnel Management (1981) U.S. Office of Personnel Management (1982) 46% 46% 8% Reported an overall increase of 43% in quality, 51% in quantity, and 44% in product timeliness. 30% 60% 10% From approximately 325,000 participating employees, it was determined that a more favorable rather than an unfavorable productivity impact occurred. Effect Upon Morale Other Effects Improved NC Decreased Favorable UnFavorable Suggested positive effect. Not specifically measured, but overall responses were generally positive. Morale of participating employees improved, but morale of those who had to "take up the slack," decreased. Managerial and supervisory moral unfavorably impacted. Savings in energy, travel. Reduction in leave usage. Experiments in 85 units dropped. Employee travel costs reduced slightly. Numerous problems for management. Historical Analysis In an attempt to seek a better way to make use of existing, but antiquated, public transportation system, an experiment took place in Metz, France in 1955. Public schools and all employees adopted a citywide staggered start system. Because of the success that was achieved from these alternating starting times, the concept of staggering starting times was adopted by a number of European cities. During the 1960s a number of German industrial firms initiated employee programs that permitted the use of "flexible" work schedules and "gliding time" as an inducement to expand the existing 45

PAGE 56

workforce and reduce traffic congestion. Although these proved to be very popular and effective programs, the needs of management to respond to periods of critical work load led to the establishment by these firms of certain "core" hours when all employees were required to be present. By 1972, it was estimated that about a third of the Swiss labor force, as well as substantial numbers of white collar and other workers in Germany, England, France and Scandinavia were working under flexible work schedules. Flexitime appeared in the United States in the early 1970s, introduced by a number of firms that had international affiliations By 1973 there were approximately 30 large U.S. firms and 200 smaller organizations permitting certain employees the opportunity to select, within agreed upon limits, their working hours (Hedges 1973; Newstrom and Pierce 1979; Zawacki and Johnson 1976). A variety of reports and informal accounts offered enthusiastic encouragement, and a number of early examinations and experiments involving flexitime also attested to the advantages of nearly all imaginative job design programs that were undertaken. However, many of these early flexitime examinations were descriptive andjor anecdotal, and lacked the degree of scientific rigor that is normally desired (Bolton 1971; 46

PAGE 57

Wade 1973; Narayanan and Nath 1982)). There were three early studies which did investigate flexitime in an experimental fashion (Evans and Partridge 1973; Golembiewski, Yeager, and Hilles 1975; Schein, Maurer, and Novak 1977), and each these of studies noted improvement in employee satisfaction and detected only minor changes in productivity. In 1978, the Congress of the United States found that these new trends in the usage of four-day workweeks, flexible work hours, and other variations in workday and workweek schedules in the private sector demonstrated sufficient promise, and that controlled experimentation by Federal Agencies over a three-year period was therefore warranted. With a number of extensions, experimentation continued until 1985, at which time authority for use of alternative work schedules was permanently granted. The background statement contained in the legislative history of P.L. 97-221, indicated that considerable weight was given to the data contained in the Interim and Final Report to the President and the Congress regarding the alternate work schedules' experimental program. The Interim (1981) and Final (1982) reports prepared by the Office of Personnel Management asserted that the 47

PAGE 58

program, in most cases, was beneficial to the public and to the employees themselves. However, the reports also indicated that there were some notable examples of failures. Experiments with alternative work schedules in the non-federal public sector are now receiving only limited attention and, therefore, most of the literature that is available is primarily descriptive, dated, or skewed in favor of the alternative schedules. Christensen's (1989) survey of senior human resource executives, for example, was reported to be not representative of U.S. businesses or even of very large corporations, and was considered by Christensen to be possibly skewed in favor of firms that provide flexible scheduling and staffing alternatives -since it was more probable that firms providing flexible schedules would have participated in the survey (surveys were mailed to 2,775 executives, but only 521 responded). Another survey, involving 308 of a total of 500 participating managers, conducted by the American Management Society Foundation (1989), reported that the major advantage of flextime was "improved employee morale" and about half of these respondents also stated that flextime also decreased tardiness and absenteeism. This same AMSF survey noted that there are problems. One of the respondent managers stated on page 2 of this 48

PAGE 59

survey that: "We feel flextime, rather than improving productivity, has reduced it because of lack of supervision during all hours of work. Also, there are distractions caused by co-workers coming to and leaving work at different times." In his analysis of flexitime in the public sector, Rubin (1979 280) noted that" flexitime has been implemented in the city government of Inglewood, California, with positive results .... A March 1990 telephone call to the city administrator of Inglewood indicated that flexitime was discontinued, and since it happened so long ago, only vague impressions of the reasons for this could be offered. This situation seems to have occurred with respect to experiments in both the public sector and the non-public sector, and, although by itself it is not remarkable, it does suggest that most of the literature now available, particularly as it involves the non-federal public sector, may suffer from lack of currentness. City government and counties have become involved in experiments in areawide flexitime to either reduce pollution or improve rush hour traffic problems of other such associated reasons. Rosow and Zager (1983) have identified efforts in several large cities, and in the combined area of Seattle and King County, Washington, to persuade employers to adopt flexitime for their 49

PAGE 60

employees. State government experiments were highlighted by McKann (1973) with respect to certain studies on Florida experiments, and there is a fairly comprehensive review of available experiences in a Golembiewski and Proehl article (1980) which summarizes public sector applications of flexible workhours. Golembiewski and Proehl note that although public sector flexitime applications have the same pattern of positive vs. negative effects as business applications (32 public sector studies were reviewed), they have some concern about the number and quality of the public sector studies. A field study which took place in an office of a federal government agency (Harrick, Vanek, and Michlitsch (1986), used a pretestjposttest design and a longitudinal frame covering 18 months of productivity time, had not as yet been conducted. Of particular interest is the paper prepared by Kim and Campagna (1981) on the effects of flexitime on employee attendance and performance at a county welfare office. This study attempts to justify the argument that flexitime has a positive impact on attendance and performance among public sector employees, but it also notes the potential for the "Hawthorne" (Mayo 1933) 50

PAGE 61

andjor the "Rosenthal effect," 2 since the employees were certainly aware of the experiment and were not "blind" to what was going on. In June 1977, the Kentucky Department of Personnel began a ten-week trial of flexitime (Craddock, Lewis, and Rose, 1981). This study dealt mainly with attitudes and the satisfaction of the employees themselves with the experiment. The effects of flexitime upon productivity were measured only through the comments of the employees involved and how they felt about their performance; if it improved, if they only "held their own," or they perhaps fell behind. Since these data are highly subjective, the study is significant for review purposes solely as a measure of employee attitude and interests in improving family life. The literature does not offer much insight with respect to flexitime experiences in city or municipal goverment. One study commented upon the indirect effect of recruitment for a large general hospital upon a small private hospital, and its subsequent adoption of flexitime to resolve a staffing problem with respect to nurses, was published by M. Arni, Director of Nurses at 2 (Rosenthal 1976) Rosenthal suggests that if experimenters can affect the behavior of those with whom they interact, than others also may affect behaviors based upon their expectations of subject behavior. 51

PAGE 62

Lakeland Manor Hospital (1983). This article provided some penetration as to how flexible schedules of work can be used to attract and retain skilled employees. Union involvement, with respect to alternative work schedules, in the available literature was surprisingly light. For Federal employees, this could possibly be because the Federal Services Impasses Panel had previously held that federal agencies had to meet a statutory burden of proof in rejecting union requests for flexitime workweeks on grounds of decreased productivity, diminished service to the public or increased costs.3 However, by passage of P.L. 99-196 (99 Stat. 1350 1985), collective bargaining rights to flexitime with respect to union representation have now been fully ensured. Alternative Work Schedules -Definitions and Usage Since the classic Western Electric-Hawthorne study in 1927, management has become increasingly aware of the important effects of group attitude and climate on work performance, and of the capacity of every person for work at whatever level of proficiency they are capable of achieving. Regardless of what the acceptance may be of these known behavioral characteristics, one of the 3 Federal Services Impasses Panel, Case No. 83FSIP64, 5 August, 1983. 52

PAGE 63

traditional assumptions about how organizations are best run is that everyone should come to work at the same time and leave at the same time. This assumption is consistent with the idea that standardization is important and that everyone should be treated in the same way. It is inconsistent, however, with the reality that people have different preferences about when they want to come to work and when they want to leave work. It also ignores the fact that people find themselves in different family situations, that transportation is not equally available for everyone, and that there are a number of disadvantages in having everyone arrive simultaneously, namely, overcrowding in elevators, etc., transportation problems, and so on. There are, of course, certain undeniable advantages to regular work hours from a supervisory and managerial perspective, and the literature produced to date not only describes the varied alternative work schedules available or being currently used, but also identifies the advantages and disadvantages of each. Although there are many variations of alternative work schedules discussed in available literature, two principal alternatives to the predominant 5-day, 9 to 5, 40-hour week have been essayed andjor adopted, the compressed schedule of a four-day work week and 53

PAGE 64

''flexitime." Flexible workplace arrangements will be discussed later. There are several kinds of flexitime programs that are and have been followed. The most limited of these is flexitour, under which employees choose a starting and quitting hour that they follow for a specified period of time, such as a week or a month. Another model is gliding time, in which a standard eighthour day is worked, but each employee can vary his or her own starting and quitting times daily. The variable day system is the most flexible where more or less eight hours are worked in a day, but debit and credit hours can be accumulated as long as they balance at the end of the pay period. In other instances, a more restrictive flexitime schedule is adopted which provides for an employee to select his own tour, but must stay on this schedule unless approval to change has been received. This could mean a selected starting and departure time for a eight hour day, 5-day week. Flexitime Since Messerschmidt-Bokow-Blohn introduced the concept of gliding time as a means of reducing traffic congestion in 1967, the system has come to be called flexitime. The Dictionary of Personnel Management and Labor Relations (Shafritz 1980) provides the following definition on pages 113-114: 54

PAGE 65

-flexitime, flexible work schedule in which workers can, within a prescribed band of time in the morning and again in the afternoon, start and finish work at their discretion as long as they complete the total number of hours required for a given period, usually a month. That is, the workday can vary from day to day in its length as well as in the time that it begins and ends. The morning and evening bands of time often are designated as "quiet time." Telephone calls and staff meetings are confined to "core time," which generally runs from midmorning to midafternoon. Time clocks or other mechanical controls for keeping track of the hours worked usually are a part of flexitime systems. Business Week in a 1972 article which commented briefly on flexitime, stated that" Flexitime is shorthand for flexible working hours .... Elbing, et al. (1974, 29), covered all Alternative Work Schedules as "flexible workinghour plans." Hedges (1977 62) noted that" Schedules that compressed a full 40-hour workweek into 4, or even 3, days dominated the early innovations. But before the mid-1970s, a different type of schedule-flexitime-gained prominence .... The ability to arrange schedules to fit individual and joint needs was addressed by Swart (1974), Evans and Partridge (1973), and Baum and Young (1973), and the term "flexitime" appeared in public administration and personnel text books shortly thereafter. Buchele (1977) describes "Flexitime and the Four-Day Week" as" an arrangement that allows workers a couple 55

PAGE 66

of hours latitude in starting and stopping work each day without reducing the total time worked per week .... Shafritz, Hyde, and Rosenbloom (1981) introduces flexitime in a chapter on job design, quoting Donahue's (1975) definition of flex time as it is applicable to New York office employees and how it differs from the federal designation of "flexitime." This clearly points out how pure flex time permits employee selection of the hours of reporting and departure as opposed to the federal process of mutually agreeing to a starting and stopping time that may differ from others in the work unit but cannot be changed without approval. The advantages of flexitime, specifically its effect on the quality of both work and family life, have been enthusiastically described by a number of individuals. Bernard's (1979) rather positive article, as well as the work of Newstrom and Pierce (1979), describe the potentials for the system but caution that careful planning is important prior to any implementation. Nollen (1980) suggested that flexitime will extend to 25 to 30 percent of all workers by the end of this decade since (1) firms will receive economic gains that endure -at a cheap price, (2) flexitime is broad-gauged; it can and is being used in many different work settings, and (3) flexitime's track record in Europe has already been 56

PAGE 67

established. More balanced reports are provided by Rubin (1979), Owen (1977), and Wheat (1982). Rubin and Owen both point out the critical need for planning, communications and supervision with respect to "core" time. Core time, the hours during which all employees must be present (Finkle 1979), along with "bandwidth" (the period within which all hours must be worked) may possibly present a scheduling problem for some individuals if management is too restrictive in its designations. Wheat recounts past successes, including recruitment enhancements, but decries the control mechanisms and the abandonment of some otherwise useful irregular tours. A representative cross section of the literature suggests that alternative work schedules can potentially offer a wide array of advantages, but the AWS system also has disadvantages. Reported results from a variety of scientific experiments and operational experiences range from no change to some degree of either a positive or negative effect. With the exception of job attitudes and leisure-time satisfaction, which reportedly are improved by alternate work schedules, the predicted consequences of work schedule experiments, with respect to productivity are that of no significant change. Based upon existing theory and the observations from 57

PAGE 68

a number of experiments, a variety of positive effects and outcomes can be obtained if organizational implementation of an alternate work schedule program has been appropriately designed and managed. It has been claimed that flexible work schedules can reduce tardiness and absenteeism, reduce overtime, increase employee morale, increase productivity, reduce turnover, improve planning and communication, reduce commuter congestion, increase customer service, broaden work opportunities, and achieve better utilization of recreation and service facilities. Although it may be possible to achieve some of these organizational attributes, it would be highly unlikely that they could all occur through any single experiment. The themes of the articles by Coltrin (1981), Hedges (1977), and Rosow (1983), bring this out very clearly. In the Coltrin article it is mentioned that while most research seems to demonstrate the existence of favorable results, few, if any, system failures have been studied. This same theme is expressed by Hedges, who cautions that not every environment offers the same prospect for success; in fact, she notes that a few establishments have abandoned flexitime as unworkable. 58

PAGE 69

Rosow reports that flexitime fails about five percent of the time, compressed workweeks about 25 percent. Consequently, although there are many positive claims that would assuredly seem to be significant recommendations for adoption of alternative work schedules, these also have been somewhat balanced by certain noted disadvantages. The principal disadvantage that has been expressed is that supervisors and managers are not available or present at all times lends to the fear that employees will take advantage of the situation. This problem was noted within the literature in most instances, and was probably the largest concern of management cited in research material.4 This conclusion was rather vividly 4 Simcha Ronen, "Arrival and Departure Patterns of Public Sector Employees Before and After Implementation of Flexitime," Personnel Psychology, 34, n. 4, (1981). G. W. Rainey, Jr., L. Wolf, "Flex-Time: Short-Term Benefits; Long-Term ... ?" Public Administration Review, 41, n. 1, (JanjFeb, 1981). G.W. Rainey, Jr., L. Wolf, "The Organizationally Dysfunctional Consequences of Flexible Work Hours: A General Overview," Public Personnel Management, 11, n. 2, (1982). J. N. Hedges, "Flexible schedules:problems and issues," Monthly Labor Review, 100, n. 2, (February, 1977). S. A. Coltrin, B. D. Barendse, "Is Your Organization a Good Candidate for Flexitime?" Personnel Journal, 60, n. 9 (September, 1981). J. S. Kim, A. F. Campagna, "Effects of Flexitime on Employee Attendance and Performance: A Field Experiment," Academy of Management Journal, 24, n. 4 (1981). J. Welch, D. Gordon, "Assessing The Impact of Flexitime on Productivity," Business Horizons, 23, n. 6, (December, 1980). R. Rubin, "Flexitime: Its Implementation in the Public Sector," Public 59

PAGE 70

supported in the review of productivity, leave usage and employee satisfaction in the National Park Service's Rocky Mountain Regional Office. One divisional unit that had but limited supervision was permitted almost infinite flexibility in its work schedule and eventually the employees and supervisors became almost contemptuous of their responsibilities. As a subsequent result, productivity and morale in the division significantly deteriorated in less than three years. It was suggested that flexitime creates extra costs for lighting, heating and air conditioning. A study of 300 European countries conducted by Management Centre Europe (1976) indicated that cost was their largest complaint with respect to flexible tour experiments, followed by the unavailability of key personnel and loss of organizational control. Limited applicability sometimes exists and, consequently, flexitime may not lend itself to every situation (Coltrin and Barendse 1981; Rainey and Wolf 1981). The possibility that overtime may be reduced under a flexible work schedule may be seen as a disadvantage by certain employees. Union concerns about this have been noted by Hedges (1977), Rubin (1979), and Owen (1977). Disadvantages Administration Review, 39, n. 3, (MayjJune, 1979). 60

PAGE 71

associated with flexitime, from an organizational perspective, seem to revolve around control and coordination as well as the possibility of additional cost, whereas employee concerns seem to focus on possible increased work-related transportation and loss of overtime. Narayanan and Nath (1982) report on a research experiment which suggests that contrary to previous research (Evans and Partridge 1973), and other theoretical positions which view flexitime as job enrichment (Gibson, Ivancevich, and Donnelly, 1979), employees reported no change in attitude with respect to satisfaction with work. Nollen and Martin (1978), however, suggest that job satisfaction may more closely follow experiments with flexitime when sound planning and efforts to restructure positions were initiated first. Compressed Schedule For purposes of this literature review, a compressed work schedule will mean the compression of a full work schedule, whether it is based upon a 40-hour week or a 80-hour bi-weekly, into fewer days than accustomed or normal. An example of this would be four 10-hour days. Anything less than the customary number of work hours in the work period would be part-time employment. There is a tendency to equate flexitime with such systems as the compressed work week, such as the 5/4-9 or the four 1061

PAGE 72

hour days. But they are distinctly different. Rainey and Wolf (1982) in discussing unanticipated consequences of experiments with alternative work schedules, note that the tendency to confuse the two different systems leads to certain negative impressions regarding the "pure" flexitime, whereby an employee selects hisjher own hours of work (exclusive of core time) and compressed time schedules, including the 5/4-9, which retain fixed work shifts, removing any semblance of personal autonomy in reporting and attendance. The Dictionary of Personnel Management and Labor Relations (Shafritz, 1980) provides the following definition of a four-day workweek: reallocation of the standard 40-hour workweek over four days instead of five. By lengthening the workday, employees get a 3-day weekend every week with no loss of pay. This concept differs from the 4-dayj 32-hour workweek that some union leaders advocate. Unfortunately, there is no attempt at a comprehensive definition for compressed workweek or compressed tour in the Dictionary. Newstrom and Pierce (1979) refer to the "4/40" plan which dated back to the 1940s with both the Gulf and Mobil Oil Companies and achieved some popularity during the early 1970s. This plan required employees to work four 10-hour days, with a three day weekend. The "eight-day week" is also mentioned as an innovation, but unlike the 4/40, the worker then has a four-day break rather than a three day before 62

PAGE 73

starting the cycle again. Atwood (1979) provides several examples of both techniques along with use of part time workers in assessing work schedules and their contributions to productivity. An excellent definition of compressed work schedules can be found in the U.S. Office of Personnel Management's Interim Report to the President and the Congress (1981): ... Any work schedule which enables a full-time employee to complete the basic work requirement of 80 hours, in less than 10 full workdays, in each biweekly pay period .... In Appendix C of this report, the initial use of compressed schedules by the private sector in the United States is noted: ... Unlike the flexible schedule, major activity with the compressed schedule has been in the United States where it also originated. During the 1960s some business leaders in the American Management Association studied the feasibility of reducing the number of days rather than hours worked per week. Out of this effort came the compressed schedule or 4-day workweek .... Use of the compressed schedule had expanded significantly between 1970 and 1973. In the previously cited Interim Report to the President and the Congress, it was stated that the number of companies offering a four-day compressed schedule rose from approximately 70 companies in 1970 to nearly 7,000 in 1973. This report also provided some additional support to attest to the remarkable growth of compressed schedules: By 1978 63

PAGE 74

an estimated 10,000 private sector companies with over a million workers were using compressed schedules, while approximately 7 million additional private sector employees were reported using flexible schedules ... (p. C-1). The general acceptance and use of this schedule was also noted by Elbing, Gadon, and Gordon (1974) who indicated that, according to the April 30, 1973, Wall Street Journal and other sources, some 3,000 U.S. Companies have adopted the four-day week, compared to an estimated 40 in 1970. Hedges (1981) provided a graphic presentation which suggested that in 1973 approximately 1.75 percent of all nonfarm workers were working under a compressed workweek. The 1989 AMS Flexible Work Survey, conducted by the American Management Society Foundation, reflected only a 1 percent scheduling acceptance for compressed tours from 1985 through 1988. This was increased to 2 percent in 1989, but it was stated that: "The four-day workweek shows little acceptance." Nallen (1980) in referring to the four-day workweek commented 11 Four-day and other compressed workweeks started in this country about 1970 (they were never much used in Europe) and spurted to 1.3 million workers, or 2.2 percent of the work force, by 1975 .... 11 These figures would somewhat tend to support Hedge's figures for 1973, but then Nallen goes on to 64

PAGE 75

state ... Most of the Federal government work force in Denver went to a four-day workweek in July, 1979 .... However, in commenting upon the "Denver Experiment," in its Interim Report to the President and the Congress, the Office of Personnel Management states that" Those on compressed represent approximately 20 percent of Denver's Federal workforce .... This would therefore suggest that of those agencies electing to participate in the experiment, 12,560 employees were involved and of this total, approximately 44 percent (5528) were on a compressed schedule. This differs substantially from the level of interest reported by Nallen and is reflective of the frequent disparity in statistical inferences supplied in the literature. The same situation can be detected with respect to data supplied regarding flexible work schedules. Schein (1977 464) states that ... The increasing popularity of flexible working hours and the paucity of research on organizational outcomes heightens the need for further study .... Golembiewski and Proehl (1980) also indicated that flexitime research may be criticized on technical grounds because of some rather obvious shortcomings, and he goes on to identify a variety of design deficiencies, i.e., use of post-test only, insufficient comparison groups, faulty statistical treatment, etc. Nallen (1979 18) identifies three 65

PAGE 76

specific deficiencies: (1) only successful flexitime programs are reported in the literature; (2) reports from organizations may be biased upward as positive results were expected and to fulfill these expectations only positive results were found; (3) the scientific quality of productive measurements is questionable. Bernard (1979 56) in his article also pointed out the vagueness of the statistical and research evidence supplied and they raise questions as to the quality and number of studies -particularly those involving the public sector. A particularly informative summary of the use of flexitime and compressed work schedules was produced by the Work in America Institute (1981). The summary was directed by Jerome M. Rosow and Robert Zager (1983), and it noted that flexitime options were split almost evenly among flexitour, variable day, and gliding time. Total growth in the use of flexitime had increased from 4.7 percent of all employees in 1974 to 11.9 percent in 1980. Use of compressed schedules, however, had only slightly increased, from 1.9 percent to 2.7 percent. Most literature on compressed work schedules tends to support the contention that longer daily work hours and shorter workweeks increase the time people have for family activities, recreation or educational purposes, reduce commuting costs, and give some employees the 66

PAGE 77

opportunity to take on a second job. However, the longer work days required under the compressed workweek could create the possibility of fatigue and reduced employee effectiveness toward the end of the day and thereby reduce the time available to the employees to engage in personal and domestic pursuits during most of the days in each week (Rainey and Wolf 1982). Ronald Lang, who was with the Canadian Labor Congress, was quoted in Business Week as rejecting compressed work weeks because they "go beyond the bounds of health and safety standards" by exceeding the eight-hour limit that unions fought to establish ("Europe Likes Flexi-Time Work" 1972). Olson and Brief (1978), however, point out that proponents of the four-day workweek suggest that such a work schedule not only reduces employee absenteeism, it increases job satisfaction and productivity, gives employees more time to "unwind" from their job as well as decreasing employees' requests for time off for personal matters. Nollen (1980) cautions that although the compressed workweek is not applicable in every setting, it can help businesses that have costly start-ups and shutdowns, inadequately utilized capital equipment, and travel time to work sites. The experiment by the Federal Government which took 67

PAGE 78

place as a result of P.L. 95-390, more commonly referred to as the Federal Employees Flexible and Compressed Work Schedules Act of 1978, attempted to provide a summary of results which were taken from a sample of 70 work units selected by OPM to be a part of a longitudinal crosssectional study. Work units were not required to submit the narrative evaluation report to OPM, and since this narrative non-technical assessment represented the work unit's total evaluation of the program, hard data andjor statistical evidence was not available. In addition, much of the data used to form the statistical profile was later provided by the work units themselves. Since these work units actively sought and preferred alternative work schedules, the findings of the study are probably biased. OPM provided on page 23 of this report that: According to the report, "The purpose of the profile was to enable OPM to collect a standardized set of quantitative information summarizing experimenting organizations' experiences with AWS ... "(p. 23). It would appear, however, that much of the data was collected in a posttest fashion without expected scientific rigor. It is also noted that OPM had taken a rather positive approach to this national experiment, and quite possibly elected not to devote much review or comment to unsuccessful experiments. Thus, OPM's analysis of results of 68

