The distribution of leadership in nonprofit organizations between the chair and the executive director

Material Information

The distribution of leadership in nonprofit organizations between the chair and the executive director
Leduc, Robert F
Publication Date:
Physical Description:
xii, 170 leaves : ; 28 cm

Thesis/Dissertation Information

Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public administration


Subjects / Keywords:
Nonprofit organizations -- Management -- United States ( lcsh )
Chief executive officers ( lcsh )
Directors of corporations ( lcsh )
Chief executive officers ( fast )
Directors of corporations ( fast )
Nonprofit organizations -- Management ( fast )
United States ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 151-170).
General Note:
School of Public Affairs
Statement of Responsibility:
by Robert F. Leduc.

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:
44094763 ( OCLC )
LD1190.P86 1999d .L33 ( lcc )

Full Text
Robert F. Leduc
B.S., Wayne State University, 1972
M.B.A., Wayne State University, 1974
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Administration

1999 by Robert F. Leduc
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
Robert F. Leduc
has been approved


Leduc, Robert F. (Ph.D., Public Administration)
The Distribution of Leadership in Nonprofit Organizations between the Chair and the
Executive Director
Thesis directed by Assistant Professor Michael Cortes
Americas nonprofit sector, with more than one million organizations,
employs ten percent of all Americans and provides many of the countrys charitable,
civic, health, educational and cultural services. The critical role nonprofit
play in providing needed services, and the difficulties in funding these services,
demands effective leadership.
Traditionally, nonprofit leadership and management roles have been divided
between the volunteer board of directors headed by a chairperson, and a paid staff
headed by an executive director. An on-going issue of debate concerns how the
leadership roles are operationally divided between the chair and the executive
Most of the literature on management of nonprofit organizations is
prescriptive, stating that the chair is paramount, with the executive director at the
chairs service. Yet most descriptive literature on the subject, where writers have
observed what actually happens in nonprofit organizations, describes the executive
director as paramount in both management and leadership.

This research concerns the actual division of responsibility between the chair
and the executive director in 17 medium-sized, nonprofit, human service and arts
organizations with substantial external funding. The research was exploratory in
nature and designed primarily to develop hypotheses rather than test them. It used
purposive sampling, and quantitative and qualitative information was gathered and
analyzed. The t-test determined whether there were differences in the perceived roles
of chair and executive director for successful and less-successful organizations.
Discriminant analysis provided further information about what discriminates successful
and less-successful nonprofit organizations.
Of the three hypotheses developed, two were supported. The respondents
agreed that the role of executive director, rather than the role of chair, is more
critical to the success of nonprofit organizations. The chairs and executive
directors in successful organizations demonstrated greater congruence in their
perceptions and actions regarding the distribution of leadership functions, than did
the chairs and executive directors of less-successful organizations.
Board members must acknowledge the legal burdens which the chair must
bear; yet this research suggests that nonprofit leaders need more information about
working in partnerships for the long-term good of nonprofit organizations, and
that the executive director is the day-to-day and ultimate long-term leader.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Michael Cortes

To my wife, Diane, who knows the many reasons. And to my daughters, Elizabeth
and Kristen, who grew to realize that learning is a life-long pursuit.

I acknowledge the staff of the Graduate School of Public Affairs at the University of
Colorado at Denver for their limitless patience. Special thanks also to my Chairman,
Mike Cortes, for his encouragement and valuable direction to see me through.
Respectful appreciation is also given to my committee members, Philip Burgess,
Linda deLeon, Franklin James and Marshall Kaplan for their wisdom and their
A special thank you to my selfless mentor, Dr. Jamal Abedi, professor at the
University of California at Los Angeles, who explained methodology in practical
terms and helped me apply it effectively. His ongoing advice and counsel proved
My appreciation to Cynthia Taskesen, research assistant at UCLA, for her analytical
suggestions; to Marla Knutsen, for her very capable editing of final drafts; and Diane
Leduc for her help with all parts of this thesis.
Also, I wish to thank Dawn Savage of GSPA, who ran interference for me with the
bureaucracy and especially my colleagues at UCLA from the advanced research class,
who served as the backboard on which I bounced ideas and received inspiration.