PAGE 79

experiments with compressed tours seem only to reveal positive comments -as was the case with flexible work schedules. On the other hand, individual agency reports (p. 125) offer a conflicting picture. Agencies reported that under the 4-10 plan 55 percent of their participating work units experienced success, 35 percent of the work units experienced only partial success, and 10 percent experienced total failure. Under the 5/4/9 plan, 62 percent experience success, 28 percent experienced partial success and 10 percent experienced failure. Flexible Workplace Arrangements Although flexible workplace arrangements are not a part of this study, recognition for their potential influence on work schedules has been noted in this study since the actual location of the work performed would be shifted away from what is now the primary, and the traditional, work site. Thus, telecommuting could eventually impact residual work schedules at the traditional work site. The idea of "telecommuting" reportedly began nearly 20 years ago with Jack Nilles, an engineer who wanted to reduce auto pollution in Los Angeles. He proposed 69

PAGE 80

setting up satellite offices in the suburbs near the homes of clerical workers. Instead of driving downtown, clerks would use computer terminals from these satellite offices. In a six-month experiment in 1973, Nilles claimed that turnover dropped to zero from 33 percent, office rental costs were reduced, and productivity rose 18 percent (Nilles, et al. 1976). Although telecommuting is still rather new, an estimated 13 to 15 million people now work at home at least part of the time (Atkinson 1985). This accelerated interest in telecommuting stems not only from its being a popular option from an employee standpoint, but also from its mitigating effect on peakhour traffic, work travel, increased energy consumption, additional office space, and environmental concerns. Basically, telecommuting allows people to work part or full time at home-based or satellite offices, instead of commuting every day to work. It can also somewhat be compared to an updated version of the age-old "cottage industry" concept (Edwards and Edwards 1985). Like other alternative work options, the process offers a wide array of advantages and disadvantages. One major problem that telecommuting does present, which is not found in other alternative work options, is the difficulty that the worker will face in attempting to successfully merge work and home life. When one is working at home, there will 70

PAGE 81

be a number of distractions and diversions normal to a domestic environment that is not found in the typical office setting. How the worker effectively deals with this problem is unquestionably the major issue impacting telecommuting. Gil E. Gordon and Marcia M. Kelly, who are regular contributors to the Telecommuting Review (The Gordon Report), rather thoroughly covered most advantages and disadvantages in Telecommuting: How to Make it Work for You and Your Company (1986), while also stressing that telecommuting is basically a social phenomenon. Thus, most of those human/emotional responses and circumstances related to change become equally applicable with respect to telecommuting implementation. The term telecommuting additionally refers to the process of electronically transferring the results of job-related work from a home, or satellite site, to the principal office. It is questionable that such extreme decentralization would have been possible without the availability of the current vast array of sophisticated new communications and computerized equipment. It therefore becomes possible for firms, such as the J.C. Penney Company, to relocate from large cities to small towns like Plano, Texas (Lewis 1988). Thomas B. Cross (1986) has suggested that telecommuting is now in the same position today that data processing was in 25 years 71

PAGE 82

ago, i.e., executives do not grasp the significance of the new technology and thereby cost their organizations a competitive edge by losing time in adopting it. There is also, in all probability, more managerial resistance to the basic concept of telecommuting than there had been earlier during the initial consideration processes of flexitime and compressed time. Trust certainly is important, but many managers equate employee performance with such non-merit factors as punctuality, sociability, and appearance rather than with quality and quantity of work produced (Downing-Faircloth 1982). Unlike flexible work schedules, flexiplace would appear to have aroused in the government a more active and earlier interest than did flexible work schedules. In addition to the work of Jack Nilles for the Department of General Services, State of California (Martin 1989), a telecommuting pilot project was initiated by the South Coast Air Quality Management District in El Monte, California on October 7, 1988, involving 30 employees; and the County of Los Angles is currently moving forward with a pilot project that could involve up to 2,000 of the county's 17,000 employees who work in downtown Los Angeles (Gordon 1989). Other similar projects have been initiated, or the telecommuting process is being carefully considered, by other county and city 72

PAGE 83

governments. The city of Ft. Collins, Colorado, for example, has recently adopted a policy regarding telecommuting which very clearly points out that the principal goal of the program is improvement in productivity. This aspect of proposed flexiplace application is unique to the degree that most proposals heretofore have stressed either cost reduction or employee retention. A governmentwide project is now underway which will permit Federal employees to participate in flexible workplace arrangements. Sponsored by The President's Council on Management Improvement (PCMI), this project will test the feasibility and utility of flexible workplace arrangements, on a limited basis, through pilot tests scheduled to begin in early 1990. The project codirectors are Thomas Cowley, General Services Administration, and Dr. Wendell Joice, Office of Personnel Management. Dr. Joice had earlier been responsible for the Office of Personnel Management's evaluation of the flexible and compressed work schedules project in the Federal Government. Guidelines have been issued (President's Council on Management Improvement 1990), and most Federal agencies are participating in the project. It will be up to individual agencies to decide how many and the type of employees that may participate. 73

PAGE 84

The Office of Personnel Management is publishing a bimonthly newsletter, entitled Flexiplace Focus for the benefit of participants in the Federal flexible workplace pilot study, and Federal agencies are also associated with certain pilot projects being conducted by state agencies.5 Productivity The overall objective of behavioral science is to find out how to increase the productivity of organizations through optimal use of their human resources. The amount of literature available which touches upon human productivity is enormous, and it is not the intention of this review to attempt to cover all of the which has been published with respect to productivity, but rather to focus upon that which addresses the effect of alternate work schedules upon productivity. In comparing techniques of applying Human Resource Management, Teasley (1983 363) measures and defines 5 In addition to very visible support in the State of Colorado, the Federal Government is actively involved in telecommunications programs in other states, such as Washington, and Colorado. The recent Colorado Symposium on Telecommuting, was presented in conjunction with the Governor's Office of Policy and Research and the U.S. Department of Energy. 74

PAGE 85

productivity as ... the ratio of an organizational output to an input (P=O/I) .... Hatry (1982) suggests that although that may be the most common definition for productivity, to apply it to any particular government service is a complex task and subject to controversy. He further offers on page 427: ... Productivity measurement generally has been defined in the public sector as encompassing both efficiency and effectiveness. Efficiency indicates the extent to which the government produces a given output with the least possible use of resources. Effectiveness indicates the amount of the end product, the real service to the public, that the government is providing .... The Office of Personnel Management (1978), defines productivity as: "the sum of the efficiency, effectiveness, quality and responsiveness with which products and services are delivered. . ( p. 4) In a special study conducted by the u.s. Congress (Congressional Budget Office, 1987), labor productivity of the sampled federal civilian workforce was reported to have increased at an annual rate of 1.4 percent from 1977 through 1986. It should be noted here that in the federal government, productivity is measured in two different ways. The first method measures determinate labor quantities against the amount of services provided. The second method, which was initiated in 1986, uses a more comprehensive measure that considers labor along 75

PAGE 86

with capital and other resources involved in production. An analysis of productivity measurement in the federal government (U.S. Office of Management and Budget 1964), provided a more simplistic explanation of how the federal government measures productivity improvement; ... carrying out programs with the minimum number of employees and getting a dollar's worth for a dollar spent .... The definition which I feel more suitably fits the situation and circumstances that will be presented in this paper is that provided by Sutermeister (1976, 5): ... output per employee-hour, quality considered." One productivity measurement alone, however, is of little use. It is also important to consider the effect that individual productivity may have upon group productivity andjor organizational productivity. For this reason, productivity consideration will include the worker, the work process and the managerial consequences. Since the advent of alternate work schedules in private and public institutions, a variety of experiments have taken place with an equally diverse number of outcomes. Since our emphasis here is on the effect of these schedules on productivity, the literature applicable to the study andjor effect of such work 76

PAGE 87

schedules were reviewed. Although numerous authors suggest that AWS may have an effect upon productivity, as noted earlier, there were only six attempts to assess this phenomenon -and of these, only three are empirical studies. The literature does identify this weakness, and notes that the data supplied have been obtained largely from employee opinions rather than as a result of extended term pretesting and posttesting of alternative work schedule experiments. Harrick, Vanek, and Michlitsch (1986), Kim and Campagna (1981) and Orpen (1981) did conduct empirical studies, but only Harrick, et al., and Orpen employed a pretestjposttest design. Even the Interim report (OPM 1981) was suspect by the General Accounting Office (U.S. Comptroller General, 1980) because of weaknesses in testing, and the General Accounting Office also questioned whether the methods of analysis used would provide Congress the information it needed to decide the extent to which variable work hours should be allowed in the Federal work force. The Interim Report (OPM 1981) indicated that although there were some negative revelations from a "small" number of experimenting organizations, these revelations should not obscure the fact that information supplied through the overall study ... overwhelmingly suggest[s] that the benefits derived from reducing 77

PAGE 88

unproductive staff time and enhancing productivity among the large majority of experimenting organizations far outweigh the few problems encountered in administering AWS programs" (p.43). The Interim Report tied productivity to "efficiency of operations," and relied primarily upon narrative statements since statistical measurement of productivity was not otherwise available. No pre-or posttesting was taken of the sample work units selected, nor was the selection of these units random (willingness to participate was a deciding factor). In the 1982 OPM Alternative Work Schedules ExPeriment -Final Report, deficiencies in the interim report study were also noted and qualifying information was subsequently provided on page 2: The experimental program demonstrates that AWS can not be mandated for all organizations. As the General Accounting Office (GAO) pointed out in its evaluation of the AWS experimental design, the experiment selection was not a random sample of government organizations. Instead, organizations that thought the experiments would be useful and effective in their operations formally submitted requests to OPM to experiment. Such self selection is the opposite of the random sample concept. The 1500 organizations and 325,000 employees participating nationwide is still well less than 20 percent of the total Federal workforce. Because of the self selection feature and the limited share of the universe, it is not possible statistically to generalize results to the entire Federal Government 78

PAGE 89

The Final Report also noted that: .... some organizations did experience adverse effects in the key areas of productivity and service to the public and all alternative work schedule options were not equally successful in all organizational settings. (p.4) Literature considering alternative work schedules and their effect upon productivity or performance primarily dwells upon impressionable generalities associated with new approaches to human resource management. With respect to the concept of telecommuting, Jack M. Nilles, president of JALA Associates, a Los Angeles consulting firm in management uses of information technology, notes with respect to telecommuting, "Even the most hard-nosed skeptical managers say, 'At least productivity hasn't gone down'" (McGee 1988 60). Job design is another avenue where opportunities for job enrichment are frequently explored. Time management, in this context, offers the job holder through alternative work schedules the greatest possible freedom to decide when to start work, when to stop work, when to take lunch or a break -and, if taken to the extreme degree, how to assign priorities. Principal emphasis in the literature on alternative work schedules is how such schedules improve the quality of life. Morale is often mentioned as the beneficiary of the adoption of alternative schedules. Peter F. Drucker ties 79

PAGE 90

morale to performance in the following manner: ... 'Morale' in an organization does not mean that 'people get along together;' the test is performance, not conformance ... (1974, 455). The Dictionary of Personnel Management and Labor Relations (Shafritz 1980) takes a similar approach in defining morale: ... collective attitude of the workforce toward their work environment and a crude measure of the organizational climate .... With this in mind, most literature that discusses the application of alternative work schedules touches only lightly upon specific and "hard" examples of productivity improvements but instead attests to improvement in job satisfaction by the employee; less gathering, talking and gossiping, more work concentration and improved "morale." Others point out the "biological clock" process whereby any one individual will probably not be equally productive throughout various periods of the day. This concept is explained in more depth by Pierce and Newstrom (1980, 121) who posited" a flexible working hours arrangement can provide the context for a more efficient utilization of the human 24-hour clock (circadian or diurnal rhythms) and can decrease the amount of stress experienced by some employees .... A more recent article (McKendrick 1989) 80

PAGE 91

which discussed a recent AMS Foundation Survey (1989), did not even touch upon productivity, but instead indicated that respondents were more or less evenly split between citing morale or decreased tardiness/absenteeism as being the main advantages of flexitime. In compliance with the requirements of P.L. 95-390, the Office of Personnel Management analyzed what they felt was a unique use of alternative work schedules which they called the "Denver Experiment." In reporting on the results of this experiment (OPM, 1981), productivity is only lightly touched upon in summation, but these limited comments in Appendix D appear to be most interesting: ... the "Denver Experiment" has: ... Improved or had a neutral effect on efficiency in government operations, including productivity .... ... have been four terminations to date, with two more planned prior to the March, 1982 cut-off. Two of these were the result of uncontrollable factors unrelated to AWS. The other four were the result of difficulties encountered with AWS. For the group, these included productivity declines, increased supervisory workload, and problems in ensuring adequate employee and supervisory coverage .... ... overall, participating installations in Denver perceived AWS as generally having positive or neutral effects on the productivity of their organizations. Forty-one percent reported no change in the quality, or timeliness of work as a result of AWS experimentation. Another 41% reported increases in one or more ... aspects (no declines in others). Fourteen percent indicated declines .. 4% reported offsetting effects .... Other observations on this very perplexing effect 81

PAGE 92

attest to the lack of specificity with which the relationship between alternate work schedules and productivity is measured. Rubin (1979) refers to the biological clock effect, but offers only a general statement that "productivity has generally increased when flexitime is used" (p.278). Examples of productive uses of time are limited to observations that employees normally start to work on time and come to work intent on doing the job rather than "just putting in hours." Owen (1977, 155) notes that "The new systems of flexible hours have not been in operation long enough and are not sufficiently widespread to provide definite measures of their social potential." But then, he also warns of potential problems with productivity by adding, ... productivity must be maintained in each case without undermining the earnings, hours, and working conditions of the work force." There is a definite change in the tenor of the literature from the earlier to the later publications, which is only reasonable since with increased experimentation more research occurred, and the research was more objective and scientifically obtained. This is not to say that the early literature was not appropriately scholarly, but it was presented more often in the context of what was then known about alternative 82

PAGE 93

work schedules and their possible effects upon the work force and productivity. This is particularly true of the enormous amount of material which was developed in the pre-1980 period. The work of Bernard (1979, 54) in examining flexitime potential notes: "As of this writing, no analytical study exists which clearly indicates the net value of FWH to a firm .... Yet, in commenting upon productivity, Bernard states: "Flexitime primarily affects productivity through one or more of three channels: improvement of morale, reduction in stress, or reductions of frictions such as turnover or absenteeism." This finding seems to agree with later observations of the potential or real effects. The same theme is mentioned by Golembiewski and Proehl(1978), Newstrom and Pierce (1979), Olson and Brief(1978), and Nallen (1979), who report upon the information as supplied, but conclude in a manner similar to that of Schein (1977, 465) earlier, "Given the differing outcomes of the experimental groups, as well as experimental design limitations, no clear cut conclusions with regard to the impact of flexible working hours on productivity can be made .... A review of post 1980 literature reveals that although there are still some obvious carryovers of the earlier enthusiasm and credulity, later articles provide 83

PAGE 94

a more balanced approach, still positive, but somewhat more cautious. Coltrin and Barendse (1981, 713), for example, point out that ... The apparent success record of flexitime may be simply the result of lack of reporting of failures ... Managers should keep in mind that while most research seems to demonstrate favorable results, few failures of the system have been studied .... Ronen (1981, 818) in his observations with respect to the public sector, notes that tardiness is virtually eliminated under flexitime, but cautions, "Despite the advantages associated with flexible working hours, many supervisors have expressed anxiety about the degree of variation in an employee's arrival and departure times ... With the implementation of a flexible working hours program, employees tailor their schedule to suit their personal needs and ... these needs dictate the characteristics of arrival and departure patterns .... Rainey and Wolf (1982, 56) also observe that Federal employees" will occasionally admit that 'not much gets done around here between 4:30 and 6:30' .... The specific impact upon productivity appears to still be an unknown but presumedly neutral factor. Numerous studies have now taken place which attempt to measure the effect, but they tend to vary in their results. In commenting upon this observation, Narayanan 84

PAGE 95

and Nath (1982, 228) report as follows on the impact of flexitime on productivity: The study found no change in reported productivity as a result of the introduction of flexitime -a finding that was true for all levels .... (The evidence here corroborates the 1977 findings of Schein, Maurer, and Novak.) Narayanan and Nath go on to say that "Convergence of these two studies regarding the impact on productivity leads us to suggest that the introduction of flexitime by itself does not result in any significant productivity gains .... In another study, Rainey and Wolf (1981, 60) conducted an experiment on productivity at a bureau of the Social Security Administration. Their findings indicate a substantial and approximately equal increase in accuracy of output for both experimental and control sections, which may be attributed to the new processing procedure. Quantity of output, however, fell significantly ... in the control unit and not in the experimental unit .... There is some reason to believe that the improvement reflects improved attitudinal factors-i.e., commitment, enthusiasm, or sense of responsibility -rather than mechanical factors such as reduced tardiness .... Narayanan and Nath (1982 226) in testing attitudinal factors in a flexitime experiment conducted at a multinational corporation in the northeast 85

PAGE 96

note the absence of research and state that more needs to be conducted" before any generalization can be made with respect to the impact of flexitime on quality of supervision, job satisfaction and productivity .... It would therefore appear that attitude and personal perceptions have to a great extent influenced what may or may not be the impact of alternate work schedules on productivity. Management andjor supervisory employees may view the alternate work schedules as primarily an accommodation for the employees, and probable cause of future operational difficulties and breakdown in control functions. The employees, on the other hand, may view this as a challenge, an opportunity, or in some instances a device by the supervisory hierarchy to reduce costs at their expense. If the latter occurs, it is possible, even probable that productivity could suffer as the employee attempts to beat "the system," may malinger or even consider or encourage sabotage. Flexitime is popular among employees and some managers, but what about the impact on first-line supervisors, and how they perceive the effect of the alternative work schedules on productivity? Only one article fully addressed this issue, but the thoroughness of the work of Golembiewski, Fox, and Proehl (1980 46-47), would seem to adequately fill this void in the literature available. After 86

PAGE 97

recognizing the problems, and there were many, Golembiewski reports in his summation: ... Despite all the problems, however, most supervisors saw definite upbeat effects of flexitime on productivity and morale. Although only a few of the supervisors could make objective measurements of productivity, almost all found it easy to make a confident estimate of trends in output. Of the 43 supervisors who participated in the sensing groups, more than 70 percent believed that productivity had increased under flexitime, about 20 percent saw no change, and only about 7 percent felt that a noticeable deterioration in productivity had occurred .... Golembiewski, Fox, and Proehl, also note that flexitime will not transform ineffective supervisors into managerial paragons. On the contrary, flexitime is likely to exaggerate some managerial flaws: lack of confidence in judging performance, the need to overcontrol, lack of trust, and so on .... Earlier, it was mentioned that only six articles in the literature available relate solely to the impact of alternative work schedules on productivity, although others in their analysis of alternative work schedules touch upon the productivity outcome. The Schein, Maurer, and Novak article (1977), describes the sampling of 246 clerical-level employees over a four month period during an experimental flexible working hours program. It was reported that the results were mixed but, overall, the introduction of flexible working hours did not have an 87

PAGE 98

adverse impact on productivity. Stanley D. Nollen (1979), in evaluating the effects, worked with completed surveys and studies, and the results deprived of the original raw data may reflect, in part, bias or inaccurate reportingjsampling processes but, in reviewing these 14 case studies and eight surveys, Nollen rationalizes that the answer to the question of whether or not increased productivity will result from alternate work schedules lies with how the alternate work schedules are managed. Welch and Gordon (1980) analyzed the impact of flexitime implementation on productivity in the claims processing department of a medium-sized insurance company in the southwestern United States. The study covered a period which began in November, 1977, and ran through December, 1980. The study indicated that there were improvements in productivity, which were applicable to claims processing, the primary problems appeared to be centered around personnel: ... It is probable that firms which use such a system will require supervisors to work longer hours. During this study, one of the three supervisors resisted the system and attempted to negatively influence the attitude of other supervisors. In order to overcome these problems with lower management, it is recommended that supervisors also be offered a flexible work schedule ... (page 65) The fourth and last comprehensive study solely related to productivity andjor performance considered all 88

PAGE 99

previous work in this field and benefitted from the consultive arrangement with other source matter experts. Kim and Campagna (1981) conducted a study over a fourmonth period on a flexible work schedule experiment at a county welfare agency involving 353 employees within 4 divisions. After noting the possibility of "Hawthorne" andjor "Rosenthal" effects, and that the employees were also aware that the results of the study could potentially affect the entire agency, they still find that: the results of this study appear to justify the argument that the flexitime program has a positive impact on attendance and performance among public sector employees. It should be noted that this study investigated only one of several forms of the flexitime program. Additional research on various other forms of the flexitime program in different organizational contexts is clearly warranted .... (page 740) With this last study, it can be established that it is possible, from literature available, to determine that the use of alternative work schedules can have a beneficial effect upon productivity, but only when approved for appropriate work units. Approval of an alternate work schedule without considerable research and planning could lead to dysfunctional consequences. A few studies did utilize some levels of scientific rigor, and did deal with alternative work pattern applications in an experimental manner. It is useful to consider the work of 89

PAGE 100

Dr. Ben Burdetsky and Dr. Marvin Katzman, who examined through survey the experiences of organizations who had previously implemented alternative work schedules. The results of this survey was published in the Journal of Systems Management, December, 1981, under the title of "Alternative Work Pattern Applications," and the results indicated that those responding felt good about the work schedules, reported improved morale, but little tangible improvement in productivity occurred, although it might otherwise have been expected. The survey information only, quite naturally, included data from those who elected to respond, so it could be argued that opinions supporting improved productivity or faltering morale may have not been obtained from the sample available. The work of Golembiewski in measuring professed workplace enhancements, such as productivity, morale, etc., from adoption of flexitime is significant. It is interesting to compare his article (1982) "Do Flexible Workhour Effects Decay over Time?," with some of his earlier observations. He looks at what he calls flexitime at T=S years, and from survey data obtained from a large firm who had earlier experimented with flexitime, he suggests that although a" substantial proportion of respondents believe theF-T program is abused . (p.39), he still is convinced from the survey data supplied that 90

PAGE 101

it is still felt to be overall desirable. He calls for frequent reviews, levels of checks and balances and trust between supervisors and employees. Unfortunately, productivity improvements could not be measured, but he did note that only 41 percent of those responding did not feel that "goofing off" was the price for adoption of flexitime. The studies, research and literature on the effect of alternative work schedules on productivity are far from being complete. Prior to the passage of Federal alternative work schedule legislation, it was estimated that over 100 bureaus and agencies had experimented with various forms of flexitime (Ronen and Primps 1980), not including earlier experiences with compressed tours. The extension of the experiment (until July 23, 1985) was designed to encourage agencies to track in a more effective and efficient way measures of performance under these alternative schedules. Continued labor union interest will trigger management to more carefully consider and weigh the advantages and disadvantages. It is not enough to say, "we just do not want to experiment with alternative schedules," it is now the law that agency heads must negotiate this with unions. That productivity is affected is not questioned, but to what degree? And, in what manner? Previous governmental 91

PAGE 102

research admittedly lacked the objective quality needed; early claims that alternative work schedules improved productivity were amended to say that productivity was not adversely affected. The Legislative History accompanying P.L. 97-221, contains a statement by Senator Rudman concerning earlier documentation of alleged productivity benefits accruing to the Government from use of alternative work schedules. This statement adequately covers the quality of research and literature with respect to the Federal experiment: .... I think it is interesting to note that the OPM report on flexitime indicates that flexitime did not significantly reduce productivity. This is interesting in light of the fact that the argument has been that flexitime would raise morale and increase productivity. If there has been no appreciable increase in productivity due to flexitime, I question the merit of extending the experimental program to the status of permanent authorization. (U.S. Congress, Senate 1982) In consideration of the above statement and the associated implication, it is felt that recognition should also be given to the morale benefits achieved, i.e., improvement in the quality of working life, through use of alternative work schedules. If productivity is subsequently improved through more effective management of human resources, that then would be an additional benefit. If productivity is adversely affected it would 92

PAGE 103

appear that the value of the alternative work schedule would then be in question. Summary Although alternative work schedules have been in use for approximately 25 years, they were not introduced into the Federal workforce until September 29, 1978. Experiments with flexible and compressed work schedules were determined by the Office of Personnel Management to be successful, and P.L. 99-196 was subsequently enacted to convert the previous temporary authority to permanent. Early literature overwhelmingly lauded the beneficial aspects of the flexible workhours, but those studies that were conducted were primarily anecdotal, testimonial, and post hoc in nature; few were based on rigorous empirical investigations. In those few studies which attempted to deal with the impact, positive or negative, upon productivity, the methodology was hampered by the lack of solid quantitative and qualitative measurements. Irrespective of the degree of research involved, much of the earlier material published promised quantum leaps in productivity, significant reductions in absenteeism, favorable employee retention results, decreased overtime and increased efficiency. Later research did not, 93

PAGE 104

however, convincingly demonstrate the productivity and absenteeism claims, but it did attest to the potential for other work related benefits. Employees find that alternative work schedules are most helpful in their continued search for life satisfaction. The changing needs, aspirations, and values of workers are being sought through a combination of on-job and off-job activities. Although there may be individual differences, the majority of employees are placing increased emphasis on the off-work aspects of life. Quality-of-work-life demands have been fueled by these changes in employees. One of the more profound changes is the increase of women in the workforce and their need, in many instances, to accommodate what may be considered a dual career. Changes in family status are perhaps equally significant because of the number of families where both spouses work. The more subtle change is that of employee values and their preferences about the terms and conditions of work. Much of the material now being published attests to employee expectations and the efforts of management to accommodate employee needs and yet make it work for the organization. The concept of employees telecommuting, or otherwise participating full or part time with flexiplace work arrangements, is now being considered. The Federal Government is 94

PAGE 105

experimenting with these arrangements on a limited basis, and therefore the extent of literature on this subject is not yet very extensive. As employee needs change, in addition to the more traditional, but budgetarily limited, monetary rewards, employees have turned to, and now also seek, other employment rewards andjor benefits such as more favorable work schedules, career counseling, and on-the-job fitness programs. Once acquired, it becomes extremely difficult for management to renegotiate what have become employee rights. Research as to the impact that these nonmonetary rewards may have on other effects than employee satisfaction within the Federal Government is rather limited, and does not exist within the National Park Service. Another aspect, however, with respect to government employment that should be considered, but is not part of available literature, is the impact of other significant forces that helped influence the size and makeup of the Federal workforce. It is difficult, if not impossible, to measure precisely the effect of AWS on productivity, recruitment, performance, and employee demographics when such factors as extended employment freezes and protected group outreach programs are considered. Efforts by the Executive Branch to control the size of government did 95

PAGE 106

not serve to encourage young people into government. And, with the obvious satisfaction that most Federal employees have expressed regarding employment, the trend has seemingly been towards a workforce that is older and less willing to accept radical changes in work schedules. 96

PAGE 107

CHAPTER IV RESEARCH DESIGN General Structure and Methodology Framing Despite the adoption of alternative work schedules by many organizations, research efforts documenting the impact of these schedules upon productivity and performance has been rather limited. Originating in West Germany and shortly adopted by other European countries, alternate work schedules were introduced in the United States in 1972. Early support for this program was extended through impressionistic evaluations in the press and in certain non-experimental journals. A review of available experience in the application of flexible workhours was provided by Golembiewski and Proehl (1980) identified 31 studies which had taken place before 1980, and, of these studies, two used comparison or control groups (they were of the prejposttest design) and only one provided any statistical treatment (Golembiewski, Hilles, and Kagno 1974). More recent studies attempted to demonstrate that alternative work schedules dependably improve the

PAGE 108

performance of organizations, employee productivity and satisfaction, and relationships between employees and management. However, this research largely dwelt upon subjective measurements and ignored, or preferred not to deal with, some of the more important dimensions --such as effects upon organizational behavior. The effect that alternative work schedules may have upon productivity, from a quantitative perspective, would also appear to be in need of further review and analysis in view of the number of supervisory and managerial concerns that are still expressed, and the possibility that many of the conclusions previously reached regarding the effects of alternate work schedules upon productivity are unsupported by hard data. The National Park Service (NPS), as well as other Federal Bureaus and Offices, experimented with alternative work schedules from 1978 until 1985, when these schedules became a permanent option. Although most eligible employees favor the program, many supervisors and managers in the NPS are of the opinion that these schedules may have an adverse effect upon productivity, performance, andjor acceptable organizational response to its mission responsibilities. Since there are a number of opportunities to examine the effects of alternative work schedules upon performance and productivity within 98

PAGE 109

the National Park Service, and since the results of such research would be of interest beyond the immediate agency, evaluation of this phenomenon is considered to have substance and worth to the total public sector. In order that the research design could be properly structured, it was determined that the problem should be framed in a series of research hypotheses: 1. Productivity measurements for experimental (flexible work hours -flexitime employees) and control groups indicate that employees not under a compressed work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. 2. Productivity measurements for experimental (alternative eight-hour work schedules) and control groups indicate that employees not under an alternative eight-hour work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. 3. Use of annual leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups during the pretest-posttest time periods. 4. Use of sick leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups during the pretest-posttest time periods. 5. Employment turnover is lower with respect to offices andjor parks who offer liberal work schedule opportunities (experimental), than it is for offices andjor parks who offer primarily only traditional work schedules (control). 6. Employees who work under an alternative work schedule (experimental group) tend to have higher morale, and, on an average, higher 99

PAGE 110

performance evaluations than those employees who work under more traditional work schedules (control group). The calculation of data that is relevant to the above hypotheses, and the evaluation and eventual acceptance or rejection of each hypothesis is based upon a research design which: o includes and identifies the people to be studied o sets the timing of the investigation o establishes procedures for the selection of data o describes how data will be processed o provides for thorough analysis In addition to the structure of the design, another concern of internal validity that has been addressed is the matter of definition. It is not enough to say that productivity is "the efficiency with which resources are used to produce a government service or product at specified levels of quality and timeliness," as defined in Sec 2.(a) of E.O. 12552, or, "output per employee-hour, quality considered," as suggested by Sutermeister (1976). We need to know if we are addressing the productivity (performance) of an individual or the productivity of an organization. Other factors, such as the economists' term of "quantity of goods produced to the quantity of resources used," as well as the quality and worth of that produced, also should be considered. 100

PAGE 111

Since all of these factors are important to the research results, the design plan that has been formulated is structured around the concept that the productivity of any organization is principally determined by the organization's technology, the job performance of its employees, and further, that an employee's productivity is a function of ability and motivation. This has been traditionally expressed as; Performance = f (Ability, Motivation). This concept is also rather succinctly stated in the Facts on File Dictionary of Public Administration (Shafritz 1985), which goes on to define productivity as: ... [the] measured relationship between the quantity (and quality) of results produced and the quantity of resources required for production. Productivity is, in essence, a measure of the work efficiency of an individual, a work unit, or a whole organization .... Because of the number of employees and the variety of occupations, it is certainly possible that any number of experiments could be conducted in the NPS under a quasi-experimental framework to; (1) test the effect of alternative work schedules on productivity, or (2) test the effect of any other specific employee performance issue. The basic criterion for an acceptable tested outcome would largely depend upon the extent to which the study or experiment protects against the effect of 101

PAGE 112

extraneous variables. In this particular instance, the primary concern would be whether or not any other independent variable, other than an alternative work schedule, exercised some level of effect upon the desired dependent variable outcome, i.e., productivity, use of leave, personal performance, or employee motivation. Much of the literature which discusses accomplishments under alternative work schedules clearly points out that only limited productivity measurement actually took place prior to the schedules being implemented, and after implementation the primary measure was employee satisfaction. Consequently, if any studies were made of the effect of an alternative work schedule (intervention) on productivity, it generally was of the after-only variety. This, although interesting, does not provide reliable support to any conclusions that may be drawn with respect to the effects of alternative work schedules upon productivity. An somewhat current study by Ralston, Anthony and Gustafson (1985) supported a hypothesis that "flextime" had a positive effect on productivity when there are limited resources being shared by a work group. This study, which did use a control group, made extensive use of what appeared to be rather modest pretest and posttest statistics. An earlier study conducted by Rainey and Wolf (1981), also 102

PAGE 113

utilized a control group, but yielded mixed results by identifying some short-term benefits but nothing specific on long-term. Barrick, Vanek, and Michlitsch (1986) did incorporate a pretest/posttest design and a longitudinal frame of 18 months, with 16 months between measures of satisfaction, but, control groups were not used due to limitations imposed by a union contract. Specific Procedures Concern with respect to construct validity suggests that three different analyses (experiments) would be appropriate: (1) comparison of production from selected experimental and control groups, (2) evaluation of data obtained from employee performance indicators such as supervisory performance evaluation reports, use of leave, etc., (3) measures of satisfaction, i.e., employee satisfaction, supervisory satisfaction, and client satisfaction. Each of these approaches would demonstrate some of the characteristics common to experimental, survey, and/or interpretive methods. The interest is primarily in comparing the productivity and performance attributes of employees under flexible and compressed work schedules with previous performance (pretest) and/or a control group who are performing similar assignments but under traditional work schedules. 103

PAGE 114

There are certain human characteristics, i.e., age, sex, family status andjor circumstances, that represent antecedent variables to those dependent variables (attributes) previously identified as productivity, satisfaction, performance, etc., and the independent variables associated with the adoption of an alternative work schedule. In addition, the implementation policy of each individual park or office unit as well as the culture applicable to that organizational unit can have an intervening relationship between the dependent and independent variables. To better illustrate this relationship between all of the associated variables, an elaboration analysis model is provided as figure 4.1. The first approach tested productivity by comparing the accomplishments of experimental and control groups on a series of similar tasks which were partially tracked through existing systems. National Park Service supervisors in these selected groupings were asked to record individual productivity as well as group productivity over the past three years. Employees have consistently expressed their interest in alternate work schedules, but they have not in all instances received 104

PAGE 115

Age Sex Antecedent Family Status Figure 4.1. Variable Elaboration Model 105 Independent Alternate Work Schedule: Flexltime Compressed Flexiplace I I Intervening Management Policy Organizational Culture I Dependent Productivity Satisfaction Performance Motivation

PAGE 116

approval because of managerial concerns, and perhaps inexplicit supervisory apprehension, over the efficiency and effectiveness of flexible work schedules. Six control and experimental groups were identified from organizational units within one park area (Division of Maintenance, Rocky Mountain National Park) and five regional offices. This selective geographical organizational array thus provides a representative analysis of experience from major National Park Service field management units. Each of the ten regional offices have a functional organizational unit identified as either the Division of Personnel or the Division of Human Resources. Each of these units perform the same basic personnel service functions for park areas within their geographic regions. Individuals working under flexible work schedule tours are found in eight of the ten regional personnel offices as are employees doing the same basic work in nine of the ten offices who are working under traditional work schedules. On October 1, 1990, there were 263 employees involved in human resource/EO work in these ten offices, over 30 percent, 87 employees, were working under compressed (5/4/9 and 4/10) or flexible tours. The AWS distribution is identified below in table 4.1. 106

PAGE 117

Table 4.1. AWS Human Resource Distribution by Regions As of October 1, 1990 Region/Office EO/HR Authorized Percent employees AWSW/S Authorized Alaska 12 2 16.67 Mid-Atlantic 26 11 42.31 Midwest 28 2 7. 14 North-Atlantic 26 9 34.62 National Capital 42 17 40.48 Pacific Northwest 13 13 100.00 Rocky Mountain 41 14 34.15 Southeast 20 0 .00 Southwest 23 3 13.04 Western 32 ...1._ 50.00 263 87 33.08 Using a pretest (observation) as a covariate in an ANCOVA model, any observed antecedent differences in selected control and experimental groups were statistically examined. If ANOVA only were used it would be presumed that the treatment would be the sore cause of difference in posttest performance between experimental and control groups. Using for an example an experiment involving the Division of Human Resources, Rocky Mountain Region, National Park Service, an initial research design is as follows: 107

PAGE 118

x1 represents Alternative Work Schedule Intervention oel is experimental group pretest (productivity measure) ocl is control group pretest oe2 is experimental group posttest (6 mo) oc2 is control group posttest Because there are always the possibilities of potential internal validity threats arising from one circumstance or another, it is proposed that other experimental groups performing similar work assignments in other Regions be added to form a multi-group design. By adding these experimental and control groups from different Regions, the design would now be as follows: ocl oe2 oc2 x2 oorel oel xl oe2 oci oc2 oorei xi oore2 oorci oorc2 represents Alternative Work Schedule Intervention is experimental group pretest (productivity measure) is control group pretest is experimental group posttest is control group posttest is other regions AWS intervention is other regions experimental group pretest (productivity measure) is other regions control group pretest is other regions experimental group posttest is other regions control group posttest 108

PAGE 119

The second approach tests employee productivity by whatever measure may be available (independent variable), individual performance evaluations, as well as other measurable performance indicators such as use of sick leave and annual leave for brief unscheduled periods, but in a quasi-experimental mode because of the lack of true random assignments. The third approach deals with the element of satisfaction, i.e., employee satisfaction, supervisory satisfaction, and client satisfaction. A questionnaire (see Appendix B) was used to secure responses from a random sampling of employees who are working under either compressed or flexible schedules, and of those who are working under traditional schedules. The questionnaire was designed to assess individual satisfaction with the working environment, including their own work schedule, and their feelings regarding productivity, performance, and organizational culture. This questionnaire was given to the supervisors of these same randomly selected employees to ensure that information was also obtained with respect to supervisory satisfaction with matters of performance, productivity, and the appropriateness of work schedules. To measure client satisfaction, park area managers were solicited as to whether or not their park areas are being suitably served by park staff andjor 109

PAGE 120

regional support facilities, and whether or not alternative work schedules have any impact upon the level of services received. Another measure of employee satisfaction that was considered was available organizational turnover statistics. Validity and reliability of the survey/questionnaire instruments used was extremely important. Consequently, those questions selected previously formed a part of instruments which were subject to prior testing for construct validity. Of the 46 questions used, 23 were taken from previously tested instruments used by the Office of Personnel Management which were specifically designed to measure employee satisfaction, attitude, morale and motivation (three of these questions were also used in a rather extensive 1983 National Park Service employee survey). From the instrument designed by Jon L. Pierce and Randall B. Runham, "Attitudes toward Work Schedules: Construct Definition" (1986), 18 questions were selected with respect to measurement of attitudes towards work schedules. The remaining five questions had not been previously tested, but they were designed primarily to measure personal beliefs/attitudes/ perceptions regarding certain alternative work schedules 110

PAGE 121

and the effect that these schedules may have upon productivity. OPM-designed questionnaires have been often used in past National Park Service personnel management evaluations, and they contain a variety of questions regarding personnel management, including the following: 62. Do you work under an Alternative Work Schedule (AWS), e.g., flexitime or compressed work schedule? Answer choice -yes or no. 63. If YES, what effect does it have on the productivity of your work unit? Answer choice -positive effect negative effect no apparent effect Although supervisors have additional questions to respond to with respect to the cited OPM questionnaire, they are not asked whether or not they support or favor AWS's, or any other questions regarding AWS's except that dealing with their own personal use of AWS. Although the use of these questionnaires might be helpful in determining employee attitude concerning alternative work schedules and productivity (pretest), it is presumed that since most, if not all, employees who are under an AWS requested such a schedule, they would not be likely to respond that the AWS had a negative effect upon productivity. For this reason, the questionnaire that 111

PAGE 122

was used in this research did, of necessity, contain more than just direct questions as to what possible effect an AWS may have on productivity. Research Population A varied population was used for research purposes. There were approximately 17,923 employees in the National Park Service as of April 25, 1990; of these, at least 1,725 worked under some type of compressed work schedule. A random sampling, sufficient to provide a 95 percent confidence level would permit either an acceptance or rejection of the selected hypotheses. Those employees not under an alternate work schedule were furnished with a questionnaire asking specific questions with respect to satisfaction with their job, their work schedule, and their thoughts with respect to productivity and AWS's. Since it was estimated that at least 30 percent would not return the questionnaires, the sample size required based upon an untested questionnaire was provided through the National Park Service's Office of Statistics. The computation can be illustrated as follows: 2 choice answer: n' = <22-1> [100&1.96)12 12(5) = 115248 = 384.16 = 385 300 n = 16198(385) = 377 x 30% = 490 16198+385 112

PAGE 123

3074 5 choice answer (Likert Scale): n' = <52-1> [100(1.96)12 n = 16198(3074) 16198+3074 = 921984 = 3073.28 = 300 = 2504 X 30% = 3255 For an error of no more than 1/2 category or scale point: (f = .5) n' = (52-1> (5+1>2 (1.96>2 48( .5)2 = 3319.1424 = 277 12 For an error of no more than 5% of the full span of 5 scale points: n = 16198(1107) 16198+1107 300 = 1,036 X 30% = 1346 A sample size of 1,346 would subsequently be necessary, if an untested instrument had been selected, in order to maintain a 95 percent confidence within the confidence intervals and standard errors previously stated. For this reason, a 46 question, pretested questionnaire/survey instrument was used, and by using the pretested questionnaire a maximum sample size for an infinite population, 385 people, was determined to be appropriate. 113

PAGE 124

As a further step towards a modified stratification sampling process, those employees under an alternative work schedule were randomly selected within work schedule types, i.e., 4-10, 5/4/9, flexible, etc., in order to obtain a greater degree of representativeness. By using the same questionnaire, these employees were thereby asked the same questions regarding satisfaction with their job, work schedule, and thoughts with respect to the effect of AWS's on productivity, as those selected from the total non-AWS population. The sample sizes for these groups are computed as follows using the previously stated formulas for sample sizes for an infinite population: 2 choice answer: n = 1725(385) = 315 X 30% = 410 1725+385 5 choice answer: n = 1725(3074) = 1105 X 30% = 1435 1725+3074 For an error of no more than 5% of the full span of 5 scale points: n = 1725(1107) = 1725+1107 674 X 30% = 876 A sample size (stratified) of 876 would therefore be suggested as being sufficient to maintain a 95 percent confidence level for an untested questionnairejsurvey 114

PAGE 125

instrument. In order to avoid such a large sample, it was determined that only a tested instrument would be used, and a sample size of at least 41, using the same ratio as proposed for the non-AWS grouping be used. It order to increase strength of relationship in the analysis of supervisory response to AWS's and their impact upon productivity, the supervisor for each person randomly selected in the stratified sampling of AWS employees was also furnished a questionnaire seeking information with respect to the resultant impact, if any, of the employee's AWS upon hisjher own productivity and the productivity of the total work group. No further sampling process was contemplated with respect to this group, although it is recognized that certain supervisors selected may be reporting on more than one employee. Instrumentation In order to maximize the reliability and validity of the study's findings, the single questionnaire was issued to all non-AWS employees randomly selected from the nationwide employee listing. Only permanent employees were selected since temporary employees have had only limited experiences with AWS and are usually terminated within four months. This questionnaire measured the degree of work satisfaction (morale), attitudes towards 115

PAGE 126

performance, supervision, the job and workplace, work schedules, and productivity. The same instrument was given to randomly selected AWS employees. The selectees names were further used in order to secure performance evaluation data. The correlation measures obtained from these two exercises were subsequently used to either reject or accept the following hypothesis: Employees who work under an alternative work schedule (experimental group} tend to have higher H6 morale, and on an average, higher performance than those who work under more traditional work schedules (control group). Since this questionnaire/survey contained questions selected from Dunham and Pierce instruments (1986 and 1989), Federal Employee Attitude Surveys, and, when applicable, from the National Park Service Employee Survey of 1983, dealing with issues of productivity, the questioning process was used to affirm conclusions arrived at through appropriate analysis of experimental and control groups selected within the National Park Service. These measures were used to accept or reject the following hypotheses: H1 Productivity measurements for experimental (compressed work schedule employees) and control groups indicate that employees not under a compressed work schedule (control groups) demonstrate, on an average, higher productive levels between pretest-posttest time periods. 116

PAGE 127

H2 Productivity measurements for experimental (alternative eight-hour work schedules) and control groups indicate that employee not under an alternative eight-hour work schedule (control groups) demonstrate, on an average, higher productive levels between pretestposttest time periods. Behavioral expression of employee attendance behavior is demonstrated, so to speak, by the extent of employee tardiness, absenteeism, and turnover. Although it is not possible to precisely measure tardiness, it has been suggested that use of alternative work schedules, particularly that of compressed tours, reduce the amount of absenteeism andjor other unscheduled absences. The customary payroll process for National Park Service employees provides for recording the use and accumulation of both annual and sick leave. The same random selection of non-AWS and AWS employees was used to compare annual and sick leave usage to determine if unscheduled use of annual andjor sick leave is more characteristic of nonAWS employees that it is of AWS employees. Information was secured from payroll leave records. This information was subsequently used to affirm or reject conclusions arrived at from experimental/control group testings designed to accept or reject the following hypotheses: H3 Use of annual leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups between the pretest-posttest time period. 117

PAGE 128

H4 Use of sick leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups between the pretest-posttest time period. This study did not consider if differences do exist between part-time and full-time employment with respect to turnover. Consequently, the concentration and focus was limited to permanent full-time employment, and separation data was assembled for all those permanent full-time employees who left the National Park Service voluntarily, except for purpose of retirement, within one calendar year. A comparison was made between separated employees who had been working under an AWS, with those who had not. Reason for termination is frequently, but not always, provided by the employee, nor are the reasons that are stated necessarily the sole andjor complete reason for the employee leaving the National Park Service. In some instances an employee will resign to avoid some form of proposed disciplinary action or because of some other ongoing supervisorjemployee conflict. Comparisons were made between those organizations who offer a variety of AWS's, and those that offer only traditional work schedules. These measures assisted in determining whether or not the following hypothesis should be accepted or rejected: 118

PAGE 129

H5 Employment turnover is lower with respect to offices andjor parks who offer liberal work schedule opportunities (experimental), than it is for offices andjor parks who offer primarily only traditional work schedules (control). Data Collection Personnel and payroll data within the Department of Interior has been processed centrally for all its bureaus by the Bureau of Reclamation since 1986. Responsibility for the National Park Service was assumed on March 18, 1984. Payroll reporting within the National Park Service is basically decentralized, i.e., each park area or organizational unit transmits its data directly to the Bureau of Reclamation Service Center. Personnel processing is more decentralized, but all formal actions, such as accessions, losses, changes in pay status or work schedule, are eventually electronically reported to the Reclamation Service Center. Consequently, a complete payroll/personnel data base file is maintained by the Bureau of Reclamation for each employee in the Department of Interior. Access to this data base file obviously has certain restrictions. Queries to the data base file must be accompanied by a certain code(s) which activates the initiating processes leading to delivery of the data requested. The above data base was a critical part of the total 119

PAGE 130

random selection process. Employees who have been assigned a flexible or compressed work schedule, such as a 4/10, 5/4/9, etc., are identified so that automatic overtime payment, or rejection processes are not evoked whenever time in excess of the normal 8-hour day is reported on the time and attendance report. In order to obtain the information provided in Table 2.1, it was necessary to query not only for compressed work schedules by major organizational units, but also for total number of employees. In selecting employees in a random fashion for mailing of questionnaires, a detailed printout, including name, organizational code, last performance evaluation, etc., was essential. Using the suggested sample sizes, a minimum of 385 for non-AWS employees and 41 for AWS employees (stratified by AWS type), selections were randomly made, proportionately by major organizational unit, and the questionnaires were addressed by name and mailed to the appropriate regional administrative officer for distribution and direct return by the addressee to the researcher. The latter process, use of the regional administrative officer as the delivery agent, was to ensure return and organizational cooperation. It is known, for example, that the response rate on previous 120

PAGE 131

questionnaires were lower from eastern organizational units than they were for western or mid-western. It was clear, from investigating the cause for the low return, that the suspicion level was higher in the east, and supervisory support was lower. It is also noted that fewer AWS's (see table 2.1) were authorized in eastern areas and offices as opposed to those in the west. As the questions in this instument were not controversial, protection of employee rights andjor interests did not present a significant problem. If such circumstances would have presented themselves, information would not have been released that would invade the privacy of any individual or subject himjher to occupational risk. Where performance evaluations were part of the research effort, they will be subject to the provisions of P.L. 93-579, the Privacy Act. Information will not be released to supervisors with respect to any employee statement(s); but, information was sought from each AWS employee supervisor regarding his or her perception of that employee's productivity. Data collections from all quantitative quasiexperiments consisted of a summation of output by each individual, which were basically selected work units, over a measured period of time. This information was obtained either from official record data, the survey, or 121

PAGE 132

actual observation. If, for any reason, there were a possibility that this data would lack accuracy or reliability, that particular experiment to which it pertains would have been dropped from this research program, and greater reliance would have been placed upon the other quasi-experiments and qualitative research. Treatment of the Data The focus of the research was the six research hypotheses which framed the research design itself. A few selected figures and tables to demonstrate the effect of flexitime -particularly compressed work schedules, upon employee satisfaction, productivity, and management comfort, has been provided, but these descriptive devices, particularly those statistical inferences derived from the proposed ANCOVA model, are limited and have been presented only when a specific point is to be made with respect to a research observation. The literature review chapter provides important guidance with respect to structure and background, and, in support of this, a table (3.1) was furnished that identified the conclusions of prior research, with respect to attitudinal and performance measures. Some of the earlier studies, particularly those which did not make use of experimetal groupings, tended to project more 122

PAGE 133

positive results. Use of demographic variables were not explored to any extent. If it appeared that these workforce characteristics would have had significant impact influencing or directing research outcomes, then these demographic variables would have been addressed. The fact that an employee is married, or single, of a specific race or age group, or even what sex they may be, should not logically be a consideration since these factors should not influence management decision making options or processes concerning flexitime. However, that fact is that there are demographic differences that may influence work schedule productivity measurements. Information obtained from payroll/personnel sources have been presented through a variety of tables in tabular and in summary form. Raw data has not been expressed unless it is absolutely necessary for the reader's understanding of research conclusions. The research analysis of data sources immediately follow the presented tables. Results of experiments are expressed in linear fashion, and contained in individual tables, with respect to productivity comparisons. Correlations between productivity measures and sick leave, annual leave, turnover, etc., have been presented by either scattergram or bar chart. Wherever possible, the independent 123

PAGE 134

variable, the type of work schedule, has been presented for comparison with experimental outcomes (dependent variables) in the most understandable data array. Summary The six research hypotheses that were identified in the problem section of this proposal focused the search for data and framed the purpose of this research. This research concerns alternative work schedules and their effect upon productivity in the National Park Service. Several factors were therefore recognized in the design construct: (1) The National Park Service does not have a homogenous workforce. (2) With significant authority delegation, central direction and guidance is thus limited and notable differences in policy and procedures exist between major organizational units. (3) Use of alternative work schedules is legislatively encouraged; employees normally are the requestors for these schedules, and as such, along with employee unions, consider reasonable deviations from traditional schedules to be an employee right. (4) Program management exists at numerous levels, and thus managerial satisfaction with alternative work schedules would have to be polled at other than just the immediate supervisory level. 124

PAGE 135

A single project or experiment would only observe effects at that particular time within that particular organization. It would be difficult to accept or reject any of the six presented hypotheses based upon a few isolated experiments, or quasi-experiments, as being generally true and applicable to a major governmental bureau. For this reason, a survey document, or questionnaire, was issued to random samplings of employees working under alternative work schedules, and those who are not, for comparative purposes. Because productivity was extremely difficult to quantify, some experiments (more properly quasi-experiments), were also undertaken to compare and contrast with survey findings. It is also noted that this research program should consider previous studiesjresearch if for only the purpose of testing theory and replication of previous research results. It is recognized that previous research has been limited, and in many instances anecdotal (but may be considered as qualitative data), but nevertheless each has developed a variety of conclusions with respect to the dependent variables that were addressed or considered. 125

PAGE 136

CHAPTER V ANALYSIS AND RESULTS This chapter describes the results of the research that has been outlined in the preceding chapters. The analysis begins with a limited discussion of the variables that are pertinent to the experiments in this study. The relationships between these demographic (antecedent), independent, intervening and dependent variables, were described briefly in Chapter IV, and how the variables affected the study and research will be further explained in this chapter. More specifically, information contained in this chapter will include the results of the analysis of data acquired from the various experimental and control groups (as well as the questionnaire instrument), and report on the influence of alternate work schedules upon productivity. As a refinement of presented research and study findings, the statistical tests that were used, and the outcomes from these tests have been reported in hypothesis order.