Tables............................................................ xii
1. THE PROBLEM.................................................1
Importance of the Nonprofit Sector.......................1
Nonprofits and Public Administration.....................3
Leadership and Success...................................5
Significance of the Problem..............................5
Purposes and Objectives of This Study....................6
2. LITERATURE REVIEW..........................................12
Leadership in Private Sector Corporate Organizations....16
The Leadership Role in Nonprofit Organization Management....19
Dividing Leadership Roles between the Chair and the
Executive Director......................................23
The Division of Leadership Roles in Nonprofit Organizations.26
Models Found in the Literature..........................30
3. METHODOLOGY................................................34

Population and Sample........................................36
A Taxonomy of Nonprofit Organizations........................37
Identification of Successful and Less-Successful Organizations.41
Data-Collection Instruments..................................47
Pre-Testing of the Questionnaire and the Interview Tools.....51
Mail Survey..................................................55
Structured Interviews........................................57
Analysis of the Data.........................................59
Analysis Techniques/Statistical Approaches...................59
Limitations of This Study....................................61
4. FINDINGS.........................................................65
Demographic Information on the Respondents to the Mail Survey.. 64
Responses to the Leadership Questions in the Mail Survey.....65
Discriminant Analysis........................................70
Securing Qualitative Data from the Interviews................73
Responses to Questions in Interview Section One..............75
Responses to Questions in Interview Section Two..............79
Responses to Questions in Interview Section Three............86
Responses to Questions in Interview Section Four.............89

The Hypotheses
The Division-of-Leadership Functions...............100
Congruence of Responses within the Dyads...........105
5. SUMMARY AND RECOMMENDATIONS........................108
A Review of the Significant Findings of This Study.109
Thoughts Concerning the Findings...................113
What Has Been Learned..............................114
Comparing the Findings with the Literature.........120
A. EXPERT JUDGES CONTACT LETTER......................127
EXECUTIVE DIRECTORS.............................. 129
C. MAIL SURVEY........................................131
D. INTERVIEW GUIDE FOR CHAIR..........................136
F. TABLE F.l..........................................140
G. TABLE G.i:.........................................143

LIST OF WORKS CITED...............................145
ANNOTATED BIBLIOGRAPHY............................151

2.1 Three Models for the Division of Leadership Role Functions
Between Chair and Executive Director in Nonprofit Organizations....33
3.1 National Taxonomy of Exempt Entities................................38
3.2 Consistency of Judges' Choices......................................47
3.3 Percentages of Potential Participation to Actual Participation......57
4.1 Means, Standard Deviations t-value, and p for All 17 Leadership
4.2 Discriminant Analysis Structure Matrix for the 17 Leadership
4.3 A Comparison of the Differences of Scores within Dyads for
Successful and Less-Successful Nonprofits.........................106
F. 1 A Comparison of the Differences in Scores Between the Leadership
of Successful and Less-Successful Organizations Concerning the
Appropriate Division of Leadership Functions......................140
G. 1 A Comparison of the Differences in Values Between the Leadership
of Successful and Less-Successful Organizations with Regard to the
Appropriate Division of Leadership Functions......................143

There is an area of administration, which falls
between public and business administration as these
are currently studied and taught. This area might be
identified as private and non-economic (churches, for
example, would be so classified as well as most
pressure groups) but it includes also some
administrative systems or activities which are on the
shadowy, fluctuating borderline between public and
private administration. Probably this area deserves
more attention than it now receives (Waldo, 1948).
This study examines leadership as practiced in nonprofit organizations,
specifically the division of leadership responsibility between the chief elected
officer (chair) and the chief appointed officer (executive director). This study
tests principles that define the roles of board chair and executive director and
helps determine the most desirable division of the leadership function between
the two positions.
Importance of the Nonprofit Sector
While the often quoted Alexis de Tocqueville (Tocqueville, 1945)
considered volunteerism and nonprofit initiative in his work of 1831, and while
some attention has been paid to the management and leadership of nonprofit