PAGE 137

Study Variables Although the factors under study in this research project are related specifically to the measurement of the impact of alternative work schedules on productivity, each of these factors (variables), in the experiments have virtually assumed different possible values andjor outcomes. Certain antecedent variables, such as age, sex, family status, etc., may possibly have an effect on both the dependent variables (productivity, satisfaction, performance, motivation, etc.), and the independent variables considered in this study (flexitime, compressed work schedules, flexiplace, etc.). But the effect of these antecedent variables on the dependent variables may only be secondary in importance to the effect of the independent variables. In other words, the cause and effect relationship would therefore require closer scrutiny. The Antecedent Variables This study had basically two major research activities, the conduct of a series of experiments involving individuals performing similar tasks in a matched pairing of control and experimental groupings, and the distribution and review of the responses to a questionnaire dealing with alternate work schedules, 127

PAGE 138

productivity, work satisfaction, and performance. The demographic personal variables tested on those individuals who were included in the study experiments were: age, marital status, educational level, and sex. Respondents to the questionnaire were asked to provide all of the above personal demographics except marital status. For lack of a more effective way of obtaining some vital organizational demographics for those individuals randomly selected to receive a questionnaire, the survey participants were also asked to provide, (1) their job classification, (2) length of time employed with National Park Service, (3) their current work schedule, and how long they have worked under this schedule, and (4) their preference as to work schedule. Data with respect to leave accumulations and use were obtained from either bi-weekly time and attendance reports or the Department of Interior Special Annual-Sick Restored Leave Report. Absenteeism of participants in the experiments would have been included, but no reports of tardiness were noted by immediate supervisors. Turnover in employment was initially reviewed with respect to research processes associated with the experimental/control groupings, but during the period of review only three out of the 84 employees included in the 128

PAGE 139

experimental/control groups terminated their National Park Service employment. The Intervening Variables As is the case in most (if not all) organizations, organizational culture is significantly affected by its leaders. Horace M. Albright, the second Director of the National Park Service, presided over the initial expansion of the fledgling bureau in 1923. In drafting a form letter to accompany the first announcement of park ranger positions, Mr. Albright indicated that: The duties are exacting and require the utmost patience and tact at all times. A ranger's job is no place for a nervous, quick-tempered man, nor for the laggard, nor for one who is unaccustomed to hard work. If you cannot work hard ten or twelve hours a day, and always with patience and a smile on your face, don't fill out the attached blank. You have perhaps believed Government jobs to be "soft" and "easy." Most of them are not, and certainly there are no such jobs in the National Park Service. (Albright 1985, 145) Alfred A. Knopf, who served as chairman of the Secretary's Advisory Board on National Parks from 1950 to 1956, stated: It is hard to imagine more dedicated people than those who run the parks. I have never met a single one that I would not be glad to meet again, and I have invariably regretted the time to say goodby. The range of their interests, their high intelligence, their devotion, makes them a separate and wonderful breed. (Hartzog 1988, 275) 129

PAGE 140

With greater sophistication and adoption of a new age cynicism, certain changes have perhaps taken place with respect to what would have been considered to be the National Park Service work ethic and culture. Historian William c. Everhart, recognizing these changes, expressed the following thought: More recently the Park Service has experienced troubling changes with the result that some of the vitality and the excitement of the earlier years seem to have worn away . . The old mystique is nearly dormant, and anyone who today asked an employee to go the extra mile without overtime pay for the "good of the Service" might well receive a mocking response. The energy and idealism that have characterized the Park Service for so long have not vanished, but they do seem to be the victims of a slight recession. (Everhart 1983, 3) Other aspects of the organizational culture of the National Park Service that has an effect upon alternative work schedules and productivity are (1) the esteem accorded to individuals who forfeit, rather than use, annual leave that would exceed the carryover limit of 240 hours, and (2) the aversion that is universally meted out to individuals who cannot accumulate, for one reason or another, extensive hours of unused sick leave. Until it became a prohibited practice by the Fair Labor Standards Act, it was also common for supervisors to expect even low-graded, non-supervisory employees to donate work beyond their normal work hours. 130

PAGE 141

Although there are obviously other cultural processes that could be mentioned, the association of productivity with hours on the job rather than actual output is the most significant with respect to this study. The National Park Service is one of the more decentralized bureaus in the Federal Government. Management policy is determined at the bureau level (Washington, D.C.), the regional level, and at the park or office level. Nothing can illustrate this better than management policies currently in effect with respect to approval of individual work schedules. The Director of the National Park Service supports the use of alternative work schedules, but through the delegations of authority, implementation is left to the Regional Directors. The Regional Directors of the Western and Pacific Northwest Regions have strongly urged managers and supervisors to authorize alternative work schedules whenever possible, but the remaining eight Regional Directors preferred to take a more cautious approach. The Southwest and Alaskan regions do not have a alternative work schedule plan, but they report "that they have been permitting AWS's for several years and found them to be useful. All of our parks and offices may have alternative schedules. At present, most of them 131

PAGE 142

have one or more employees on such a schedule. The Regional Director of the Southwest Region also stated, however, that "Employees are aware that such a work schedule is a privilege that can be easily taken from them if they abuse it." As of April 25, 1990 the Southwest Region had 1,210 employees, 58 of these were under a compressed work schedule (see Table 1.1 for all region/office authorizations). Two other regions, the Southeast and the MidAtlantic, and the Harpers Ferry Interpretive Planning Center, also do not have alternative work schedule plans. The Regional Directors of these two regions, and the Center Manager, have stated that alternative work schedules are "incompatible with service to the public and agency mission." Although a few compressed work schedules are authorized, these authorizations amount to only slightly more than 2 percent of their combined workforce. The Mid-Atlantic Region does "suffer and permit" flexitime. Compressed work tours are only reluctantly approved. No AWS plan has been prepared, and no specific accountability over use of these flexible work schedules exists. Except for the human resource employees who were participants in the experiments discussed in this paper, no supervisory reviews or overviews are performed. 132

PAGE 143

A new Regional Director was recently assigned to the North-Atlantic Region. The region still does not have a plan, but alternative schedules are now being authorized when they were formerly subject to routine rejection. The Regional Director of the Midwest Region does have an alternative work schedule plan and he has delegated the authority to members of his directorate and park superintendents to approve all schedules. No restrictions were expressed in the use of this authority except that the schedules must be "compatible with an office's work requirement." National Capital Region was slow to implement alternate work schedules. Prior to 1988, there were no management initiated alternative work schedule plans in any unit of the National Capital Region. Although at this time they have only a few plans in effect, they do give tacit support. Managers and supervisors cannot request an alternative work schedule "because of staff shortages." The Regional Director of the Rocky Mountain Region developed a plan in 1987, and authorized only the 5/4/9 and flexible eight-hour work schedule. She does not permit supervisory employees and specialists in one-of-akind jobs to work under alternative work schedules. Although she has been requested by an employee 133

PAGE 144

association within the Regional Office to permit supervisors to authorize compressed tours, she does not at this time see any reason to change the current policy. The Denver Service Center, which is served administratively by the Rocky Mountain Regional Office and headquartered in the same building, has an alternate work schedule plan and the manager of the Center has taken a very liberal approach with respect to use of nontraditional schedules. The Center is also seeking approval from the Department of Interior to implement a flexiplace schedule for a few of their professionals. The Office of the Director, along with two satellite offices located in Reston, Virginia and Denver, Colorado, does not have an alternative work schedule plan, but the individual supervisors have full delegated authority to authorize any individual work schedule they may feel is appropriate. From the above, it can be fairly well established that employee access to and effective use of alternate work schedules has been influenced significantly by management and management policy. And further, while several previous studies would seem to support the contention that alternative work schedules influence use of leave, absenteeism, and perhaps turnover, is also possible that organizational culture within the National 134

PAGE 145

Park Service may equally exercise a mitigating effect upon the use of leave and, thus, also perhaps motivated employees toward more positive work habits. The Independent Variables In the National Park Service, flexible work schedules were used prior to the Federal Employees Flexible and Compressed Work Schedules Act of 1978 to accommodate a variety of program objectives. Seasonal trail crews in the 1950s would regularly work 10 days on and four days off in order to minimize the amount of time spent travelling to and from back country work sites. The normal tour of conscripted fire crews were suspended and a first forty-hour work schedule was substituted to optimize the advantages from a massive first attack, and search and rescue teams were also placed on a first forty-hour work schedule. The predominate alternative work schedule in the National Park Service is flexitime (flexitour), a 5-day, variable 40-hour work week. All supervisory employees over the general schedule grade of GS-9, as well as any other employee who preselects a starting and ending time within acceptable limits, may adopt a flexible work schedule. Managers and supervisors are generally free to exercise a reasonable level of independence with respect to their work schedule --provided they meet previous 135

PAGE 146

agreed upon commitments. The second most frequently used alternative work schedule is the 5/4/9. This schedule is a moderate version of a compressed schedule whereby an employee will work 9 hours a day for eight days in one pay-period (two weeks), 8 hours for the ninth day, and then will take the tenth day as a lieu day (day off). The 5/4/9 and 4/10 distribution between Regions and Offices is identified'ln Table 5.1. Table 5.1 Compressed AWS by Regions and Offices As of April 25, 1990 Region/Office 5/4/9 4/10 Work Schedule Work Schedule Alaska 17 18 Mid-Atlantic 2 8 Midwest 57 65 North-Atlantic 25 70 National Capital 21 162 Pacific Northwest 80 104 Rocky Mountain 39 73 Southeast 29 49 Southwest 14 44 Western 131 477 DenverSC 16 83 Harpers Ferry 1 2 Washington Office 29 109 Total Employees 461 1,264 136 Percent of Total 2.03 .58 7.07 5.51 10.61 10.67 6.49 4.52 3.36 35.25 5.74 .17 8.00 100.00

PAGE 147

The Dependent Variables From the experiments and the questionnaire, four phenomena of interest (dependent variables) are addressed in this study: productivity, satisfaction, performance, and motivation. The primary focus of this study is the effect of alternative work schedules on productivity however, employee satisfaction, performance, and motivation, are essential ingredients to good employee performance and, accordingly, increased productivity. The questionnaire, for example, was designed to assess the state of the National Park Service's motivational system, and whether or not the availability of alternative work schedules have enhanced general employee job satisfaction, and if this, in turn has had a beneficial effect on productivity. The experiments directly measure productivity as it relates to the use, or non-use, of alternative work schedules. Productivity. Measurement of accomplishments in government is decidedly more cumbersome than it is in the private sector because of the lack of a true "profit and loss" statement and because of government's failure to properly identify what are truly critical tasks. If a person is busy all day accomplishing work assigned by a supervisor, but nothing that is mission worthy --should these efforts by that employee really be considered 137

PAGE 148

productive? The experiments selected for this study compare direct measures of productivity for individuals assigned (1) human resource management tasks from both experimental and control groups, and (2) facility maintenance tasks for an experimental and a control group at Rocky Mountain National Park. The human resource data was gathered from logs and registers that were maintained on recruitment actions, appointments, separations, and employee developmental programs. Maintenance data was obtained from the computerized Rocky Mountain National Park Maintenance Management system. Each maintenance project was assigned a work order number and equivalent projects were assigned each day by the assistant chiefs of maintenance. If an employee was keeping him/herself "busy," but not actually performing assigned product work, that work was not measured or included as productive work. Human resource data was collected manually from logs and registers maintained by the Division of Human Resources, Rocky Mountain Region, and the Division of Human Resources, Mid-Atlantic Region. The Mid-Atlantic's Division of Human Resources was the only office that experimented with flextime as opposed to compressed work schedules. Data was obtained by telephone through the Chief, Division of Human Resources, Rocky Mountain 138

PAGE 149

Region, for the other three regional offices. Maintenance data was obtained from review of MMS records. Performance. Each employee classified in a position below general schedule grade GM-13 is formally evaluated by hisjher supervisor at the close of the calendar year. The process is based on the preparation of performance standards and the evaluation of the employee's performance against these agreed upon standards. A graphic rating scale is applied, from level 1 to level 5, with level 1 representing unsatisfactory work performance and level 5 outstanding. The eventual product is an evaluation document prepared by the immediate supervisor, reviewed by the second-line supervisor, and discussed with the employee. A level 1 summary rating is tantamount to removal, and a level 5 (and occasionally a level 4) summary rating results in some form of monetary award. The performance evaluation process is a combination of objective as well as subjective supervisory perceptions. As an example, those individuals included in the experiments, and performing human resource management work, were evaluated on the quality and quantity of the products they generated individually or as a member of a team. They were also evaluated on their effectiveness as a communicator, a team member, and, 139

PAGE 150

although not specifically identified as such, their ability to "get along with people." It is more commonly the subjective evaluative characteristics to which employees take exception. The numeric summary level ratings for 1989 and 1990, contained in performance evaluations given to those employees participating in the experiments, were taken directly from regional performance evaluation reports. The questionnaire that was randomly distributed Servicewide contained five specific questions directly related to performance evaluation. Employee response to these questions has been used to measure, and subsequently record, general employee acceptance of the effectiveness and accuracy of the current process used to measure employee performance. Satisfaction. It has been fairly well established that a satisfied worker is not necessarily a high performer, nor is a high performer necessarily a satisfied employee. However, the fact that turnover in the National Park Service is extremely low, would certainly be one indicator of the existence of some rather positive factors related to employee job satisfaction. According to Herzberg (1959), achievement, recognition, responsibility, advancement, the work itself, and the possibility of growth, are all factors of 140

PAGE 151

a total set which he refers to as "satisfiers.'' Thus, the ability to select a preferred work schedule, the ability to work with a significant degree of independence, and to be recognized by supervisors and peers for individual efforts, should also be predictors of employee job satisfaction and/or attitudes towards the job and the workplace. In the questionnaire, a number of questions (presented as a matrix of items) were included that addressed employee perceptions of matters influencing general job satisfaction and/or their feelings regarding work and its environment. There are, however, two obvious limitations with respect to use of this data. It is assumed, first, that the respondents gave accurate and honest replies, and second, that the responses represent employees attitudes at a point in time. Motivation. While job satisfaction is an end-state resulting from attainment of either work or personal goals, motivation is primarily concerned with the individual's desires and how they can be fulfilled in the work situation. The opportunity for self-actualization, and/or the absence of such opportunity, was presented to employees in a series of questions that sought responses that measured the level of perceived fulfillment, the positive feelings that were held about his/her job, and 141

PAGE 152

as such, the content of the job itself. Initial Analysis Three analyses were used. Multivariate analysis of variance (MANOVA), analysis of covariance (ANCOVA), and analysis of variance (ANOVA). MANOVA was used to examine, as a function of the independent variable AWS, the three varieties of productivity measures assembled from the experiments; use of leave, employee performance evaluation, and assigned product completion. If differences existed between the experimental and control groups with respect to the demographic variables, the ANCOVA was used for this purpose. And finally, the ANOVA was used to measure productivity, and to evaluate the demographic attributes that were previously identified in an individual manner. Results of Preliminary Analysis Comparability of Demographic Variables for Experimental and Control Groups Analysis of the demographic data suggests that there are some statistical differences between the groups. Because of the sample size of the two primary experimental groups (Rocky Mountain National Park n=5 and Human Resources five region analysis n=37), a single 142

PAGE 153

sample t-test was used to compare mean age, sex, race, marital status, education, and years of employment in the organizational unit (see appendix D). No statistically significant difference was detected between groups on any variable except age. In addition to the single sample ttests, an ANOVA F-test, shown in Table 5.2, was used to determine the significance of difference between means. The lack of comparability in ages of group members, however, is not considered to represent a serious flaw because it is representative of a condition in the total workforce. For purposes of ease of comparison, mean ages for all groups are further represented by Figure 5.1, mean ages for all groups of human resource employees (control and experimental) that were selected with respect to their use of compressed schedules are depicted by Figure 5.2, and mean ages for the Mid-Atlantic Region flexitime experiment by Figure 5.3. Comparison of the Rocky Mountain National Park maintenance group is illustrated by Figure 5.4. 143

PAGE 154

Table 5.2. Age Comparisons for Experimental and Control Groups -F-Ratio Analysis Univariate Anal. Group Mean S.D. F-Ratio Human Resources: Compressed-Control 40.24 9.27 Compressed Exper. 45.97 10.30 Flexitime Control 36.00 10.25 Flexitime Exper. 38.63 12.30 AM Maintenance: Compressed Control 52.80 8.61 CompressedExper. 40.80 5.89 3.07 Figure 5.1. Ages -Human Resource Groups Appx P 0.014 14 ____, I ..... 0 .... al .Q E 4-::::. :c: 0 Ages21-25 Ages26-30 Ages31-40 Ages41-50 Ages51-60 0119r60 Human Resource Groups Control C Experimental 144

PAGE 155

Figure 5.2. Ages Compressed AWS Experiments Ages 21-25 Ages 26-30 AgM 31-40 Ages 41-50 Ages 51-flO Over60 Compressed AWS -Human Resource Groups D Control Experimental Figure 5.3. Ages MARO Flexitime Experiment Ages 21-25 Ages 31-40 Ages 41-50 Ages Flexitime MARO Human Resource Group 0 Control Experimental 145

PAGE 156

Figure 5.4. Ages RMNP AWS Experiment l I Age41-50 Age 51-60 Over60 RMNP Maintenance Control C Experimental Review of figures 5.1., 5.2., 5.3., and 5.4., does demonstrate rather clearly that although there are statistically significant differences in the demographic variable of age, that this difference is uniform in both the compressed and flexitime experiments, and that these groups each peak at the age 41-50 category. To determine whether or not age did influence productivity, ANCOVA was used to plot regression lines that would illustrate the effect of age on pre and 146

PAGE 157

posttest productivity. As can be discerned by review of the figure 5.5. scatterplot, the two down-slopes indicate that as control group individual age increased, individual productivity slightly decreased. Figure 5.6. demonstrates an opposite effect for the experimental groups, i.e., the two up-slopes indicate that higher productivity levels can be minimally attributed to higher individual age. .! 525 -c ::;:) 351 > :g ::J 175 0 ... 11. 0 0 0 I 25 0 0 0 Figure 5.5. Age -Productivity Correlation Human Resources Control Groups 0 0 .0 0 0 II o. 11!!1 0 S ,. 0 0 34 43 52 Ages of Employees 147 .-u

PAGE 158

Figure 5.6. Age Productivity Correlation Human Resources Experimental Groups 788 525 -0 c :::::> a-351 = :;; I .... ,_ (,) I :::s 175 "tJ 0 .. a. I 21 33 4s Ages of Employees The Effect of Alternative Work Schedules upon Productivity :a 0 57 ., 0 = ., ... !. CD !. As previously mentioned with respect to construct validity, the first approach was an exploration of the relationship between an employee work schedule and productivity through the comparison of production from selected experimental and control groups. The series of experiments that were conducted involve people who performed human resource management assignments (personnel) within five National Park Service Regions, and a group of maintenance employees (n=lO) at Rocky 148

PAGE 159

Mountain National Park. The research design provided for; (1) the pretesting of experimental and control groups one year prior and immediately before the intervention, (2) a selected AWS intervention, and (3) a posttest of the experimental and control groups. The first approach compares productivity of the selected experimental and control groups on a series of similar tasks, these comparisons are statistically described in table 5.3, which also provides for regional comparisons, and graphically illustrated by Figure 5.7. The second approach tests the productivity of the experimental and control group employees by other available measures (independent variables), i.e., individual performance evaluations, and use of sick and annual leave (See Appendix D) Table 5.3 Mean Productivity by Organization and Experimental/Control Groups Org./Group Mean(Pretest) Mean(Posttest) t p RMRO (E) 275.750 307.500 0.90 0.398 RMRO (C) 270.625 363.250 1.62 0.149 WRO (E) 294.125 347.750 0.90 0.397 WRO (C) 219.625 295.125 2.06 0.078 PNWRO (E) 268.846 285.308 0.39 0.704 SERO (C) 295.462 400.000 3.74 0.003 Compressed (E) 277.724 308.655 1.17 0.251 Control (C) 267.690 360.931 4.04 0.000 149

PAGE 160

Table 5.3. (contd.) Org./Group Mean(Pretest) Mean(Posttest) t p MARO (Flex-E) 224.000 236.00 0.31 0.767 MARO (C) 250.250 241.625 -0.19 0.852 RMNP (E) 376.000 323.400 -1.18 0.304 RMNP (C) 489.800 444.000 -0.67 0.542 The use of MANOVA to examine performance and productivity measures assembled from the human resource experiments, indicated a significant difference in F-ratio values only between human resource experimental and control groups associated with the compressed work schedule experiments (F = 4.89). The F-values depicted in table 5.4. for the other experimental and control groups were not statistically significant. Table 5.4. Productivity and Performance Results for Control and Experimental Group Employees Productivity Work Schedule Experimental Groups PrePostTest Test Compressed X 277.72 308.65 (142.3) S.D. (129.7) Control Groups PrePostTest Test 267.69 (143.1) 360.93 (124.4) 150 Multiple Analyses F Ratios for Interaction 4.89*

PAGE 161

Table 5.4. (continued) Experimental Control Multiple Analyses FWork Schedule PrePostPrePostRatios for Test Test Test Test Interaction Flextime X 224.00 236.00 250.25 241.63 0.07 S.D. (87.8) (110.1) (143.9) (126.4) -RMNP Maint. X 376.00 323.40 489.80 444.00 1.94 Compressed S.D. (94.8) (99.6) (169.9) (153.9) Performance Ratings Compressed X 298.35 305.07 306.62 298.10 0.49 S.D. (38.6) (40.4) (40.3) (37.8) Flextime X 269.13 284.75 271.88 311.25 1.21 S.D. (46.5) (24.3) (73.9) (51.3) -RMNP Maint. X 262.40 271.00 290.20 293.00 2.14 Compressed S.D. (35.5) (13.9) (27.9) (29.9) < .05 The differences that resulted within groups between the pretest and the posttest periods are particularly significant. The mean differences and subsequent t-tests presented in table 5.5. reveal some rather remarkable contrasts with respect to improvements and/or regressions in productivity and performance evaluations between pretest and posttest scores for experimental and control groups. Mean productivity for the human resource control 151

PAGE 162

group increased by 93.24 units, yet the mean performance evaluation decreased by 8.86 points. However, a different situation is noted with respect to the human resource flexitime control group whereas mean productivity decreased by 8.63 units, and the mean performance evaluation increased by 39.38 points. It would therefore appear that supervisory performance evaluations are not particularly reflective of changes in individual andjor group productivity. Table 5.5 T-Tests-Difference in Means Between Pretest and Posttest Scores Human Resources Compressed: (n = 29) Experimental T2-T1 Control T2-T1 Flextime: (n =8) Experimental T2-T1 Control T2-T1 RMNP Maintenance Compressed: (n-5) Experimental T2-T1 Control T2-T1 Mean 30.93 93.24 12.00 -8.63 -52.60 -45.80 Productivity -5.67 0.000 1.94 0.094 3.63 0.022 152

PAGE 163

Table 5.5 (contd.) Performance Mean _L_ ___f_ Human Resources Flextime: (n =8) Experimental T2-T1 15.62 Control T2-T1 39.38 -2.14 0.070 RMNP Maintenance Compressed: (n-5) Experimental T2-T1 8.60 Control T2-T1 2.80 0.29 0.783 At the close of the 1989 calendar year, the number of productive units completed by the Rocky Mountain National Park maintenance control/experimental groups and the human resource control/experimental groups were assembled. AWS programs were placed into effect at the beginning of the 1991 calendar year (the intervention), and the productive units completed by these same individuals/groups for the 1990 calendar year, prior to the AWS intervention, were again assembled. After one full year of experience under an AWS schedule, productive units were again totalled. By comparing productivity results from the second preceding year to reported productivity of the preceding year, and that of the AWS experimental year, an unexpected research outcome with respect to the Rocky Mountain National Park maintenance 153

PAGE 164

groups, was detected. This is illustrated by figures 5.7. and 5.8 .. Productivity for both Rocky Mountain National Park groups increased slightly in 1990 over that of 1989 --but both groups, experimental and control, declined in production in 1991, after the AWS intervention. on the other hand, the human resources groups productivity for 1989 and 1990 were quite similar, and productivity for both human resource groups increased in 1991 over 1990. Figure 5.7. Productivity Human Resource Groups 550 500 fl) 450 400 -0 ::;, "tl 350 300 1989 1990 1991 All Experimental Groups RMNP.CONTROL II RMNP-EXPR [J HRM-CONTROL HRM-EXPR 154

PAGE 165

550 500 II) 450 c: 400 0 :3 "tl 350 300 Figure 5.8. Productivity Compressed HR Experiments 1989 1990 1991 AWS Compressed Experiments RMNP..CONTROL RMNP-EXPR 0 HRM-CONTROL HRM-EXPR The experiment conducted within the Division of Human Resources Management, Mid-Atlantic Region, was the only experiment that examines the use of flexitime. Eight divisional employees, under a flexitime work schedule, and eight who were working under traditional schedules were monitored. These employees were performing basically the same tasks as those human resource employees selected in the other four regions, but inevitably, some minor differences can be observed between regions with respect to workflow and volume. Productivity gradually decreased for the MARO control 155

PAGE 166

group from 1989 to 1991 (note the gradual down slope illustrated by figure 5.9.), and also decreased for the experimental group from 1989 to 1990. Productivity did increase slightly from 1990 to 1991, after the flexitime intervention, to a point where the output of each group was nearly similar. By removing productivity of the MARO experiment from total AWS productivity, (illustrated by Figure 5.8.), mean productivity from compressed AWS experiments (figure 5.9.) is not significantly different from that of total AWS productivity means. Figure 5.9. Productivity MARO Flexitime Experiment 1989 1990 Flexitime Experiment MARO-Control MARO-Exper 156 1991

PAGE 167

It is suggested that more significant contrasts, with respect to the Mid-Atlantic experiment, would be found in initial appointment, supervision and training, and organizational culture. The mean employee longevity for the control and experimental groups were far lower than that found in the other four regions, yet no formalized effort to train new employees had yet been established. Supervisors failed to conduct required midyear employee evaluations, nor did they make other visible and active efforts to counsel and assist subordinates. As a consequence, a greater number of employees are required to accomplish personnel work programs than would have been necessary if the employees were adequately trained andjor committed. What the employees are providing is participation without any real content. Questionnaire/Survey Responses It had been initially determined that a sample size of 385 for an infinite population was appropriate. To allow for address changes, a total of 433 questionnaires were distributed to randomly selected National Park Service employees and supervisors in six regional and two service center offices. Of the 433 questionnaires, 388 were mailed to employees not working under an AWS, and 45 to employees working under an AWS. An addressed and 157

PAGE 168

stamped envelope was included with the questionnaire and explanatory cover letter to encourage the return of the survey document. Thirty-six questionnaires were returned because the employee had either moved or terminated (8 percent), 330 were returned completed (76 percent), and there were 67 partial non responses (16 percent). The basic thrust and purpose of the questionnaire was to sample employee attitudes and beliefs regarding their work, work schedules, performance, productivity, and supervision. Eliminating nondeliverables, the rate of return (83 percent) is considered good; nonetheless, it remains that 17 percent of the original random sample are not included in the data files. When the return rate is compared to that of similar governmental employee surveys, however, as illustrated in table 5.6, the difference does not appear to be significant. The questionnaire solicited information regarding the personal characteristics and work situation of the respondent, and then presented a series of closed-ended questions in a five-level Likert scale format requesting the respondent's opinions and feelings concerning work schedules, productivity, performance and other issues that are applicable to the work place and employee satisfaction. The average would thus vary between 1.0, 158

PAGE 169

Table 5.6 Response Rate of Employee Surveys Applicable to the NPS Percent Name and Date of Survey Questionnaires Returned OPM Fed. Empl. Attitude Survey of 1979 71.9 OPM Fed. Empl. Attitude Survey of 1980 75.7 OPM Fed. Empl. Attitude Survey of 1983 71.1 NPS Employee Survey of 1983 77.1 NPS Employee Spouse Survey of 1983 65.8 NPS Temporary Employee Survey of 1985 73.9 USGAO 1984-1985 Survey of AWS 75.0 MSPB Merit Principles Survey of 1986 77.0 MSPB Merit Principles Survey of 1989 74.3 indicating the lowest level of response, to 5.0, the highest level of response. However, it would not be correct to say, as an example, that if the mean response for a question was 4.6, that the average value would lie between "agree" and "strongly agree." Using the Newman-Keuls multiple comparison test, at the 0.05 significance level for personal characteristics and work situations, the reported value means for all work schedule groups reflected only minor differences. These differences were primarily not between employees who were under an AWS or not under an AWS, but between employees who were working under different AWS schedules. The only work situation which reflected significant 159

PAGE 170

opinion difference between AWS and non-AWS employees was with respect to preferred work schedule. The question as presented stated: "In terms of your overall personal situation, which of the following work schedules would you prefer to work under? (Circle one number) 1. 5-day, fixed 40-hour week 2. 4-day, fixed 40-hour week 3. flexitime (5-day, variable 40-hour week) 4. 5-4/9 plan s. permanent part-time 6. other (please indicate what kind) It is noted that 31.3 percent of the responding employees who reported that they were currently working under a non-AWS work schedule indicated that they would prefer the 5-day, fixed 40-hour week. However, 17.7 percent indicated that they would like to work a 4-day, fixed 40hour week, and the remainder (51 percent) professed an interest in flexitime or some other form of work schedule. On the other hand, only 5.8 percent of those employees who reported that they were working under some form of AWS work schedule expressed an interest in working a 5-day, fixed 40-hour week. Two of the appendices included at the end of this dissertation are applicable with respect to findings in this chapter. Appendix A is a summary of responses by non-AWS, AWS employees, and in total. Appendix B is a copy of the questionnaire as distributed. The 160

PAGE 171

questionnaire sought responses that could be summarized as attitudes towards; (1) the job and workplace, (2) work effort, (3) productivity and performance, (4) supervisors and performance appraisal, and (5) work schedules. A number of inferential statistical data analysis procedures were used and no significant differences were observed with respect to measured responses from AWS and non-AWS employees except with those responses that measured employee attitude towards work schedules. At the 0.05 significance level, the mean of the non-AWS employee responses and the means of all five AWS responses were not significantly different except for those responses provided for questions 9, 16, 17, 20, 21, 25, 31, and 32; all of which were applicable to measurement of employee attitude toward work schedules and their effect upon work productivity, performance, andjor the quality of the employee's life. When a single sample t-test was applied to each question within the "attitude toward work schedule" summary grouping comparing AWS responses to that of non AWS employees, calculated t and probability (p= > 0.005) resulted in a greater number of questions demonstrating statistical differences. An independent group summary analysis was also conducted using a non-parametric comparison module (Kruskal-Wallis), and this test also 161

PAGE 172

indicated that there is evidence to conclude that a significant difference in opinion with respect to the effectiveness of alternate work schedules exists between non-AWS employees and AWS employees. (See Appendix C). There were 20 questions presented that were applicable to attitudes regarding work schedules. The results, as displayed in table 5.7, indicate that although there may be significant differences between AWS and non-AWS employees with respect to individual questions, by use of the theme measure approach to responses, these differences were not statisically significant. Table 5.7 Theme Responses to the Employee Questionnaire Regarding Work Schedule Questions _I_ .. Approx-P Alternative Work Schedules: Questions 8-12, 17-19, 1.17 0.1648 Equal variance (328DF) 2.81 0.0050 Unequal variance (308DF) 2.85 0.0050 Work Schedule Influences: Questions 15-16, 20, 23-25 1.06 0.3427 Equal variance (328DF) -0.29 0.7710 Unequal variance (290DF) -0.29 0.7720 Traditional Work Schedules: 13-14, 21-22 1.23 0.0976 Equal variance (328DF) -4.72 0.0000 Unequal variance (277DF) -4.64 0.0000 162

PAGE 173

Hypothesis Testing To determine what the effect that alternate work schedules have upon productivity in the National Park Service, six research hypotheses were presented and examined in the context of each hypothesis being represented by an existing null and research theorem. Hypotheses one and two investigate, through the use of selected experimental and control groups, quantitative productive outcomes. Hypotheses three and four examine the use of annual and sick leave, and hypothesis five addresses turnover. Hypothesis six, the final hypothesis, compares the effect of alternate work schedules upon other dependent variables, i.e., morale and performance evaluation. Hypothesis Testing for Hypotheses l, 2, 3 and 4. Five separate experiments were conducted that measured (1) the productivity of 29 employees performing human resource (personnel) work within four regional offices and were assigned to a compressed work schedule; (2) the productivity of 29 other employees who were performing similar work, but under a traditional work schedule (a non-flexible 8-hour day, 5-day week); (3) the 163

PAGE 174

productivity of 8 employees performing human resource work in the Mid-Atlantic Regional Office assigned to a flexitime work schedule, and (4) the productivity of 8 other employees in the Mid-Atlantic Regional Office who were performing similar work, but under a traditional work schedule. A sixth experiment was conducted at a field installation (Rocky Mountain National Park) that measured the productivity of 5 maintenance employees assigned to a compressed work schedule, and 5 maintenance employees who were working under a traditional schedule. Hypothesis 1. This hypothesis involves a pretest and a posttest comparison of the previously identified control and experimental groups. Hypothesis 1 states the following: Productivity measurements for experimental (compressed work schedule employees) and control groups indicate that employees not under a compressed work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. Group productivity of employees performing human resource tasks in the four regional offices increased during the test period whether they were working under a compressed work schedule or a traditional work schedule. The productivity of the 29 employees assigned to compressed work schedules (5/4/9 and 4/10) increased by slightly more than 10 percent, however; the productivity 164

PAGE 175

of the 29 employees assigned to traditional work schedules increased by nearly 27 percent. To determine if this difference is statistically significant, several tests were applied. T-Test for Difference of Means. Appendix D provides a comparison of the equality of means for such variables as productivity, performance, use of leave, education, job experience, as well as certain personal attributes, i,e., age, sex, race, and marital status. Since no effort was made to assign individuals to groups, one of the consequences, or perhaps benefits, of this random selection was the occurrence of statistically significant differences in mean age between groups. This was a particularly pronounced with respect to the Rocky Mountain National Park Maintenance groups, where the mean difference in age between the experimental (compressed AWS) group and the control group was 12.8 years. It also had been noted earlier that age did exercise a slight negative effect upon productivity with respect to control groups. A p-value of 0.1129 was obtained through ANOVA in testing the regression relationship between experimental and control groups. Although this is not necessarily small enough to declare that a significant relationship exists, it's a determination of no critical difference; this observation with respect to the effect 165

PAGE 176

of age upon productivity will consequently be used as a mitigating factor in determining productivity comparisons. Thus far, it has been shown, primarily through ANOVA and T-Test, that productivity was not significantly different. Pretesting of productivity levels prior to the AWS intervention indicated that human resource employees associated with compressed AWS experimental groups (n=29) produced at a slightly higher mean output (277.724) than did their control counterparts who were assigned to a traditional work schedule (267.690). The experimental compressed maintenance group at Rocky Mountain National Park (n=5) reflected a lower pretest mean (376.000) than did the control group (489.000). The AWS interventions occurred on 1/1/91, and at the close of the 1991 calendar year the productivity levels (posttest) for 1991 were compared to that of 1990. Mean productivity increased for both groups. The experimental group productivity increased by 11.14 percent and the control group by 34.83 percent. Although the increases experienced by the experimental and control groups are certainly significant, Table 5.8 indicates that the equality of means between pretest and posttest is not statistically significant at the 95 percent confidence interval, but 166

PAGE 177

that the mean increase in productivity for the human resource groups is statistically significant. Table 5.8. T-Tests -Compressed AWS Productivity Pretest and Posttest Means Mean _I_ __f_ Human Resources Pretest: Experimental 266.108 Control 263.919 0.11 0.914 Posttest: Experimental 308.655 Control 360.931 -1.98 0.058 Increase /Decrease: Experimental 32.621 Control 97.012 -3.67 0.001 RMNP Maintenance Pretest: Experimental 376.000 Control 489.800 -2.68 0.055 Posttest: Experimental 323.400 Control 444.000 -2.71 0.054 Increase /Decrease Experimental -52.600 Control -45.800 1.87 0.135 Although pretest and posttest comparisons did not reflect (ANOVA) significant statistical differences, the productivity increase for the human resource control group from pretest to posttest was significantly greater than that of the experimental group. It is quite 167

PAGE 178

possible that this may in part be due to an observed greater teamwork effort on the part of control group employees. Hypothesis 2. This hypothesis involves a pretest and a posttest comparison of a human resource experimental, and a control group, in the Mid-Atlantic Regional Office of the National Park Service. Hypothesis 2 states the following: Productivity measurements for experimental (alternative eight-hour work schedules) and control groups indicate that employees not under an alternative eight-hour work schedule (control groups) demonstrate, on an average, higher productive levels during pretest-posttest time periods. To this point, it has been shown that group productivity of employees performing human resource tasks in the Mid-Atlantic Regional Office under a flexible work schedule increased during the test period, while the productivity of employees performing the same tasks under a traditional schedule during the test period decreased. The productivity of the 8 employees assigned to a flexible work schedule increased by slightly more than 5 percent, and the productivity of the 8 employees assigned to a traditional work schedule decreased by slightly more than 3 percent. To determine if this difference is statistically significant, those tests used with respect 168

PAGE 179

to Hypothesis 1 were also applied. T-Test for Difference of Means. As was the case with respect to Hypothesis 1, the research design provided for random selection of individuals. A significant mean age difference between the experimental and control groups was noted. The results of this test, t = 3.10 and p< 0.017 is further supported by an ANOVA test of the regression relationship, i.e., 0.0577 rsquare and p< 0.2095. Thus, the relationship of age between the control and experimental group is quite small, but that is significant only to the degree that age has an impact upon productivity. Table 5.9. provides an analysis of the Pearson r correlation coefficient and t and p values for pretest, posttest, and productivity increase/decreases. Although correlation and p demonstrates significant relationship for pretestjposttest between age and productivity for the experimental group, the increasejdecrease between pretest and posttest reflects a more significant relationship for the control group. The age and productivity relationship for the MARO flexitime experiment is graphically illustrated by figure 5.10 (pretest), figure 5.11 (posttest), and figure 5.12 (productivity increasejdecrease). 169

PAGE 180

Table 5.9 Age Productivity Correlation MARC Flexitime Experiment r_ _t __g_ Pretest: Experimental 0.1463 0.36 0.730 Control -0.7433 2.72 0.035 Posttest: Experimental 0.0182 0.04 0.966 Control -0.8058 3.33 0.016 Increase /Decrease: Experimental -0.3604 0.95 0.380 Control 0.1477 0.37 0.727 Figure 5.10 MARC Productivity Pretest 472 i 354 ., E i 236 111 I 0 0 I 31 37 43 49 55 Ages Experimental and Control 170

PAGE 181

Figure 5.11 MARO Productivity Posttest 382 (I) :t: 299 c :; 2U -... u ::I '8 133 ... a. i 51 0 0 31 37 43 49 55 Ages Experimental and Control Figure 5.12 MAROProductivity Increases/Decreases 62 I ., ........ 0 24 Q. 0 c 0 n ... S-14 c: f ::I ., E i II( w -91 31 37 43 e 55 Ages Experimental and Control 171

PAGE 182

As was the case with the experiments with compressed schedules, ANOVA and t-tests indicate the equality of means should not be rejected. Pretesting of productivity levels prior to the AWS intervention indicated that the MARO human resource employees working under a flexible work schedule (n=S) produced at a lower mean output (224.000) than did their control counterparts who were assigned to a traditional work schedule (250.250). The AWS intervention occurred on 1/1/91, and at the close of the 1991 calendar year the productivity levels (posttest) for 1991 were compared to that of 1990. Table 5.10 indicates that the equality of means is not statistically significant at the 95 percent confidence interval. Table 5.10 T-Tests Flexible AWS Productivity Pretest and Posttest Means Mean Human Resources MARO Pretest: Experimental Control Posttest: Experimental Control Increase /Decrease: Experimental Control 224.000 250.250 236.000 241.625 12.000 -8.625 -0.85 0.426 -0.14 0.889 1.94 0.094 172

PAGE 183

Although pretest and posttest comparisons did not reflect (ANOVA) significant statistical differences, the productivity increase for the human resource experimental group from pretest to posttest was significantly better than the decrease that was reflected by the control group. Hypotheses 3 and 4. These hypotheses involve a pretest and a posttest comparison of the use of annual and sick leave by control and experimental groups performing human resource assignments in five National Park Service Regional Offices and a experimental and control group performing maintenance work in Rocky Mountain National Park. Hypotheses 3 and 4 state the following: Use of annual leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups during the pretest-posttest time period. Use of sick leave for brief unscheduled periods is, on an average, less for employees in experimental groups than for those in control groups during the pretest-posttest time period. Unscheduled use of annual leave by both human resource and Rocky Mountain National Park maintenance employees assigned to an alternate work schedule was significantly greater than those employees assigned to a traditional work schedule during the test period. The 173

PAGE 184

mean use of sick leave by the AWS employees (experimental group) during the test period was also larger than that of employees who were working under traditional work schedules. However, the mean use of sick leave during the test leave year decreased for experimental group human resource employees from the mean of the prior two years and increased for control group employees. Experimental and control group employees at Rocky Mountain National Park reflected an increase in the mean use of sick leave over the mean of the prior two years. To determine if these differences are statistically significant, the following tests were applied. T-Test for Difference of Means. In order that annual leave may be property scheduled, each employee is asked toward the beginning of the leave year to provide to their supervisor their proposal for use of leave during the leave year. Annual leave that is not scheduled can only be taken if there are not existing and compelling reasons for the employee to otherwise remain on duty. Sick leave is never scheduled. Any use of sick leave is considered to be emergency need only. Employees (particularly the acknowledged career employees) will frequently resist any use of sick leave and request annual leave instead of sick leave. A possible reason for this position is that sick leave is not of the "use 174

PAGE 185

or lose" variety, and is creditable toward years of service at the time of retirement. Table 5.11 illustrates this point with respect to use of sick leave. Control and experimental groups reflect little statistical difference is usage during the test period although the unscheduled use of annual leave by experimental groups is significantly greater than that of the control groups. The small sample size of the maintenance group may have influenced the p outcomes. Table 5.11 Use of Annual and Sick Leave Mean T __f_ Unscheduled Annual Leave: Human Resources Experimental 46.243 Control 29.838 4.37 0.000 RMNP Maintenance Experimental 45.800 Control 25.800 7.24 0.002 Sick Leave Used: Human Resources Experimental 57.919 Control 53.324 0.76 0.451 RMNP Maintenance Experimental 80.600 Control 35.800 1.90 0.131 175

PAGE 186

Table 5.11 (contd.) Mean Mean Increase/Decrease Sick Leave: Human Resources Experimental Control RMNP Maintenance Experimental Control Hypothesis Testing, -2.108 7.784 30.600 8.400 for Hypotheses 5 and 6. -2.06 0.047 1.06 0.350 These two research hypothesis suggest that there is a relationship between employee performance, work satisfaction, and work schedules. The experiments that were conducted with human relations employees in five regional offices and the maintenance group at Rocky Mountain National Park will be used to examine the significance of differences in performance evaluations between experimental and control groups; and a questionnaire mailed to 433 randomly selected National Park Service employees will be used to determine the extent of attitudinal difference. Hypothesis s. This hypothesis requires an analysis of employee turnover prior to and after the AWS intervention. For purposes of the experiments, the 176

PAGE 187

number of separations that occurred during the test period were retained and compared between groups and for the preceding year. Hypothesis 5 states the following: Employment turnover is lower with respect to offices andjor parks who offer liberal work schedule opportunities (experimental), than it is for offices andjor parks who offer primarily only traditional work schedules (control). Table 5.12 identifies the number of National Park Service terminations in the 1990 and 1991 years, and the rate of turnover. A number of studies have reported that federal turnover is lower than that reported by private firms. In the federal government, the white collar quit rate for natural resource agencies (3.2 percent) in 1984 was significantly less than that for all types of work (4.9). Table 5.12 Terminations 1990 and 1991 1990 % 1991 % Retirements 43 7.6 36 5.9 Deaths 16 2.8 25 4.1 Terminations 504 89.6 548 90.0 (AWS Terminations) (94) 18.7 11.1ID 20.6 Total 563 100.0 609 100.0 NPS Employment 17,923 3.1 18,161 3.4 177

PAGE 188

A recent study conducted by the Mid-Atlantic Regional Office indicated that out of 37 employee complaints, only one even mentioned a problem with work scheduling. During the period of the AWS intervention testing, only 3 out of the 84 participating employees terminated (2 experimental and 1 control), and that occurred at the end of the test period. There does not appear to be any correlation between the availability or nonavailability of AWS and turnover in NPS employment. The source of the raw information which was summarized in table 5.12 would suggest that geographic location andjor the absence of large metropolitan employment markets within the region is the primary factor rather than AWS policies. Hypothesis 6. This hypothesis deals with two primary issues, employee performance and the collective attitude of the workforce toward their work environment. Performance evaluation rating were obtained for those 84 employees who participated as a member of an experimental or control group for the evaluation period ending prior to the AWS intervention, and at the close of the intervention year. In addition, a questionnaire was mailed to 433 employees; 260 who were not working under 178

PAGE 189

an AWS, and 173 to employees working under an AWS. The questionnaire sought responses that would identify employee attitudes toward; (a) job and workplace, (b) work effort, (c) productivity and performance, (d) supervisors and performance appraisals, (e) consequences of performance, and (f) work schedules. Hypothesis 6 states the following: Employees who work under an alternative work schedule (experimental group) tend to have higher morale, and, on an average, higher performance evaluations than those employees who work under more traditional work schedules (control group). The mean group performance of employees performing human resource tasks in the five regional offices increased whether they were working under an AWS or a traditional work schedule. To determine if this increase was statistically significant several tests were applied. T-Test for Difference of Means, Experimental and Control Groups. Consistent with observed past employee performance evaluation practices, only modest changes in valuative ratings occurred at the end of the year after the AWS intervention. In a few individual circumstances there were some rather extreme changes, but this was principally in the form of adjustments to a prior year preemptive rating. Table 5.13 indicates that the 179

PAGE 190

equality of means between pretest and posttest is not statistically significant at the 95 percent confidence interval. The only exception was the mean performance evaluation posttest for the Rocky Mountain National Park maintenance group. Table 5.13 TTests Performance Evaluation Pretest and Posttest Means Mean Human Resources Pretest: Experimental 292.027 Control 299.108 -1.04 0.307 Posttest: Experimental 300.676 Control 300.946 -0.04 0.966 Increase /Decrease: Experimental 5.514 Control 2.000 0.78 0.442 RMNP Maintenance Pretest Experimental 262.400 Control 290.200 -1.75 0.155 Posttest: Experimental 271.000 Control 293.000 -3.55 0.024 Increase /Decrease Experimental 8.600 Control 2.800 0.29 0.783 T-Test for Difference of Means, Questionnaire Responses. The questionnaire that was mailed to the 180

PAGE 191

randomly selected NPS employees contained 46 questions which could be grouped into the following categories: 1. Attitude toward job and workplace. 2. Attitude toward work effort. 3. Attitude toward productivity and performance. 4. Attitude toward supervisors and performance appraisal. S. Attitude toward work schedule. a. alternate work schedules. b. traditional work schedules. c. work schedule influences Appendix A provides an index of the questions which form the basis of categories 1-S. Table S.7 provides an index of the questions which form the basis of categories Sa-Sc. Table S.14 provides an ANOVA group analysis of each category. Questionnaire responses from AWS and non-AWS employees do not differ significantly with respect to categories 1, 2, 4, S and Sc. Statistically significant differences were recorded with responses to questions that have been placed in categories 3, Sa and Sb. Category 1. AWS Non-AWS Category 2. AWS Non-AWS Table 5.14 TTest and AN OVA Independent Group Analysis Summary Group Mean 3.6368 3.6687 3.6752 3.7206 f 1.08 1.08 181 t -0.32 0.753 -0.47 0.640

PAGE 192

Table 5.14 (contd.) Group Mean _f_ __t 3. AWS 3.1613 Non-AWS 3.5298 1.12 -3.33 0.001 4. AWS 3.3030 Non-AWS 3.3812 1.05 -0.68 0.495 5. AWS 3.1698 Non-AWS 3.1604 1.06 0.08 0.934 Sa. AWS 3.5915 Non-AWS 3.1614 1.01 2.96 0.003 Sb. AWS 2.5525 Non-AWS 3.0888 1.23 -4.72 0.000 Sc. AWS 2.8647 Non-AWS 2.8958 1.06 -0.29 0.770 Category 3 is represented by questions 28, 29, 43 and 44, which seek responses with respect to work as a matter of personal satisfaction, and the quality of the evaluation they receive from their supervisor for their personal accomplishments. The p value for Q28 and Q43 were 0.458 and 0.255. The p for Q29 "To get my work done I usually work more than 40 hours per week," was 0.018, with a mean value of 3.911 for non-AWS responses and 3.667 for AWS. The p for Q44, 11There is a tendency for supervisors here to give the same performance ratings 182