organizations within schools of social work since World War II, it has only been
in the last thirty years that the nonprofit sector has begun to be discussed
seriously by scholars and policy makers. Michael ONeill indicates that this
increased attention to the nonprofit sector is a result of such efforts as the Report
of the Commission on Private Philanthropy and Public Needs, the extensive
research conducted by Yale Universitys Program on Nonprofit Organizations,
and the consciousness-raising work of Independent Sector, a national coalition of
nonprofit groups. ONeill further indicates, The immense size and impact of the
sector has only recently been recognized, and serious theorizing about the
dynamics of the sector has only just begun (1989, p. xii).
Hodgkinson and Weitzman say that:
The nonprofit sector is often called the Third Sector.
The First Sector, that of business, commands 80 percent
of our economy, while the Second Sector, the government
sector, weighs in at about 14 percent of the economys bulk.
The nonprofit sector constitutes the remaining six percent of
the national economy (1989, p. 9).
Yet, according to Van Til, this six percent of the national economy accounts
for nine percent of the total national employment (1988, p. 3).
Richard L. Moyers notes that:
Even though the nonprofit sector is the smallest of the sectors,
it is growing the fastest. From 1977 to 1982, the nonprofit sector
grew faster than the other sectors of the economy; from 1982 to
1987, it grew at the same pace as the rest of the economy; and
from 1987 to 1990, the sectors growth rate again surpassed that

of the other sectors of the economy. In 1989, nonprofits
expenditures of $295 billion represented approximately six
percent of the national income (1994, p. 7).
In order to calculate the true importance of the nonprofit sector, one must
add to this cash economy the tremendous amount of voluntary action from
Americans. OConnell comments that forty-six percent of all women in
America volunteer, as well as forty-seven percent of all men (OConnell, 1985,
p. 10). Moyers indicates, in 1990, the value of the volunteered time totaled
$110.4 billion, or 2.4 percent of the GNP (Moyers, 1994, p. 7). Obviously, the
American nonprofit sector is a consequential part of the American economy.
Nonprofits and Public Administration
Salamon notes, Nonprofit organizations are important to the field
of public administration because they deliver 40 percent of the services that
governments, at all levels, now provide (Salamon, et al, 1986, p. 123). In
another discussion, Salamon states:
Government is also important to many types of nonprofit
organizations because governments provide about 40 percent
of the income of human service nonprofit organizations, for
example, compared to 30 percent derived from fees and 20
percent from donations (1987, p. 176).
The nonprofit sector is also important to the field of public administration
because significant encouragements and discouragements of the flow of income

and expenditures to and from nonprofit organizations are provided through a
complex of tax deductions and credits. By definition, nonprofit organizations are
tax- exempt organizations. They are not required to pay taxes on corporate
income, although their employees pay income and social security taxes. In the
case of certain charitable, religious, educational, scientific and research
organizations, American society allows tax deductions to donors/taxpayers for
their contributions to some of these organizations. This preferential treatment
speaks to the importance attributed to nonprofit organizations and to the critical
relationship they have to the exercise of public management. In The Law of Tax-
Exempt Organizations, Bruce Hopkins notes that:
While there is ongoing criticism in some quarters about
this preferential treatment for nonprofit organizations,
tax exemption is generally deemed worthy of continuance,
, despite the revenue Toss and occasional abuses, because
of the basic acceptance of the importance of individual,
pluralistic endeavors in our society (1979, p. 29).
Finally, it would be inappropriate to speak of the nonprofit sector and
public administration without mentioning the critical relationship between
nonprofits and democracy. When the previously mentioned French aristocrat,
Alexis de Tocqueville, visited the United States in the early nineteenth century,
he was impressed with the effect of voluntary organizations in shaping American
society. To de Tocqueville, democracy not only engendered and nourished, but

also required for its survival, a healthy and diverse nonprofit sector (cited in
Moyers, 1994, p. 5).
Leadership and Success
A review of the literature, as provided in Chapter 2, on determinants of
success such as management skills, motivation, planning, attitude, personality,
time management and others is abundant but little related to the organizational
arena. Similarly, there is also a small body of literature on leadership as a
determinant of success but it relates to success in a broad range of topics such as
physical education, undergraduate success, predicting college success, Chicanos
and Angelinos in community settings, but very little considered organizational
success through leadership.
Significance of the Problem
This study attempts to shed more light on a continuing debate that is
of great importance to managers and board members of the American nonprofit
sector. Considerable tension and contradiction exist within the literature of
nonprofit management about the division of leadership between the chair and the
executive director. The prescriptive literature says the chair is supreme, while the
empirical literature suggests the opposite. This tension needs to be better
understood if not resolved in order to help practitioners know what they should
do to best accomplish their jobs and their organizational missions.