PAGE 193

regardless of how well people perform their jobs," was 0.03S, with a mean value of 3.341 for AWS responses and 3.146 for non-AWS. Category Sa is represented by Questions 8-12, 17-19, and 31-32. These 10 questions seek responses with respect to the use of other than a traditional work schedule. It was expected that there would be a significant difference in response between the AWS and non-AWS employees. The p values for all questions were 0.000, except for Q11 "I would bejam very productive under a compressed schedule," which was 0.001. The mean values were all substantially higher for AWS respondences except for Q8 "Alternative work schedules (e.g., flexitime, compressed work week) tend to hinder productivity in this organization," the mean value for non-AWS was 2.S36, and the mean for AWS was 2.109. These both are comparatively low mean scores, which would indicate that their is not much general belief that AWS hinders productivity. Category Sb relates to employee responses to questions 13-14, and 21-22. These four questions relate to attitudes about the use of traditional work schedules. As was the case with category Sa, it was expected that significant differences would emerge from the evaluation of AWS and non-AWS responses. The p values for all 183

PAGE 194

questions were 0.000. The mean values of responses were all substantially higher from non-AWS employees. Based upon the above analysis, the only area of difference between AWS and non-AWS responses were the personal perceptions regarding the acceptability of traditional or alternate work schedules and the quality of employee performance evaluations. There is no evidence to suggest that morale/employee satisfaction is higher for either AWS or non-AWS employees. Valuative ratings for performance applicable to the six experimental/control groups were not statistically different. Summary of the Findings Analysis of the results from the six experiments and the employee questionnaire leads to the following findings with respect to the six hypotheses which were the focus of this study: Hypothesis 1. Employees not under a compressed work schedule demonstrate higher productive levels during the pretest-posttest time period. Finding. Although the human resource employee control group (non-compressed work schedule) did demonstrate higher performance levels during the pretest 184

PAGE 195

-posttest time period, the difference was not statistically significant. Hypothesis 2. Employees not under a flexible work schedule demonstrate higher productive levels during the pretest-posttest time period. Finding. The production of non-flexible work schedule employees decreased during the pretest-posttest work period. A slight productivity increase occurred with respect to flexible work schedule employees. However, the results were not statistically significant. Hypothesis 3. Employees who work under an AWS use less annual leave for brief unscheduled periods during the pretest-posttest time period. Finding. The use of unscheduled annual leave during the pretest-posttest time period was significantly greater by employees who worked under an AWS. Non-AWS employees submitted more accurate projections of their proposed use of annual leave. Hypothesis 4. Employees who work under an AWS use less sick leave during the pretest-posttest time period. Finding. Although employees under an AWS used more sick leave than non-AWS employees during the pretest-posttest time period, the difference was not statistically significant. Non-AWS employees experienced a slight 185

PAGE 196

increase in mean use of sick leave during the test period over that of the mean of two preceding years. A slight decrease did occur in mean use from the preceding period for the AWS employees. Hypothesis 5. Employee turnover is lower for those offices who offer liberal work schedule opportunities. Finding. Employee turnover is not affected by work schedule policies. Since turnover is extremely low, it appears that geographic location and presence of large metropolitan areas exercise a greater impact on employee turnover. Hypothesis 6. Employees who work under AWS have higher morale and, on an average, receive higher performance evaluations Finding. At the 95 percent confidence interval, performance evaluation ratings were not significantly different between AWS and non-AWS employees. Analysis of questionnaire responses also do not indicate that any significant difference exists between the morale of AWS and non-AWS employees. 186

PAGE 197

CHAPTER VI CONCLUSIONS AND SUMMARY The initiative for implementing alternative work schedules was taken in the late 1960's by the European industrial community. Initially built around a concept of a "gliding time" which would avoid rigid starting and stopping times, flexitime was later modified to add the idea of a "core" time, a work period that all employees would be expected to be present. With the expansion of flexitime to the United States and other countries, the use of compressed work schedules and other variations in the customary workday and workweek schedule evolved. Currently, experiments are being conducted by the Federal Government to consider the feasibility of authorizing "flexiplace" work programs; permitting certain employees to work at home or in other locations to eliminate commuting problems and mitigate office space requirements. Alternative work schedules are now considered by many to be a right of employment. Union negotiations will cover the availability of other than traditional (fixed) schedules, and once management has conceded their rights with respect to work scheduling, it would be

PAGE 198

extremely difficult, if not impossible, to regain this rights and limit the use of alternative work schedules. The Federal Government initially authorized the use of alternate work schedule arrangements in 1978, and, since that time, nearly every agency has approved their use. The National Park Service is the largest bureau in the Department of Interior. In 1990 it had nearly 18,000 permanent employees and required a near equal number of temporary employees to manage its seasonal operations. Alternative work schedules were approved without adequate research or investigation as to the effect that approval of such schedules would have upon the specific work programs involved. The effect of alternative work schedules upon productivity in the Federal Government has been limited, but the body of research and literature available suggests that AWS's can lead to; (1) improved employee work satisfaction, (2) higher morale, (3) decreased absenteeism, and (4) reduction in employee turnover. Significant differences exist between regions, offices, and park areas, not only with regard to the approval, use, and management of alternate work schedules, but also with respect to management environment and organizational culture. In order to consider the effect of alternative work schedules on 188

PAGE 199

productivity, performance, leave use, employee turnover, and work satisfaction, six hypotheses were framed and were tested through the use of experiments and an employee questionnaire. Work Schedule Productivity Results Two specific work schedule alternatives to the traditional fixed hour five-day week were examined; the compressed work schedule (4 10-hour days, and 5/4/9), and flexitime. Experimental and control groups involving human resource employees were established in five National Park Service regional offices. In addition, a control/experimental experiment was conducted in a National Park Service field unit (National Park) involving maintenance employees. The results of these experiments indicated that there were no significant statistical differences between control and experimental groups with respect to productivity. Reports from pretest-posttest productivity summations indicated that productivity increased during the test period for both groups, and although a comparison of production did not permit a finding of significant statistical difference, it is noted that the extent of increase was greater for the human resource 189

PAGE 200

control group. The maintenance groups each experienced a decrease in productivity during the test period. One antecedent variable exercised some possible influence in the outcomes, and that was employee age. As individual age increased for members of the control group, individual productivity slightly decreased. This phenomenon was not observed with respect to experimental group employees. Based on the analysis of the productivity outcomes, it is concluded that perhaps other factors, i.e., intervening variables, such as managerial andjor supervisory style may have equally contributed to the the measurable productivity and performance outcomes from the observed work schedule experiments. Usage of Sick and Unscheduled Annual Leave All sick leave usage is considered to be unscheduled. Although individuals would request from time-to-time sick leave in advance to attend to routine dental or medical examinations, use is more normally associated with an unplanned illness or medical emergency. To avoid comparison with a possible nontypical prior year cG .. of sick leave, the mean of the previous two years was used to test equality of means 190

PAGE 201

between prior year usage and the year of the AWS intervention. It was expected that the need for sick leave would have been less for AWS employees who perhaps had greater opportunities to schedule their medical appointments during off duty hours. Instead, the findings revealed that AWS employees used slightly more sick leave than non-AWS employees during the pretest-posttest period; however, this difference was not statistically relevant. Since unused sick leave is creditable toward length of service, it is certainly possible that use of sick leave would be, in part, mitigated by its possible future value at the time of retirement. However, this potential future value would certainly be considered by both AWS and non-AWS employees. Employees are requested to prepare some form of schedule projecting their use of annual leave throughout the leave year. This is to permit supervisors and managers to adjust the requests, wherever possible, to avoid "bunching" or "clumping" of prolonged employee absences. AWS employees apparently were not as successful, or perhaps as concerned, as non-AWS employees in developing their leave schedules as statistically significant differences did result because of the AWS employee's expanded use of unscheduled annual leave. The 191

PAGE 202

unscheduled use predominately occurred toward the close of the leave year where an underscheduling led to a "use or lose" situation, and the supervisors were requested to approve annual leave over the last few weeks of the leave year. Employee Turnover in Employment Employee turnover in the National Park Service is remarkably low (3.4% in 1991). A study that was conducted in the Mid-Atlantic Region with respect to employee complaints indicated that out of 37 employee complaints, only one even mentioned a problem with work scheduling. There were 84 employees involved in the experiments, 3 terminated at the end of the experimental period, 2 from experimental groups and 1 from a control group. The reasons for the terminations were not related to the experiments or their work schedules. No apparent correlation between termination and work schedules could be uncovered, and this included the consideration of whether or not the NPS office took a liberal or conservative approach to the availability of AWS. Employee Performance Evaluation and Morale Two primary issues are involved here. Employee satisfaction with conditions and circumstances 192

PAGE 203

surrounding their work environment, and a recorded numeric evaluation of their performance by their supervisor. The six experiments were used to determine if there were significant statistical differences between performance evaluations received by AWS employees and non-AWS employees; and an employee questionnaire mailed to 433 randomly selected NPS employees was used to assess employee morale/satisfaction. Testing of performance evaluation ratings did not reveal that there were any significant statistical difference in those performance ratings given to experimental and control group employees at the close of the experimental performance year. Pretest and posttest means were near identical, each group increased their mean valuation; the experimental group slightly more. It is interesting to note that one of the questions contained in the questionnaire asked if the performance rating presented a fair and accurate picture of actual job performance. Through use of the Newman-Keuls Multiple Comparisons Summary, at the 0.05 significance level the rather positive means of all work schedule groups were not significantly different. The questionnaire contained 46 questions which were grouped into 8 categories: 193

PAGE 204

1. Attitude toward job and workplace. 2. Attitude toward work effort. 3. Attitude toward productivity and performance. 4. Attitude toward supervisors and performance appraisal. 5. Attitude toward work schedule. a. alternate work schedules. b. traditional work schedules. c. work schedule influences Questionnaire responses from AWS and non-AWS employees did not differ significantly with respect to categories 1, 2, 4, 5 and Sc. Mean responses were all on the high side indicating comfort and satisfaction with current conditions. Statistically significant differences were recorded with responses to questions that were placed in categories 3, Sa and Sb. Category 3 contained four questions which sought responses with respect to work as a matter of personal satisfaction, and the quality of the evaluation their supervisor gave them for their personal accomplishments. Two of the questions did not result in any significant difference between AWS and non-AWS employees, but the responses to the other two were significantly different. Q29 (p = 0.018) "To get my work done I usually work more than 40 hours a week," was responded to more affirmatively by non-AWS employees. This outcome was not unexpected since AWS employees seemingly observe the limits of their work schedule hours more closely than non-AWS employees -particularly those who may be working 194

PAGE 205

a compressed tour. Q44 (p = 0.03S) "There is a tendency for supervisors here to give the same performance ratings regardless of how well people perform their jobs," was responded to more affirmatively by AWS employees. Discussions with several supervisors of employees involved in the experiments indicated that they felt that they were somewhat restricted by management in their ability to award outstanding performance ratings. In addition, they admitted that they were reluctant to lower ratings from the previous year (based upon employee performance), because of the potential they may have to respond to formal complaints andjor grievances. Categories Sa and Sb pertain to attitudes about traditional work schedules (Sb) and AWS work schedules (Sa). As may be expected, those employees working under an AWS responded more positively to category Sa, and those working under a traditional work schedule responded more positively to category Sb. Employees who are working under an AWS feel quite strongly about their continued use of these schedules. There is no evidence that their morale has been significantly enhanced because they are working under an AWS, but it is clear that it would be significantly diminished if the AWS would not be permitted. The assignment of a new Regional Director to the Mid-Atlantic 19S

PAGE 206

Regional Office, who are now actively using flextime schedules, was a matter of considerable employee concern as the new Regional Director had a reputation for disliking AWS's. Initial efforts of the Regional Director to reduce the number of flexitime schedules for supervisory employees led to a number of complaints, and a quite obvious reduction in organizational effectiveness. The results of this study and the findings suggest that productivity is neither enhanced nor diminished by the use of alternate work schedules. The use of leave is a situational condition. It is thus more associated with personal circumstances and perceptions of need than perhaps driven by use of any particular work schedule. The National Park Service does not experience a problem with employee turnover. Among Federal agencies it is valued as an employer, and although AWS's are similarly valued, it has not been demonstrated that an employee has terminated their employment over work schedule policies. Morale andjor employee satisfaction is consistently high, whether or not an employee may be working under an AWS or a traditional work schedule. The availability of AWS has perhaps reached the point to where they are considered to be a condition of employment, each person 196

PAGE 207

seeking out a schedule which satisfies their particular needs. Responses to Q1 of the questionnaire, "In general, I am satisfied with my job," resulted in an very high mean (4.170 on a 1 to 5 scale, p = 0.662) attesting to the fact that employees are indeed well satisfied. In the 1960s the rapid growth of knowledge in the social sciences, particularly that which concerns human behavior in the work environment, significantly changed the way we viewed leadership, direction, and participative management. What also emerged from the sixties, along with the utilization of new styles of democratic supervision, was an ever increasing need for more research as to what are the effects from this rather large body of motivational innovations on organizational and employee performance. The initial purpose of this research project was to confront what appeared to be a vacuous approach by the National Park Service in the implementation of alternative work schedules and consider what the effect these schedules may have on productivity. Although eventual findings supported literature contentions that alternative work schedules neither enhance or diminish productivity, some other effects and observations with respect to the use of alternative work schedules became apparent. Organizational attendance, for example, did not improve through use of alternative 197

PAGE 208

work schedules. On an average, AWS employees used more sick and unscheduled annual leave than non-AWS employees. It was also quite interesting to note that increases or decreases in employee productivity exercised little influence over individual performance evaluations. In visiting the various offices and areas that participated in AWS experiments, it seemed that employees who aggressively sought and preferred alternative work schedules also exhibited different perspectives on work and the work environment than did those employees who preferred the more traditional work schedules. In many instances this outlook affected the way that they approached the whole question of an appropriate work ethic for their office. In a regional office where an alternative work schedule was recently introduced, the announced reassignment of the regional director and his subsequent replacement with an individual who was known as an opponent of alternative work schedules, exacerbated an existing labor problem and further fueled what was a growing distrust of management by employees. Much of the literature on why alternative work schedules are implemented will generally cite a number of organizational and personal benefits. However, organizational benefits appear to be largely speculative and rely heavily upon such factors as increased employee 198

PAGE 209

morale, motivation, and perhaps a justifiable need to be competitive in the employment market. Employees, at least those in the National Park Service, now view alternative work schedules as a right of employment, and are not easily convinced that a work schedule must meet other than their physiological, psychological, andjor emotional needs. To be effective, however, work schedules must be adapted to a variety of forces including; employee performance, organization mission and objectives, and inter-group coordination and communications. It is quite evident that further research is needed to resolve the continued questions of whether or not many of the benefits from alternative work schedules are short lived in nature or perhaps of primary benefit only to the organization's employees. Introduction of alternative work schedules can also introduce in the work environment other more ubiquitous and unanticipated outcomes such as; a lack of group cohesiveness, unacceptance of organizational values, and work ethic revisionism. Research should therefore also seek answers to the complex and multidimentional process of implementing alternative work schedules. 199

PAGE 210

Appendix A Reviewer's Guide to Questionnaire Percent Strongly Strongly Disagree Disagree Undecided Agree Agree Attitude Towards Job and Workplace: Q-1 Total 0.9 4.5 6.1 53.6 34.8 Non-AWS 1.0 4.2 6.3 52.6 35.9 AWS 0.7 5.1 5.8 55.1 33.3 Q-2 Total 0.3 2.1 2.7 47.0 47.9 Non-AWS 0.5 2.1 1.6 46.4 49.5 AWS 0.0 2.2 4.3 47.8 45.7 Q-3 Total 0.6 2.1 8.5 33.3 55.5 Non-AWS 0.5 0.5 8.3 34.4 56.3 AWS 0.7 4.3 8.7 31.9 54.3 Q-35 Total 10.0 46.7 21.8 13.3 8.2 Non-AWS 9.9 49.0 21.4 11.5 8.3 AWS 10.1 43.5 22.5 15.9 8.0 Q-36 Total 5.8 33.9 31.5 23.9 4.8 Non-AWS 6.8 35.4 28.1 26.0 3.6 AWS 4.3 31.9 36.2 21.0 6.5 Q-37 Total 1.8 10.6 15.5 60.6 11.5 Non-AWS 1.6 10.9 14.6 61.5 11.5 AWS 2.2 10.1 16.7 59.4 11.6 Q-38 Total 3.6 12.4 17.0 56.1 10.9 Non-AWS 3.6 9.9 17.2 59.9 9.4 AWS 3.6 15.9 16.7 50.7 13.0 Q-39 Total 4.5 15.2 17.6 53.6 9.1 Non-AWS 4.7 13.0 16.7 55.2 10.4 AWS 4.3 18.1 18.8 51.4 7.2 Attitude Towards Work Effort: Q-4 Total 0.0 0.0 3.0 34.2 62.7 Non-AWS 0.0 0.0 3.1 30.2 66.7 AWS 0.0 0.0 2.9 39.9 57.2 200

PAGE 211

Percent Strongly Strongly Disagree Disagree Undecided Agree Agree Attitude Towards Work Effort: (Continued) Q-5 Total 0.9 7.0 17.0 51.8 23.3 Non-AWS 0.5 6.8 17.2 50.5 25.0 AWS 1.4 7.2 16.7 53.6 21.0 Q-6 Total 7.0 32.7 32.1 20.3 7.9 Non-AWS 7.8 32.8 30.7 20.8 7.8 AWS 5.8 32.6 34.1 19.6 8.0 Q-41 Total 3.9 14.2 19.4 54.5 7.9 Non-AWS 3.1 14.1 20.3 54.7 7.8 AWS 5.1 15.2 17.4 54.3 8.0 Q-42 Total 2.4 9.7 22.1 52.7 13.0 Non-AWS 2.1 8.9 21.4 56.3 11.5 AWS 2.9 10.9 23.2 47.8 15.2 Attitude Towards Productivity and Performance: Q-28 Total 0.3 1.2 6.7 54.8 37.0 Non-AWS 0.0 1.0 5.2 57.8 35.9 AWS 0.7 1.4 9.4 50.0 38.4 Q-29 Total 2.4 16.1 14.2 33.0 34.2 Non-AWS 0.5 17.2 9.9 35.4 37.0 AWS 5.1 14.5 19.6 30.4 30.4 Q-43 Total 5.8 15.5 21.2 48.8 8.8 Non-AWS 5.2 14.6 19.3 53.1 7.8 AWS 6.5 16.7 23.9 42.8 10.1 Q-44 Total 4.2 26.1 24.8 32.4 12.4 Non-AWS 5.2 28.6 23.4 31.8 10.9 AWS 2.9 22.5 26.8 33.3 14.5 Attitude Towards Supervisors and Performance Appraisal: Q-40 Total 7.3 14.8 17.3 48.5 12.1 Non-AWS 7.3 14.1 15.1 49.5 14.1 AWS 7.2 15.9 20.3 47.8 8.7 Q-41 Total 3.9 14.2 19.4 54.5 7.9 Non-AWS 3.1 14.1 20.3 54.7 7.8 AWS 5.1 15.2 17.4 54.3 8.0 201

PAGE 212

Percent Strongly Strongly Disagree Disagree Undecided Agree Agree Attitude Towards and Performance (Continued) Q-42 Total 2.4 9.7 22.1 52.7 13.0 Non-AWS 2.1 8.9 21.4 56.3 11.5 AWS 2.9 10.9 23.2 47.8 15.2 Q-45 Total 10.9 20.0 24.2 37.0 7.9 Non-AWS 9.9 22.4 20.3 39.6 7.8 AWS 12.3 16.7 29.7 33.3 8.0 Q-46 Total 7.0 20.9 33.0 35.2 3.9 Non-AWS 6.3 21.9 28.6 39.6 3.6 AWS 8.0 19.6 39.1 29.0 4.3 Attitude Towards Work Schedules: Q-7 Total 15.2 34.2 26.4 17.9 6.4 Non-AWS 15.1 38.0 25.5 14.6 6.8 AWS 15.2 29.0 27.5 22.5 5.8 Q-8 Total 23.6 40.0 19.1 11.5 5.8 Non-AWS 17.2 38.0 25.0 13.5 6.3 AWS 32.6 42.8 10.9 8.7 5.1 Q-9 Total 2.4 10.0 23.0 47.0 17.6 Non-AWS 3.1 12.0 27.1 46.4 11.5 AWS 1.4 7.2 17.4 47.8 26.1 Q-10 Total 3.0 9.4 20.6 39.4 27.6 Non-AWS 2.1 14.1 26.6 38.5 18.8 AWS 4.3 2.9 12.3 40.6 39.9 Q-11 Total 1.2 10.0 28.2 33.6 27.0 Non-AWS 1.0 9.9 34.9 33.9 20.3 AWS 1.4 10.1 18.8 33.3 36.2 Q-12 Total 0.6 3.0 23.6 43.3 29.4 Non-AWS 0.0 5.2 30.2 39.6 25.0 AWS 1.4 0.0 14.5 48.6 35.5 Q-13 Total 10.3 30.0 39.7 15.8 4.2 Non-AWS 4.7 27.1 41.7 22.4 4.2 AWS 18.1 34.1 37.0 6.5 4.3 202

PAGE 213

Percent Strongly Strongly Disagree Disagree Undecided Agree Agree Attitude Towards Work Schedules: (Continued) Q-14 Total 5.5 16.1 28.2 44.2 6.1 Non-AWS 2.1 9.9 29.2 54.7 4.2 AWS 10.1 24.6 26.8 29.7 8.7 Q-15 Total 19.1 48.5 20.3 9.4 2.7 Non-AWS 10.4 50.5 24.0 10.9 4.2 AWS 31.2 45.7 15.2 7.2 0.7 Q-16 Total 22.4 37.6 31.8 6.1 2.1 Non-AWS 15.6 37.0 39.1 5.7 2.6 AWS 31.9 38.4 21.7 6.5 1.4 Q-17 Total 6.7 26.1 40.3 20.0 7.0 Non-AWS 7.8 30.7 38.5 18.2 4.7 AWS 5.1 18.8 42.8 23.2 10.1 Q-18 Total 4.8 21.8 38.5 27.0 7.9 Non-AWS 5.2 26.6 40.1 22.4 5.7 AWS 4.3 15.2 36.2 33.3 10.9 Q-19 Total 4.5 22.1 37.0 27.3 9.1 Non-AWS 4.7 25.5 42.2 19.3 8.3 AWS 4.3 17.4 29.7 38.4 10.1 Q-20 Total 11.5 35.2 32.4 17.0 3.9 Non-AWS 5.7 34.4 35.9 19.8 4.2 AWS 19.6 36.2 27.5 13.0 3.6 Q-21 Total 12.7 41.8 19.4 21.5 4.5 Non-AWS 6.3 37.5 22.4 28.6 5.2 AWS 21.7 47.8 15.2 11.6 3.6 Q-22 Total 12.4 35.5 17.0 30.6 4.5 Non-AWS 7.8 30.2 18.2 38.5 5.2 AWS 18.8 42.8 15.2 19.6 3.6 Q-23 Total 8.8 41.2 31.2 14.5 4.2 Non-AWS 8.9 50.0 26.6 11.5 3.1 AWS 8.7 29.0 37.7 18.1 6.5 Q-24 Total 3.0 12.1 30.6 39.7 14.5 Non-AWS 3.1 16.1 32.8 35.4 12.5 AWS 2.9 6.5 27.5 46.4 16.7 203

PAGE 214

Percent Strongly Strongly Disagree Disagree Undecided Agree Agree Attitude Towards Work Schedules: (Continued) Q-25 Total 0.6 8.2 18.8 43.0 29.4 Non-AWS 0.5 8.3 26.0 44.3 20.8 AWS 0.7 8.0 8.7 42.0 40.6 Q-26 Total 2.4 17.3 22.4 34.8 23.0 Non-AWS 1.6 20.3 19.8 34.9 23.4 AWS 3.6 13.0 26.1 34.8 22.5 Q-27 Total 13.9 35.2 21.8 24.2 4.8 Non-AWS 12.5 33.3 21.9 26.6 5.7 AWS 15.9 37.7 21.7 21.0 3.6 Q-32 Total 4.2 19.4 22.7 33.0 20.6 Non-AWS 4.2 24.5 27.1 30.7 13.5 AWS 4.3 13.0 15.9 36.2 30.4 Q-33 Total 0.6 0.9 5.5 38.2 54.8 Non-AWS 0.5 0.5 5.7 39.1 54.2 AWS 0.7 1.4 5.1 37.0 55.8 Q-34 Total 6.1 12.1 18.5 47.0 16.4 Non-AWS 5.2 15.1 16.7 45.8 17.2 AWS 7.2 8.0 21.0 48.6 15.2 204

PAGE 215

"" 0 U1 Deer SliMly PatiJclpent I am lnteresled In fJinployee lhoughts tBfliJidlng the use ol altematlve worlr schedules. Under the provisions ol Public Law 99-196, December 23, 1985, pennanent authority was granted to Fedetal agencies to apptOtl8 nexJble and compressed work schedules. Allemlllltle WOt1r schedules.,. penn/lied by the Director ol the Natlonlll Park Setvice prcNided that ... apptfN8d schedules do nollnlettere wllh lhe mission d lhe ..... Alternative worlr schedules have now been used for some lime, and it would be amamely uselul to know what employees and supetvisors think about cettaln WDrlr related Issues as well as what your opinion may be with respect to use o1 altematlve warfr achedule8 (tladble 01' compre888d wort schedules). The Information will be traated confldentJally, and will only be used to prepare awregate statistical repotts. While .,oor input and cooperation Is appt8Cialed, patflclpatlon Is voluntary but your response Is Important to the quality olltJis Sftldy. Please use the encloeed pre adell 1 ssect end lllllmped fltrtlfllope fo retum your responses. Thank You. Hlltold P. Danz What Is your age? (Citcle one) 1. under 21 2. 21-25 3. 28-30 4. 31-40 6. 41-60 8. 61 80 7. over 80 What Is the highest education level you have achieved? (Circle one) 1. some high school 5. college graduate (Bachelor's degree) 2. high-school graduate or equivalent 6. some graduate school 3 some college, but no degree 1. graduate degree (Master's or higher) 4. associate or technical degree

PAGE 216

N 0 0\ What is your job classification? (Citde one) 1. nonsupervisory clerical 6. supervisory operations 2. supervisory administrative 5. nonsupervlsory professional 3. nonsupervisory technical 6. supervisory professional 4. supervisory technical 7. management 5. nonsupervisory operations How long have you worked for the National Park Service? (Ckcle one) 1. less than 1 year 2. 1 3 years 3. 4 8 years Are you ? (Circle one) 1. Male 2. Female 4. 9 15 years 5. 16 25 years tJ. OWif 25 ....,a What type of work schedule do you have at present? (Circle one numbeT) 1. 5-day. fixed 4(}hour week 2. 4-day, fixed 40-hour WHir 3. flexitime variable 40-hour WfHIIc) 4. 5-4/9 plan 5. permanent part-time 6. other (plsrMindlcaiB whallcind) ----------------How long have you worked under your current schedule? (Ckcle one numbeT) 1. less than 30 days 2. between one to three months 3. over three months but less U.n six months 4. over six months but las than one )'88f 5. over one year

PAGE 217

N 0 ...a In terms of your overall personnel situation, which of the following work schedules would you ptelet to WOt1c under? (Citcle ane nutrlbw} 1. 5-0ay, fixed 40-hour week 2. 4-day, fixed UUour WNir 3. flexitime (5-dey, vart.ble Utour W8M) 4. 5-4/9 plan 5. pertfNinent part-time 6. other (p#81181ndfc ... --kind}---------------Are compressed work schedules (such as 8 ten-hour work days per payperiod) allowed In your organization? (Circle,.,_ ,.e no} 1. Yes. 2. No. Are limited compressed work schedules (such as 8 nlnhour and one eight-hour work day) allowed In your organization? (Circle fiJIIhar ,.e ex no) 1. Yes. 2. No. Are flexible work schedules (8 hours per day but worked at other than normal office work hours) allowed in your organization. (Circle..,_ ,.e no) 1. Yes. 2. No.