These tensions between prescribed and actual practice are also reflected in
the discrepancy between what is taught in the academy and in training sessions
versus what actual board chairpersons and actual executive directors need to do
in the effective leadership of a nonprofit organization. There is a need for further
research, as Chapter 5 will argue.
Purposes and Objectives of This Study
This research has five objectives:
1. To begin to test the models of leadership division in nonprofit organizations in
the literature by comparing them with the actual practice in successful, and
less-successful, nonprofit organizations.
2. Consider, on the basis of new data gathered through this study, how
leadership in nonprofit organizations, as practiced by the chair and the
executive director, is divided between them.
3. Consider, on the basis of new data gathered by this study, whether the chair or
the executive director is more important to the leadership of nonprofit
organizations. More important will be used in this study, instead of more
effective or more critical or other similar terms, to mean that one of the
two necessary leaders, the chair as prescribed by law or the executive director
as prescribed by management practice, must take the lead even though the
other is also needed to make any nonprofit organization successful.

4. Consider, on the basis of new data gathered by this study, if and how the
chair and the executive director are supportive of each others leadership.
5. Consider, on the basis of new data gathered by this study, if there is a
difference in the division of leadership and responsibility over the short
(daily administration) and long (policy-level management) term within
nonprofit organizations.
Before proceeding with the discussion of leadership in nonprofit
organizations, it is important to clarify terminology to be used. Throughout this
study, for purposes of simplification, I will use the term chair synonymously
with the term chief elected officer, rather than such other choices as
president, supreme guardian, chancellor (deck, 1980, p. 661) or chief
volunteer officer (OConnell, 1988, p. 3). Similarly, I will use the term
executive director in place of the term chief appointed officer, rather than
such other choices as chief staff officer, president, executive vice president
(Nason, 1990, p. 1), or executive secretary, general manager or association
executive (deck, 80, p. 671).
A second important semantic choice relates to the central issue of this
study: the balance of responsibility and power between the chair and the

executive director. Like all of the research reviewed (cf. Chapter 2), this study
proceeds from the assumption that a nonprofit organization is led by both:
A chair of the board of directorsas required by law
(cf. Hopkins, 1979; Nason, 1990; Oleck, 1980) and as
recommended by OConnell and other prescriptive writers
(cf. Chait, 1992; Conrad and Glenn, 1980; Dayton, 1987) and
An executive directorwho operates the organization on a
day-to-day basis and is seen as the more critical in the empirical
(observational) research of Herman and Heimovics (1990b, p. 168)
and by practitioners.
Yet the study is spurred by the recognition that direct observation and
practice of the day-to-day workings of a nonprofit organization lead to the
conclusion that the primacy of the chair, which is prescribed by the law and
portrayed [or described] in the prescriptive literature, does not appear to
characterize effective nonprofit organizations.
The vocabulary of the research questions was pre-tested, that is to say
tried out, in discussions with more than one dozen chairs and executive directors
of organizations comparable to those chosen by the panel of judges. In
particular, these individuals were asked to state which of the following was
most acceptable to delineate the leadership distributions between the chair
and executive director:

More valuable?
More efficient?
More critical to success?
Principal leader?
Real leader?
More important?
These nonprofit leaders were most comfortable with the term more
important, so that terminology is used in all related questions in the mail survey
and in the interviews. To be clear, however, it should be emphasized that the
term more important, for the purposes of this thesis, means that the respondents
reported that the subject was more critical to the success of the organization with
respect to such issues as general leadership, motivation of the board and staff,
fund raising/marketing, finance and the figurehead/spokesman role.
Three other terms require definition because they are used in different ways
within this study. These terms are management, policy and leadership. The term
management is used by Philip Burgess (Burgess, 1995, p. 705-16) wherein he
describes public management as being composed of a number of management
functions that can be divided between policy, resource and program levels. He
describes policy management as composed of boundary spanning duties like
relations with other levels of government and citizens; resource management as
dealing with finances and personnel and program management as dealing with daily

activities like the maintenance of streets, roads, the provision of health care and
educational services. Similarly, this study will define policy management within
nonprofit organizations as boundary spanning duties, resource management duties as
fundraising and personnel and program management for the nonprofit organizations
included in this study as dealing with the provision of social services or arts
Further, the term management is used in nonprofit organizations to
differentiate administration (management and general) from such other accounting
concepts as fundraising and program expenses. These terms are used by public
auditors in the determination of the amount of time and donated financial resources
devoted to these three functions (AICPA, 1998, p. 26).
The term policy will also be used as it is used in the nonprofit sector in
discussions relating to policy and daily management responsibilities. There is an old
saying in the sector that the board makes policy and the staff carries it out
(OConnell, 1985, p. 44). This study will attempt to consider that in several ways.
Finally, the term leadership is used as it is used in the general literature on
leadership about such concepts as visioning, motivating, the use of power, values,
entrepreneurship and the like as discussed in Chapter 2. It assumes that some
manager, or some group of managers, have the responsibility to accomplish the
policy, resource and program duties attendant to the accomplishment of the mission

of the nonprofit organization. Also, and within nonprofit organizations, leadership is
accomplished by the chair, who is responsible, by law, for specific leadership and
management responsibilities as is the executive director responsible, by the science
of management, to accomplish specific duties and to insure specific results as a
condition of his or her role as the day-to-day manager of the people, finances, and
other resources of the organization.

The nonprofit sector is large and growing; so is its importance to the
American people and the American economy. How this sector and its
organizations are led is important to the people of this country. Research into the
leadership patterns of nonprofit organizations is therefore important. Since so
many authors say that nonprofits are critical to American democracy, it is
important that researchers and practitioners know what works best in different
circumstances within different nonprofit organizations so as to maximize their
Unfortunately, the literature about leadership roles within nonprofit
organizations is in conflict. The prescriptive literature says that the chair has the
ultimate power and responsibility, while the executive director exists to serve the
chair (OConnell, 1985, p. 44). But the empirical literature notes that the executive
director should take responsibility for the central leadership role that engages and
inspires the board to achieve improved performance (Herman, 1991, p. 91). Since
there is a contradiction in the literature, this study will attempt to contribute to the
literature by determining what does and does not work with a purposive sample of

successful and less-successful nonprofits. Since, however, this is a purposive
sample, it should be noted that this method does not allow for the precise testing of
statistical relationships and may limit the generalizability of the findings.
Throughout much of the prescriptive literature on the subject, the
conception of nonprofit organizations is that they are hierarchically structured
with a board of directors in the superordinate position. The board, as led by the
chair, establishes policy, oversees programs, and develops standards to ensure the
accomplishment of the organizations mission. The executive director, who is
hired to assist the board (OConnell, 1985, p. 44) in its efforts to achieve the
mission, should work according to the boards directions. This conception is the
application of what organization theorists have variously labeled the purposive-
rational model (Pfeffer, 1982) or the managed systems model (Elmore, 1978)
for nonprofit organizations. Throughout this study, this model will be referred to
as the hierarchical model. Herman and Heimovics note that:'
These models, which are undoubtedly the theories in use
of most organizational participants, treat organizations
as goal-directed instruments under the control of rational
decision makers where responsibility and authority are
hierarchically arranged (1990b, p. 168)

Leadership is immensely complex. Thousands of books and articles have
been written on the topic. James MacGregor Bums reports that a recent study
turned up 130 definitions of the word (Bums, 1978, p. 2).
The importance of leadership therefore, is obvious. This study seeks to
consider how organizations, particularly nonprofit, charitable organizations, in the
field of human services and the arts, are lead. More specifically, who provides the
leadership in such organizations and how. We know that corporate law in the 50
states declares that the voluntary chairperson of the board serves as the chief
executive officer and holds the legal responsibility, with the other board members,
for leadership within the nonprofit organization (California Corporations Code,
1998, section 5213). But as these organizations grow, there is an increasing need to
have a full-time, professionally trained and experienced leader to serve as the day-to-
day manager/leader, usually with the title of executive director. This study focuses
on how these two people, the volunteer chair and the paid executive director, share
leadership responsibility in selected nonprofits toward the accomplishment of the
charitable mission of these target organizations. To understand that, it is helpful to
better understand the topic of leadership.
John Gardner defines leadership as the process of persuasion or example by
which an individual (or leadership team) induces a group to pursue objectives held