PAGE 218

The questions presented In this questionnaire ask your opinions and l8fllngs concerning work scheduln, productivity, performance end other lssun that ere applicable to the work place end employH utllfectlon. For eKh .-......m, circle number to Indicate the degr .. to whk:tl you ar or dlugrM wt1t1 the .-.emem. ....w ,.._ ......, .... .,..,.. .... ..... nd...,, Job. 1 2 3 4 5 2. Th8 wart I do an fill Job Ia m 11 VtgU 10 me. 1 2 3 4 5 3. My job,. c:IMI ...... 1 2 3 4 5 4. I wort lwd an Jab. 1 2 3 4 5 5. MCMII o1 lite errpa, 111 In Ofgatiilf&Mion aJao watt lrMl an lhak 1 2 3 4 5 Job. 6. TheAJ Is need II) ,. watt .-..In .. cwv-*-"'n. 1 2 3 4 5 N 0 7. Fleldplace {wolldng In anollter audl ltomeJ could 1 2 3 4 5 CD 1 2 3 4 5 e. An_,.,.. ...n pmdtt:Mv#lt 1n 1 2 3 4 5 10. An...,..,_. wort .:lt8dule WOIMt/'*-enltMce .. 1 2 3 4 5 ql.,.ol, ... 1 2 3 4 5

PAGE 219

13. I am mcxe produclitle CJitt& a tradillonal worlr achedule. 14. Taking evetylhlng Into consldenllion, I would be SIIIJslied willr my lite in general while Mriir'll nclfbtal Mri ..... 15. I am disslllislied wilh my cwrant Mri achedule. 16. A compressed worlr schedule would/'*-have 1111 unltNonlble lnlfuence on my phyak:al ,_.. 11. A compressed worlr schedule would/doellmpnJIIIB clienl acc.s to the seMcel al my organlzlllion. 18. A compressed worlr schedule would/'*-help fhla organlzlllion N bellsr meet the needs a/Ita cMeneL 0 \0 19. A compressed worlr schedule would/doea make II euler lot me ro meet the ..mc:e needs al my clieniJJ. 20. A compressed worlr achedule would/dcel hutt lhe qualify al client .... .no.. 2 1. The use al ather than worlr .atedules mayfdDee cause pi'OlJietna In ... .......... 22. The use ol otheT than worlr .chedules may/does cauae problemS In ooordinllling wort ... my CIOMJrlranl. sttongly dis8f11N 1 1 1 1 1 1 1 1 1 1 1 diNgtH agrMnor disaarM 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 strongly ... .. 5 .. .. 5 .. 5 .. 5 .. 5 .. 5 .. 5 .. 5 .. 5 .. 5

PAGE 220

' ... ... ... ... ... ... ... ... ... ... ... tl I u 210

PAGE 221

,, ., ., ., ., ., ., ., .. ., ., ., ., ., H J n 211

PAGE 222

APPENDIX C The Significance of the Tests of Responses Applicable to Attitude Toward Work Schedules (Likert Scale -1 strongly disagree to 5 strongly agree) Question Mean Q-8. Alternative work schedules tend to hinder productivity in this organization. 2.358 2.536 2.109 Non-AWS (n=192) AWS (n=138) Q-9. An alternative work schedule would/does improve productivity in this organization. Non-AWS (n=192) AWS (n=138) Q-10. An alternative 3.673 3.510 3.899 work schedule would/does significantly enhance the quality of my life. 3.791 Non-AWS (n=192) 3.578 AWS (n=138) 4.087 Q-11. I would bejam very productive under a compressed work schedule. Non-AWS (n=192) AWS (n=138) Q-12. I would bejam very productive under a flexible work schedule. Non-AWS (n=192) AWS (n=138) 3.752 3.625 3.928 3.979 3.844 4.167 212 F_ _T p 4.68 -4.52 0.000 3.37 4.94 0.006 5.80 5.89 0.000 7.81 3.40 0.000 3.46 4.87 0.005

PAGE 223

Q-13. I am more productive under a traditional work schedule. 2.736 6.25 -5.77 0.000 Non-AWS (n=192) 2.943 AWS (n=138) 2.449 Q-14. Taking everything into consideration, I would be satisfied with my life in general while working a traditional work week. 3.294 6.66 -4.81 0.000 Non-AWS (n=192) 3.490 AWS (n=138) 3.022 Q-15. I am dissatisfied with my current work schedule. 2.282 6.09 -6.10 0.000 Non-AWS (n=192) 2.479 AWS (n=138) 2.007 Q-16. A compressed work schedule would/does have an unfavorable influence on my physical health. 2.279 5.16 -4.32 0.000 Non-AWS (n=192) 2.427 AWS (n=138) 2.072 Q-17. A compressed work schedule would/does improve client access to the services of my organization. 2.945 5.90 4.00 0.000 Non-AWS (n=192) 2.802 AWS 3.145 Q-18. A compressed work schedule would/does help this organization better meet the needs of its clients. 3.112 5.51 4.02 0.000 Non-AWS (n=192) 2.969 AWS (n=138) 3.312 213

PAGE 224

Q-19. A compressed work schedule would/does make it easier for me to meet the service needs of my clients. 3.142 Non-AWS (n=192) 3.010 AWS (n=138) 3.326 Q-20. A compressed work schedule would/does hurt the quality of client services. 2.667 2.823 2.449 Non-AWS (n=192) AWS (n=138) Q-21. The use of other than traditional work schedules mayjdoes cause problems in coordinating work with my supervisor. 2.633 Non-AWS (n=192) 2.891 AWS (n=138) 2.275 Q-22. The use of other than traditional work schedules mayjdoes cause problems in coordinating work with my coworkers. 2.794 Non-AWS (n=192) 3.031 AWS (n=138) 2.464 Q-23. A traditional work schedule would/does have an unfavorable influence on my family life. 2.642 Non-AWS (n=192) 2.495 AWS (n=138) 2.848 Q-24. An alternative work schedule would/does make it easier for me to coordinate my schedule with the schedules of other family members. 3.506 Non-AWS (n=192) 3.385 AWS (n=138) 3.674 214 5.93 3.64 0.000 10.52 -4.14 0.000 6.88 -6.92 0.000 7.18 -5.98 0.000 4.87 4.02 0.000 1.81 3.65 0.110

PAGE 225

Q-25. Overall, alternative work scheduling has more good than bad aspects. 3.924 5.02 4.64 0.000 Non-AWS (n=192) 3.771 AWS (n=138) 4.138 Q-31. I prefer working a flexible work schedule over the regular work schedule. 3.776 8.32 9.09 0.000 Non-AWS (n=192) 3.479 AWS (n=138) 4.188 Q-32. I prefer working a compressed work schedule over the regular work schedule. 3.464 10.82 5.09 0.000 Non-AWS (n=192) AWS (n=138) 215

PAGE 226

Appendix D Comparision of Equality of Means Group ... Ca1culated Two-Tailed llull. CanDarisao T p Accel!tlB:eJect a-Resources Au Ezpert.mlta1 (a-37) 45.486 Control (n=37) 39.189 3.94 0.000 Reject Cc.pressed (n=29) 45.966 Control (n=29) 40.241 2.99 0.006 Reject F 11111 tiJDB (n=8 ) 43.750 Control (n=8) 35.375 3.10 0.017 Reject Su: Ezpert.mlta1 (n=37) 1.811 Control (D=37) 1.811 0.00 1.000 Accept Cc.pressed (n=29) 1.759 Control (n=29) 1.862 -1.28 0.212 Accept FlerltiJDa (D=8) Control (n=8) Race Ezpert.mlta1 (D-37) 1.432 Control (D"'37) 1.541 -0.52 0.607 Accept Compressed (11"'29) 1.310 Control (n=29) 1.517 -1.56 0.129 Accept FlerltiJDa (n-8) 1.875 Control (n=8) 1.625 0.71 0.499 Accept Marita1 Status Ezperimanta1 (n=37) 1.919 Control (n=37) 1.865 0.35 0.724 Accept Cc.pressed (D""29) 1.931 Control (n=29) 1.828 0.74 0.467 Accept FlerltiJDa (n=8) 1.875 Control (n=8) 2.000 -1.00 0.351 Accept Educaticm Ezperimaata1 (D=37) 4.541 Control (n=37) 4.162 1.32 0.195 Accept Compressed (n=29) 4.828 Control (n=29) 3.966 2.74 0.011 Reject FlerltiJDa (n=8) 3.500 Control (n=8) 4.875 -2.43 0.046 Reject Years f!!E!!!:!:f:ed in Unit Ezpert.mlta1 (n=37) 6.811 Control (n=37) 7.784 -1.21 0.236 Accept Caapressed (n=29) 7.724 Control (D=29) 8.621 -0.94 0.355 Accept FlerltiJDa (IF'8) 3.500 Control n=8) 4.750 -2.20 0.063 Accept Mountain liP Kaintao-ce Y! Ezperimaata1 (n=5) 40.800 Control (n=5) 52.800 -4.56 0.010 Reject 216

PAGE 227

Group Me liD Calculated 'ho-T ailed llull T p Acceetl!!,eject Mountain IP Continued Su: EzperiJDIIIltal (n=5) 2.00 Control (n=5) 2.00 0.00 1.000 Accept Race EzperiJDental (n=5) 1.00 Control (D'"5) 1.00 0.00 1.000 Accept Marital Status Ezpert.ental (D'"5) 1.80 Control (n=5) 2.00 -1.00 0.374 Accept Education Ezpert.ental (n=5) 2.800 Control (n=5) 2.400 1.07 0.345 Accept Years in Unit EzperiJDental (n=5) 10.200 Control (n=5) 18.200 2.75 0.051 Accept lhalm Resources Reaiona Productivitx Pretest: Ezpert.ental (n=37) 266.108 Control (n=37) 263.919 0.11 0.914 Accept Caalpressed (n=29) 277.724 Control (n=29) 267.690 0.42 0.680 Accept Flexi time (n=8) 224.000 Control (n=8) 250.250 -0.85 0.426 Accept Productivit% Posttest: EzperiJDIIIltal (n=37) 292.946 Control (n=37) 335.135 -1.86 0.071 Accept Caalpressed (n=29) 308.655 Control (n=29) 360.931 -1.98 0.058 Accept Flexitillle (n=8) 236.000 Control (n-8) 241.625 -0.14 0.889 Accept Maintenance HHBP Productivit% Pretest: Ezpert.ental (n=5) 376.000 Control (n=5) 489.800 -2.68 0.055 Accept Productivit% Posttest: EzperiJDIIIltal (n=5) 323.400 Control (n=5) 444.000 -2.71 0.054 Accept BmWf RESOURCES REGIORS PerfoJ:IIIIIIlce Evaluaticm Rat!ga -Pretest: Ezpert.ental (D'"37) 292.027 Control (n=37) 299.108 -1.04 0.307 Accept ec.pressed (n=29) 298.345 Control (n=29) 306.621 -1.16 0.258 Accept Flexitillle (n=8) 269.125 Control (n=8) 271.875 -0.17 0.872 Accept 217

PAGE 228

Group Mean Calculated 1Wo-Tailad Bull Cmmarhcm r p Acce:etaleject B'IMAII RESOORCES REGiatS, Ccmtinued Performance Evaluaticm RatiEZ -Posttest: bpert.antal (JF37) 300.676 Ccmtrol (JF37) 300.946 -0.04 0.966 Accept Caapressed (IF29) 305.069 Ccmtrol (JF29) 298.104 0.93 0.362 Accept (IF8) 284.750 Ccmtrol (IF&) 311.250 -3.08 0.018 Reject Almual Leave Usye S!!!!l!1e Mean l2 Yrs -Pretest: bpert.antal (D""37) 136.162 Ccmtrol (JF37) 146.405 -1.19 0.240 Accept Compressed (JF29) 148.035 Ccmtro1 (JF29) 158.138 -1.14 0.263 Accept (IF&) 93.125 Ccmtrol (IF8) 103.875 -0.64 0.541 Accept Annual Leave Usye 1991 Posttest: bperiJIIental (JF37) 145.486 Ccmtrol (JF37) 160.432 -1.87 0.070 Accept Compressed (n=29) 152.241 Ccmtrol (JF29) 170.138 -1.82 0.079 Accept (IF&) 121.000 Ccmtrol (IF&) 128.125 -1.64 0.145 Accept Sick Leave Us!Y5e S!!!!l!le Mean l2-Yrs -Pretest: bpert.antal (JF37) 60.000 Ccmtrol (JF37) 45.541 2.19 0.035 Reject Compressed (n=29) 63.310 Ccmtrol (n=29) 51.310 1.52 0.141 Accept (IF&) 48.000 Ccmtrol (IF&) 24.625 2.27 0.057 Accept Sick Leave Us!le 1991 Posttest: bpert.antal. (11"'37) 57.919 Ccmtrol (n=37) 53.324 0.76 0.451 Accept Compressed (n=29) 63.207 Ccmtrol (n=29) 60.793 0.33 0.741 Accept (IF&) 38.750 Ccmtrol (IF&) 26.250 1.96 0.091 Accept MAIBTElWtCE lHIP PerfoJ:11181lce Evaluation Rat!gs -Pretest: bpert.antal (IF5) 262.400 Cantrol (n=5) 290.200 -1.75 0.155 Accept PerfoJ:11181lce Evaluaticm RatiEZ -Posttest: bpert.antal (u-5) 271.000 Ccmtrol (u-5) 293.000 -3.55 0.024 Reject 218

PAGE 229

Group HaiiD Calculated Two-Tailed llul.l CuDarisDD y p Accee,ae iect MAiliTEKAIICE m.p 1 CODtiDued Aanual Leave Usye S!!!l!le Ha1111 2 Yrs l -Pretest: Ezperillllmtal (n=5) 148.600 Control (n=5) 194.600 -2.04 0.111 Accept Aanual Leave Usye 1991 Posttest: Ezperillllmtal (n=5) 149.800 Control (n=5) 193.800 -4.36 0.012 Reject Sick Leave Usye s-le Ha1111 -Pretest: Ezperillllmtal (u=5) 50.000 Control (u=5) 27.400 2.37 0.076 Accept Sick Leave Usye 1991 Posttest: Ezperillllmtal (n=5) 80.600 Control (D=5) 35.800 1.90 0.131 Accept BlNB RESCJIIRCES Productivit% ADal%sis: Control Posttest (IF37) 335.135 Control Pretest (n=37) 263.919 3.26 0.002 Reject Control Posttest (u=29) 360.931 Control Pretest (n=29) 267.690 4.04 0.000 Reject Control Posttest (n=8) 241.625 Control Pretest (n=8) 250.250 -0.19 0.852 Accept Ezperillllmtal Posttest (u=37) 292.946 Ezperillllmtal Pretest (u=37) 266.108 1.18 0.244 Accept Cmapressed Post test (u=29) 308.655 Caapressed Pretest (n=29) 277.724 1.17 0.251 Accept Flazihle Posttest (n=8) 236.000 Flexible Pretest (u-8) 224.000 0.31 0.767 Accept EvaluetiDDB: Control Posttest (u=37) 300.946 Control Pretest (D=37) 299.108 0.28 0.785 Accept Control Posttest (n=29) 298.103 Control Pretest (JF29) 306.621 -1.21 0.235 Accept Control Posttest (n=8) 311.250 Control Pretest (a-8) 271.775 2.17 0.067 Accept Ezperillllmtal Posttest (u=37) 300.676 Ezperillllmtal Pretest (n=37) 292.027 1.38 0.177 Accept Cmapressed Posttest (n=29) 305.069 Caapressed Pretest (u=29) 298.345 0.90 0.378 Accept Flexible Posttest (n=8) 284.750 Flazihle Pretest (n=8) 269.125 1.82 0.112 Accept Aanual Leave Usye: Control Posttest (u=37) 160.432 Control Pretest (u=37) 146.405 1.70 0.098 Accept Control Posttest (u=29) 170.138 Control Pretest (o=29 158.138 1.34 0.190 Accept Control Posttest (n=8) 128.125 Control Pretest (n=8) 103.875 1.62 0.150 Accept 219

PAGE 230

Group He an Calculated Two-Tailed Bull CclmDarillon T p Acc!m!;l!!ject llmtAJII RESOURCES BEGIOBS I Contiuuad ADDual Leave O&!le: bper:Umntal Posttest (D=37) 145.486 bper:Umntal Pretest (u=37) 136.162 1.16 0.252 Accept Caalpresaed Postteat (u=29) 152.241 Caalpreaaed Pretest (u=29) 148.035 0.43 0.611 Accept Flazihle Posttest (JF8) 121.000 Flll][ible Pretest (u=8) 93.125 6.41 0.000 Reject Sick Leave Us!le: Control Posttest (u=37) 53.324 Control Pretest (u=37) 45.541 0.93 0.358 Accept Control Posttest (u--29) 60.793 Control Pretest (u=29) 51.310 0.96 0.344 Accept Control Posttest (JF8) 26.250 Control Pretest (JF8) 24.625 0.15 0.885 Accept Posttest (u=37) 57.919 Pretest (u=37) 60.000 -0.34 0.732 Accept Caalpressed Posttest (u=29) 63.207 Caalpreased Pretest (11""29) 63.310 0.01 0.989 Accept Flazihle Posttest (JF8) 38.750 Flexible Pretest (JF8) 48.000 -1.45 0.190 Accept Productivit% Aoalxsis: Control Postteat (u=5) 444.000 Control Pretest (uz5) 489.800 -0.67 0.542 Accept Posttest (u=5) 323.400 Pretest (u=5) 376.000 -1.18 0.303 Accept Evaluations: Control Posttest (u=5) 293.000 Control Pretest (u=5) 290.200 0.21 0.845 Accept Posttest (u=5) 271.000 Pretest (u=5) 262.400 1.39 0.238 Accept ADDual Leave Ua!le: Control Postteat (u=5) 193.800 Control Pretest (u=5) 194.600 -0.13 0.905 Accept Ezpert.antal Posttest (u=5) 149.800 Ezperimental Pretest (u=5) 148.600 0.12 0.911 Accept Sick Le11V8 Os!le: Control Posttest (u=5) 35.800 Control Pretest (u=5) 27.400 0.53 0.622 Accept Posttest (o-5) 80.600 Pretest (u=5) 50.000 1.30 0.265 Accept 220

PAGE 231

BIBLIOGRAPHY Albright, Horace M. 1985. The Birth of the National Park Service. Salt Lake City: Howe Brothers. American Management Association. 1976. Flexible Working Hours in Europe. Brussels: Management Centre Europe. American Management Association (Goodmeasure, Inc.). 1985. The Changing American Workplace: Work Alternatives in the '80s. New York: AMA Publications Division. American Management Society Foundation. 1988. 1988 AMS Flexible Work Survey. Trevose, PA: AMS Foundation. American Management Society Foundation. 1989. 1989 AMS Flexible Work Survey. Trevose, PA: AMS Foundation. Arni, Marjorie A. 1983. "Solving the staffing problem in hospitals." Supervisory Management, 28, n. 1 (January):31-33. Atkinson, William. 1985. Working at Home; Is It for You? Homewood, IL: Dow Jones-Irwin. Atwood, Caleb S. 1979. "A work schedule to increase productivity." Personnel Administrator, 24 n. 10 (October):29-33. Barad, Cary B. 1980. "Flexitime under scrutiny: research on work adjustment and organizational performance." Personnel Administrator, 25 n. 5 (May):69-74. Baum, J.J., and W.M. Young. 1973 A Practical Guide to Flexible Working Hours. London: Kogan Page. Bernard, Keith E. 1979. "Flexitime's potential for management." Personnel Administrator, 24 n. 10 (October): 51-56. Biesanz, John, and Mavis. 1964. Modern Society, an Introduction to Social Science. NJ: Prentice-Hall, Inc. Bolton, Harvey J. 1971. Flexible Work Hours. Webley, England: Anbar. 221

PAGE 232

Buchele, Robert B. 1977. The Management of Business and Public Organizations. New York: McGraw Hill, Inc. Bunger, Gerald E. 1979/80. "The Effects of Flexitime." Government Accountants Journal, 28, n. 4 (Winter): 52-57. Burdetsky, Ben, and Marvin S. Katzman. 1981. "Alternative Work Pattern Applications." Journal of Systems Management, 32, n. 12 (December):6-9. Burgess, Phillip. 1980. "Telework adding flexibility to jobs." Rocky Mountain News, 7 June: 57. Burns, Tom, and G.M. Stalker. 1961. The Management of Innovation. London: Tavistock Publications. Cherrington, David J. 1980. The Work Ethic. New York: American Management Associations. Christensen, Kathleen E. 1989. "Flexible Staffing and Scheduling in U.S. Corporations." The Conference Board. New York: Research Bulletin 240 (excerpts from this study also published in Working Woman, May, 1990, 32-35). Cole, Robert E. 1979. Work, Mobility and Participation. Berkeley: University of California Press. Coltrin, Sally A., and Barbara D. Barendse. 1981. "Is Your Organization a Good Candidate for Flexitime?" Personnel Journal, 60, n. 9 (September):712-715. Craddock, Susanne, Tom Lewis, and Jack Rose. 1981. "F1exitime: The Kentucky Experiments." Public Personnel Management Journal, 10, n. 2 :245-251. Cross, Thomas B., and Marjorie Raizman. 1986. Telecommuting: The Future Technology of Work. Homewood, IL: Dow Jones-Irwin. Donahue, Robert J. 1975. "Flex Time Systems in New York." Public Personnel Management, 4 (July/August):212 -215. Downing-Faircloth, Margo. 1982. "Would Working at Home be Wise?" Personal Computing (May):42. 222

PAGE 233

Drucker, Peter F. 1974. Management: Tasks, Responsibilities, Practices. New York: Harper and Row. Dunham, Randall B., Jon L. Pierce, and Maria B. Castaneda. 1987. "Alternative Work Schedules: Two Field Quasi-experiments." Personnel Psychology, 40, n. 2 (Summer):215-242. Edwards, Paul, and Sarah Edwards. 1985. Working From Home. Los Angeles: Jeremy P. Tarcher. Elbing, A.O., H. Gadon, and J.R.M. Gordon. 1974. "Flexible Working Hours: Its About Time." Harvard Business Review, 52: 18-33; 154-155. "Escape from the office." Business. 1989. Newsweek, 24 April: 58-60. "Europe likes flexi-time work." Social Issues. 1972. Business Week, 7 October: 80-82. Evans, E.G., and B.E. Partridge. 1973. "Notes on the Impact of Flexitime in a Large Insurance Company." Occupational Psychology, 47 (June):237-240. Everhart, William C. 1983. The National Park Service. Boulder, CO: Westview Press. Finkle, Arthur L. 1979. "Flexitime in Government." Public Personnel Management, 8, n. 3 (MayjJune):153 -155. "Flexitime -A New Work Style Catches On." 1977. Dun's Review, 109 (March):61-64. Gadon, Herman. 1984. "Making Sense of Quality of Work Life Programs." Business Horizons, 27, n. 1 (JanuaryjFebruary):42-46. Gibson, James L., John M. Ivancevich, and James H. Donnelly, Jr. 1979. Organizations: Behavior, Structure, and Process. Dallas: Business Publications. Gladden, E.N. 1972. A History of Public Administration. Vol. Two. London: Frank Cass. Golembiewski, Robert T. 1982. "Do Flexible Workhour Effects Decay Over Time?" Public Productivity Review, 6, n. 1-2 (MarchjJune):35-45 223

PAGE 234

Golembiewski, Robert T., Ronald G. Fox, and Carl w. Proehl Jr. 1980. "Flexitime: The Supervisors' Verdict." Wharton Magazine, 4, n. 4 (Summer): 43-47. Golembiewski, Robert T., Richard J. Hilles, and Monro Kagno. 1974. "A Longitudinal Study of Flexi-Time Effects: Some Consequences of an OD Structural Intervention." Journal of Aplied Behavorial Science, 10: 503-532 Golembiewski, Robert T., and Richard J. Hilles. 1977. "Drug company workers like new schedules." Special Flexitime Reports. Monthly Labor Review, 100, n. 2 (February):65-69. Golembiewski, Robert T., and Carl W. Proehl Jr. 1978. "A Survey of the Empirical Literature on Flexible Workhours: Character and Consequences of a Major Innovation." Academy of Management Review, 3, n. 4 (October):837-853. Golembiewski, Robert T., and Carl W. Proehl Jr. 1980. "Public Sector Applications of Flexible Workhours: A Review of Available Experience." Public Administration Review, 40, n. 1 (JanjFeb):72-85. Golembiewski, Robert T., Samuel Yeager, and Richard J. Hilles. 1975. "Factor Analysis of Some Flexitime Effects: Attitudinal and Behavioral Consequences of a Structural Intervention." Academy of Management Journal, 18, n. 3 (September):500-509. Gordon, Gil E. 1989. "A Closer Look at Los Angeles County Pilot," Teleconunuting Review (The Gordon Report), 6, n. 11 (November): Gordon, Gil E., and Marcia M. Kelly. 1986. Teleconunuting: How to Make It Work for You and Your Company. Englewood Cliffs, NJ: Prentice-Hall. Gorlin, Harriet. 1982. "Company Experience with Flexible Work Schedules." The Conference Board, Research Bulletin number 110. Guzzo, Richard A., and JeffreyS. Bondy. 1983. A Guide to Worker Productivitv Experiments in the United States 1976-81. New York: Pergamon Press. 224

PAGE 235

Harrick, Edward J., Gene R. Vanek, and Joseph F. Michlitsch. 1986. "Alternate Work Schedules, Productivity, Leave Usage, and Employee Attitudes: A Field Study." Public Personnel Management, 15, n. 2 (Summer):159-169. Hartzog, George B., Jr. 1988. Battling for the National Parks. Mt. Kisco, NY: Moyer Bell Limited. Hatry, Harry P. 1982. "The Status of Productivity Measurement in the Public Sector," ed. FrederickS. Lane. Current Issues in Public Administration. New York: St. Martin's Press. Hausser, D. 1980. The Impact of Flexitime on Organizational Functioning: A Field Study. Washington D.C.: U.S. Office of Personnel Management, Workforce and Development Group, Productivity Research Division 1980. Hedges, Janice Neipert. 1973. "New patterns for working time." Monthly Labor Review 94, (February):3-8 ----:,---,::. 1975. "How Many Days Make A Workweek? Monthly Labor Review, 94 {April):29-36 ------.-1977. "Flexible schedules: problems and issues." Special Flexitime Reports. Monthly Labor Review, 100, n. 2 (February):62-65. Herskovits, Melville J. 1962. Man and his Works. New York: Alfred A. Knoff. Herzberg, Frederick, Bernard Mausner, and Barbara Snyderman. 1959. The Motivation to Work. New York: John Wiley and Sons. Hollmann, Robert W., and Maureen F. Ullrich. 1983. "Participative and Flexible Decision Making." Journal of Small Business Management, 21, n. 1 (January):1-7. Hopp, Michael A., and C.R. Sommerstad. 1977. "Reaction at computer firm: more pluses than minuses." Special Flexitime Reports. Monthly Labor Review, 100, n. 2 (February):69-71. "How to manage employees on alternative schedules." 1984. Keeping Informed: Updates on management. Management Review, 73, n. 8 (August):6. 225

PAGE 236

Ivancevich, J.M., and H.L. Lyon. 1977. "The Shortened Workweek: A Field Experiment." Journal of Applied Psychology, 62, n. 4:34-37. Johnson, Darryl! R., Donald R. Field, and RichardS. Converse. 1985. The National Park Service Employee Spouse Survey of 1984: Statistical Abstract. Seattle: Cooperative Park Studies Units, University of Washington, Oregon State University, University of Idaho. Johnson, Darryll R., Donald R. Field, Gary E. Machlis, and Richard S. Converse. 1984. The Organization and the Employee in an Era of Change. The National Park Service Employee Survey of 1983: Statistical Abstract. Seattle: Cooperative Park Studies Units, University of Washington, Oregon State University, University of Idaho. Johnson, Darryl! R., and William Salvi. 1986. 1985 National Park Service Temporary Employee Survey: Statistical Abstract. Seattle: Cooperative Park Studies Unit, College of Forest Resources, University of Washington, September. Johnston, William B., Stephanie Faul, Brad Huang, and Arnold H. Packer. 1988. Workforce 2000. Washington, D.C.: Prepared for the u.s. Office of Personnel Management by the Hudson Institute (June). Kilborn, Peter T. 1990. "Americans working harder, longer as jobs become an obsession." The Denver Post, 11 June: C-1; C-3. Kim, Jay S., and Anthony F. Campagna. 1981. "Effects of Flexitime on Employee Attendance and Performance: A Field Experiment." Academy of Management Journal, 24, no. 4 (December):729-741. Kleiman, Carol. 1991. "Flex work time a benefit to firms and employees." The Denver Post, 8 April: C-4. Lazer, Robert I., and WalterS. Wikstrom. 1977. Appraising Managerial Performance. New York: The Conference Board. 226

PAGE 237

Latack, Janina C., and Lawrence w. Foster. 1985. "Implementation of Compressed Work Schedules: Participation and Job Redesign as Critical Factors for Employee Acceptance." Personnel Psychology, Inc., 38:75-92. Lawler, Edward E., III. 1986. High Involvement Management. San Francisco: Jossey-Bass. Lewis, Geoff. 1988. "The Portable Executive." Business Week, 10 October, 102-107. Macut, John J. 1974. "Measuring Productivity Under a 4 -Day Week." Monthly Labor Review, 97, n. 4 (April): 55-56. Magoon, Warren, and Larry Schnicker. 1976. "Flexible hours at State Street Bank of Boston: a case study." The Personnel Administrator, 21, n. 7, (October): 34-37. Management Centre Europe. 1976. Flexible working hours in Europe. Brussels. Martin, Arlene. 1989. "There's No Place Like Home to Work." Human Resource Executive (July):50-51. Maslow, Abraham H. 1954. Motivation and Personality. New York: Harper and Row. Mayo, Elton. 1933. The Human Problems of an Industrial Civilization. New York: Viking Press. McCarthy, Maureen. 1981. "Should Your Firm Consider The Flexitime Alternative?" Industrial Engineering, January, 49-50; 55-57; 89. McGee, Lynne F. 1988. "Setting up Work at Home." Personnel Administrator (December):58-62. McKann, M.R. 1973. "Flexitime at Florida." Library Journal, November, McKendrick, Joseph E., Jr. 1989. "Stretching Time in '89." Management World (JulyjAugust):10-11. Morgan, Phillip I., and Kent H. Baker. 1984. "Taking a look at flexitime." Supervisory Management, 29, n. 2 (February):37-43 227

PAGE 238

Mowday, RichardT., Richard M. Steers, and Lyman w. Porter. 1979. "The Measurement of Organizational Commitment." Journal of Vocational Behavior, 14:234-247. Moskowitz, Milton, and Carol Townsend. 1989. "The 60 Best Companies for Working Mothers." Working Mother (October):74-100. Mueller, Oscar, and Muriel Cole. 1977. "Concept wins converts at Federal Agency." Special Flexitime Reports. Monthly Labor Review, 100, n. 2 (February): 71-74. Narayanan, V.K., and RaghuNath. 1982a. "Hierarchical Level and the Impact of Flextime." Industrial Relations, 21, no. 2 (Spring):216-230. Narayanan, V.K., and RaghuNath. 1982b. "A Field Test of Some Attitudinal and Behavioral Consequences of Flexitime." Journal of Applied Psychology, 67, no. 2:214-218. Narayanan, V.K., and RaghuNath. 1984. "The Influence of Group Cohesiveness on Some Changes Induced by Flexitime: A Quasi-Experiment." The Journal of Applied Behavioral Science, 20, n. 3:265-276. National Center for Productivity and Quality of Working Life. 1977. Improving Government Productivity: Selected Case Studies. Washington, D.C.: u.s. Government Printing Office. Newstrom, John W., and Jon L. Pierce. 1979. "Alternative work schedules: the state of the art." The Personnel Administrator, 24 n. 10 (October):19-23. "New Work Schedules For a Changing Society." 1981. Executive Summary of a Work in America Institute Policy Study, Directed by Jerome M. Rosow, and Robert Zager, Work in America Institute, Inc., Scarsdale, NY. Nilles, Lack M, F. Roy Carlson, Paul Gray, and Gerhard Hanneman. 1976. The Telecommunications -Transportation Trade Off: Options for Tomorrow. New York: John Wiley and Sons. 228

PAGE 239

Nollen, Stanley D. 1979. "Does flexitime improve productivity?" Harvard Business Review, 57, n. 5, (Septemberj0ctober):12, 16-22. Nollen, Stanley D. 1980. "What is happening to flexitime, flexitour, gliding time, the variable day? And permanent part-time employment? And the four-day week?" Across the Board, 17, n. 4 (April):6-21. Nollen, Stanley D., and Virginia H. Martin. 1978. Alternative Work Schedules: Flextime. New York: AMACOM. O'Connell, Jim, and George Flynn. 1988. "City stuck as county endorses flex time." The San Diego Union, 28 September: B-1; B-5. Olson, Dan, and Arthur P. Brief. 1978. "The Impact of Alternative Workweeks." Personnel, 55, n. 1, (JanjFeb.):73-77. Olsten, William. 1984. "Effectively managing alternative work options." Supervisory Management, 29, n. 4 (April):10-15. Orpen, C. 1981. "Effect of flexible working hours on employee satisfaction and performance: A field experiment." Journal of Applied Psychology, 66: 113-115. Owen, John D. 1977. "Flexitime: Some Problems and Solutions." Industrial and Labor Relations Review, 30, n. 2 (January):152-160. Partridge, B.E. 1973. "Notes on the Impact of Flexitime in a Large Insurance Company: II. Reactions of Supervisors and Managers." Occupational Psychology, 47 (June):241-242. Pierce, Jon L., and John W. Newstrom. 1980. "Toward A Conceptual Clarification Of Employee Responses To Flexible Working Hours: A Work Adjustment Approach." Journal of Management, 6, n. 2: 117-134 Pierce, Jon L., and John W. Newstrom. 1983. "The design of flexible work schedules and employee responses: relationships and process." Journal of Occupational Behavior, 4:247-262. 229

PAGE 240

President's Council on Management Improvement. 1987. 1987 Annual Report to the President. Government Excellence Through Partnership. Washington D.C., PCMI. President's Council on Management Improvement. Human Resources Committee. 1990. Guidelines for Pilot Flexible Workplace Arrangements. Washington D.C., PCMI (January). Rainey, Glenn W., Jr. and Lawrence Wolf. 1981. "Flex -Time: Short-Term Benefits; Long-Term ... ?" Public Administration Review, 41, n. 1, (JanjFeb):52-63. Rainey, Glenn w., Jr. and Lawrence Wolf. 1982. "The Organizationally Dysfunctional Consequences of Flexible Work Hours: A General Overview." Public Personnel Management, 11, n. 2:165-175. Ralston, David A., William P. Anthony and David J. Gustafson. 1985. "Employees May Love Flexitime, but What Does It Do to the Organization's Productivity?" Journal of Applied Psychology, 70, n. 2:272-279. Ronen, Simcha. 1981. "Arrival and Departure Patterns of Public Sector Employees Before and After Implementation of Flexitime." Personnel Psychology, 34, n. 4:817-822. Ronen, Simcha, and Sophia B. Primps. 1980. "The Impact of Flexitime on Performance and Attitudes in 25 Public Agencies." Public Personnel Management Journal, 9, n. 3:201-206. Rosenthal, Robert. 1976. Experimenter Effects in Behavioral Research. New York: Irvington Publishers. Rosow, Jerome M., and Robert Zager. 1983. "Punch out the time clocks." Harvard Business Review, 83, n. 2 (March/April):12-14. Rubin, Richard S. 1979. "Flexitime: Its Implementation in the Public Sector." Public Administration Review, 39, n. 3 (MayjJune):277-282. Ruh, Robert A., Kenneth J. White, and Robert R. Wood. 1975. "Job Involvement, Values, Personal Background, Participation in Decision Making, and Job Attitudes," Academy of Management Journal, 18, n. 2 (June):300-312. 230

PAGE 241

Saltzman Am W 11 d Y 1991. "Trouble at the top." u.s. News or Report, 17 June:40-48. Saltzman, Amy. 1991. "Sex discrimination subtleties delicate but devastating." Rocky Mountain News, July: S0-51. and are 7 Schein, Virginia E., Elizabeth H. Maurer and Jan F. Novak. 1977. "Impact of Flexible working Hours on Productivity." Journal of Applied PsychologY., 62, n. 4:463-465. Sc'nroeer susan G. 1983. "Alternative work.styles:,:aa to productivity problems?" _uperv> 28, n. 1 (July):24-30 M a ement M 198Q D_l.Ctl.ona l' ois Moore JaY oak park, 11 1.n and Labo "e -

PAGE 242

Saltzman, Amy. 1991. "Trouble at the top." u.s. News and World Report, 17 June:40-48. Saltzman, Amy. 1991. "Sex discrimination subtleties are delicate but devastating." Rocky Mountain News, 7 July: 50-51. Schein, Virginia E., Elizabeth H. Maurer and Jan F. Novak. 1977. "Impact of Flexible Working Hours on Productivity." Journal of Applied Psychology, 62, n. 4:463-465. Schroeer, Susan G. 1983. "Alternative workstyles: A solution to productivity problems?" Supervisory Management, 28, n. 7 (July):24-30. Shafritz, Jay M. 1980. Dictionary of Personnel Management and Labor Relations. Oak Park, Illinois: Moore Publishing Co. ________ 1985. Facts on File Dictionary of Public Administration. New York: Facts on File Publications. Shafritz, Jay M., Albert C. Hyde, and David H. Rosenbloom. 1981. Personnel Management in Government. New York: Marcel Dekker, Inc. Smith, A.D., D.M.W.N. Hitchens, and S. w. Davies. 1982. "International Industrial Productivity: A Comparison of Britain, America and Germany." National Institute Economic Review, n. 101 (August):13-25. Stewart, Gary M., and Arthur Guthrie. 1976. "Alternative workweek schedules: Which one fits your operation?" Supervisory Management, 21, n. 6 (June):2-14. Sutermeister, Robert A. 1976. People and Productivity. New York: McGraw Hill. Swart, J.C. 1974. "What Time Shall I Go To Work Today?" Business Horizons, 17 (October):l9-26. Teasley, C.E., III. 1983. "Human Resource Management for Public Sector Productivity: We Need Desperately to Practice what we Preach," ed. Robert T. Golembiewski and Frank Gibson. Readings in Public Administration. Boston: Houghton Mifflin Company. 231

PAGE 243

Thomas, Edward G. 1986. Flexible Work Keeps Growing." Management World (April/May):43-45. Turney, John R., and Stanley L. Cohen. 1983. "Alternative Work Schedules Increase Employee Satisfaction." Personnel Journal (March):202-207. u.s. Comptroller General of the United States. General Accounting Office. 1980. The Alternative Work Schedules Experiment: Congressional Oversite Needed to Avoid Likely Failure. Report to the Congress of the United States. GPO, Washington D.C.: November 14. U.S. Comptroller General of the United States. General Accounting Office. 1985. Alternative Work Schedules for Federal Employees. B-216021. Report to the Chair, Task Force on Economic Security, Select Committee on Children, Youth and Families, House of Representatives. GPO, Washington, D.C.: July 19. u.s. Comptroller General of the United States. General Accounting Office. 1987. Status of Agency Practices for Improving Federal Productivity. B-226307. Fact sheet for the ranking minority member, Subcommittee on Federal Services, Post Office, and Civil Service, U.S. Senate. Superintendent of Documents, Washington, D.C.: June 16. U.S. Comptroller General of the United States. General Accounting Office. 1988. Federal Workforce -A Framework for Studying its Quality Over Time. B-228638. Report to the Chairman, Committee on Post Office and Civil Service, House of Representatives. Superintendent of Documents, Washington, D.C.: August 4. U.S. Comptroller General of the United States. General Accounting Office. 1989. Poor Performers -How They Are Identified and Dealt With in the Social Security Administration. B-209522. Report to the Ranking Minority Member, Subcommittee on Federal Service, Committee on Governmental Affairs, U.S. Senate. Superintendent of Documents, Washington, D.C.: January 27. 232

PAGE 244

u.s. Comptroller General of the United States. General Accounting Office. 1990. Performance Management -How Well Is the Government Dealing With Poor Performers? B-235639. Report to the Ranking Minority Member, Subcommittee on Federal Service, Committee on Governmental Affairs, U.S. Senate. Superintendent of Documents, Washington, D.C.: October 2. u.s. Congress. House. Committee on Post Office and Civil Service. 1978. Federal Employees Flexible and Compressed Work Schedules Act of 1978. An Act to authorize Federal agencies to experiment with flexible and compressed employee work schedules over a three year period. 95th Congress, 92 Stat. 755, Public Law 95-390, 29 September. U.S. Congress. Senate. Committee on Governmental Affairs. Civil Service Reform Act of 1978. 1978. To reform the Civil Service Laws. 95th Congress, 92 Stat. 1111, Public Law 95-454, 13 October. u.s. Congress. Senate. Committee on Governmental Affairs. 1982. Federal Employees Flexible and Compressed Work Schedules Act, Extension. An Act to temporarily extend until the first day of the second pay period beginning after July 4, 1982, the authority to conduct experiments in flexible schedules under the Federal Employees Flexible and Compressed Work Schedules Acts of 1978. 97th Congress, 96 Stat. 21, Public Law 97-160, 26 March. U.S. Congress. Senate. Committee on Governmental Affairs. 1982 Federal Employees Flexible and Compressed Work Schedules Act of 1982. An Act to amend Title 5, United States Code, to provide permanent authorization for Federal Agencies to use flexible and compressed employee work schedules. The amendments made by this Act shall not be in effect three years after the date of enactment of this Act. 97th Congress, 96 Stat. 227, Public Law 97-221, 23 July. u.s. Congress. House. Committee on Post Office and Civil Service. Subcommittee on Human Resources. 1985. Federal Employees Flexible and Compressed Work Schedules Act of 1982: Hearings, H.R. 1534, Report No. 99-82, Washington, D.C.: u.s. Government Printing Office, March 19. 233

PAGE 245

U.S. Congress. House. Committee on Post Office and Civil Service. 1985. Federal Employees Flexible and Compressed Work Schedules Act of 1982. An Act to convert the temporary authority to allow Federal employees to work on a flexible or compressed work schedule under title 5, United States Code, into permanent authority (removal of section 5 of P.L.97 -221). 99th Congress, 99 Stat. 1350, Public Law 99 -196, 23 December. U.S. Congress. Congressional Budget Office. 1986. Employee Turnover in the Federal Government. Washington, D.C.: Superintendent of Documents, February. U.S. Congress. Congressional Budget Office. 1987. Federal Civilian Employment. Washington, D.C.: Superintendent of Documents, December. u.s. Congress. Committee on Post Office and Civil Service. 1989. Federal Employees Flexible Work Arrangements Act of 1989: Hearings, H.R. 2435, Washington, D.C.: U.S. Government Printing Office, May 18. U.S. Department of Interior. National Park Service. 1986. Alternative Work Schedules in the National Park Service. A handbook for supervisors. Washington, D.C.: NPS. U.S. Merit Systems Protection Board. 1987. Federal Personnel Policies and Practices --Perspectives from the Workplace. Washington, D.C.: MSPB, December. U.S. Merit Systems Protection Board. 1988. Towards Effective Performance Management in the Federal Government. Washington, D.C.: MSPB, July. u.s. Merit Systems Protection Board. 1990. Working for America: A Federal Employee Survey. Washington, D.C.: MSPB, June. u.s. Office of Management and Budget (Bureau of the Budget). 1964. Measuring Productivity of Federal Government Organizations. Washington D.C.: Superintendent of Documents. 234

PAGE 246

U.S. Office of Personnel Management. Assessment Services Division. 1991. Federal Personnel Manual System. Federal Flexible Workplace (Flexiplace) Project. FPM Letter 368-1, Washington, D.C.: 26 March. u.s. Office of Personnel Management. Office of Compensation Program Development. 1981. Alternate Work Schedules Experimental Program: Interim Report to the President and the Congress. Washington D.C.: September. u.s. Office of Personnel Management. Office of Compensation Program Development. 1982. Alternative Work Schedules Experiment: Final Report to the President and the Congress. Washington D.C.: March. u.s. Office of Personnel Management. Office of Compensation Program Development. 1990. Flexiplace Focus. Washington D.C.: March. u.s. Office of Personnel Management. Office of Planning and Evaluation, and Office of Productivity Programs. 1981. Federal Employee Attitudes. Phase 1: Baseline Survey 1979. Government-wide Report. Washington D.C.: Superintendent of Documents. U.S. Office of Personnel Management. Office of Planning and Evaluation, and Office of Productivity Programs. 1983. Federal Employee Attitudes. Phase 2: Followup Survey 1980. Government-wide Report. Washington D.C.: u.s. Government Printing Office. U.S. Office of Personnel Management. Office of Planning and Evaluation, and Office of Productivity Programs. 1984. Federal Employee Attitude Survey. Phase 3: 1983 Preliminary Report. Washington, D.C.: OPM, February. U.S. Office of Personnel Management. Workforce Effectiveness and Development Group. 1980. Measuring Federal Productivity: A Summary Report and Analysis of the FY 78 Productivity Data. Washington, D.C.: February. u.s. President. 1986. Productivity Improvement Program for the Federal Government. Executive Order 12552. Federal Register (28 February) vol. 51, no. 40, page 7041. Microfiche. 235

PAGE 247

Vroom, Victor H. 1964. Work and Motivation. New York: Wiley. Wade, Martin. 1973. Flexible Workings Hours in Practice. Essex, England: Glower Press. Walker, C.R. 1950. "The problem of the repetitive job." Harvard Business Review (May-June):54-58. w. E. Upjohn Institute for Employment Research. Report of a Special Task Force to the Secretary of Health, Education and Welfare. 1981. Work in America. Cambridge, MA: MIT Press. Weber, Max. 1958. "Die Protestantische Ethik und der Geist des Kapitalismus," Archiv fur Sozialwissenschaft, 20, 21 (1904, 1905). Translated by Talcott Parsons, The Protestant Ethic and the Spirit of Capitalism. New York: Charles Scribner's Sons. Welch, JoeL., and David Gordon. 1980. "Assessing the Impact of Flexitime on Productivity." Business Horizons, 23, n. 6 (December):61-65. Wheat, Richard A. 1982. "The Federal Flexitime System: Comparison and Implementation." Public Personnel Management Journal, 11, n. 1:22-30. Wiener, Leonard. 1988. "Your Home, The Office." u.s. News and World Report, 26 September:64-73. Winett, Richard A., and MichaelS. Neale. 1980. "Results of experimental study on flexitime and family life." Monthly Labor Review, 103, n. 11 (November):29-32. Zawacki, Robert A., and Jason S. Johnson. 1976. "Alternative workweek schedules: One company's experience with flextime." Supervisory Management, 21, n. 6 (June):15-19. Zeldner, Rita L. 1990. "Telecommuters Praise In-House Job Experiment." Federal Times, 29 October: 3; 16. Zientara, Marguerite. 1983. "More Managers Allow Varied Work Schedules -Better Productivity, Use of Resources Among Reasons for Flexible Schedules." Computerworld, 20 June: 11. 236