by the leader or shared by the leader and his or her followers (Gardner, 1990, p. 1).
James MacGregor Bums identifies two basic types of leadership: the transactional
and the transforming. The relations of most leaders and followers are transactional
leaders approach followers with an eye to exchanging one thing for another-jobs for
votes, or subsidies for campaign contributions. Such transactions comprise the bulk
of the relationship among leaders and followers, especially in groups, legislatures,
and parties. Transforming leadership, while more complex, is more potent. The
transforming leader recognizes and exploits an existing need or demand of a
potential follower. The result of transforming leadership is a relationship of mutual
stimulation and elevation that converts followers into leaders and may convert
leaders into moral agents (Bums, 1978, p. 4).
William Safire, writing in the prolegomenon to his book, Leadership,
instructs that leadership is an art but that it demands subtlety and involves
complexities (1990, p. 15). Obviously, leadership is not simple or easy to
understand. The concept of leadership is tied up with other complex issues, as
recognized by virtually all who write about leadership. They also discuss such
related topics as vision (Bennis, 1989, 99) (Gardner, 1990, p. 130) (Phillips, 1992,
p. 162-169), integrity (Bennis and Nanus, 1985, p. 184) (Bryson, 1992, p. 41-42)
(Phillips, 1992, p. 51-56), values (Bennis, 1989, p. 168) (Bennis and Nanus, 1985, p.
87-109) (Bums, 1978, p. 19-21), power (Bennis and Nanus, 1985, p. 15-18)

(Gardner, 1990, p. 57) (Lynch, 1993, p. 39-59), motivation (Gardner, 1990, p. 183-4)
(DeVille, 1984, p. 18-23), management (Bennis, 1989, p. 44-46) (Bennis and Nanus,
1985, p. 92-94) (Gardner, 1990, p. 14-16) and more.
The concept and practice of leadership is immense, complicated and difficult
to understand. But all that said, leadership is critical to the success of nonprofits and
all organizations. Bennis points out that organizations are the primary form of our
era and its executives must be social architects (1989, p. 177). If we as a people are
to accomplish what we dream about, we may do it at least in part through leadership
in organizations, despite how difficult and complicated that may be. Nonprofit
organizations are more likely to have impact if their leaders have impact. This
research hopes to increase what we know about leadership in nonprofit organizations.
Leadership in Private Sector
Corporate Organizations
Boards of Directors were invented in America in pre-revolutionary times. In
the earliest days of colonial business, the directors and the investors of a corporation
were largely the same small group. This group would meet regularly down at the
wharf, the counting house, or even under a convenient tree, usually without furniture
enough for a proper sit-down meeting. A board would be laid across some barrels to
form a crude table, and this board came to be the symbol representing the
assembled owners of an enterprise. If the business was feeling particularly hale, a

few stools might be available for the members, but usually the group could afford
only a single comfortable chair. This chair was reserved for the leader of the
assembled squires, who had the privilege of sitting his businesslike self in it during
meetings. This Chairman was usually the largest investor and led the meeting
(Monks and Minow, 1995, p. 180).
By the middle of the nineteenth century, boards were still run by a few people
close to the investors in the corporation. Owners managed and managers owned
their enterprises (Chandler, 1977, p. 37) and investors comprised no more than a
handful of people. Beginning in the 1850s, the railroads blazed a path toward
professional management (Ward, 1997, p. 36) and were soon followed by telegraphy,
meatpacking, consumer machinery and mass retailing. While there was still a chair,
there was now a day-to-day manager. Oddly, boards were still somewhat confused
about their role and this probably explains why so little was written on the subject
until the middle of this century. By the 1930s, control of the corporation began to
slowly slip from the owners of the corporation to the managers and the board of
directors (Roe, 1994, p. 21).
By the end of the 1940s, the postwar corporate board was still a subsidiary of
management. These boards were loaded with inside directors. A study in 1963
revealed that fully 60 per cent of large industry boards had a majority of inside
directors (Vance, 1964, p. 19). Also, boards definitely sought only qualified males

for outside director slots, and white males at that (Juran, 1966, p. 225). As late as
1975, there were fewer than 100 blacks on the boards of the Fortune 500 U.S.
corporations (Johnson, 1975, p. 25). Directors tended to come from the same golf
and social clubs as their CEOs and CEOs sat on each others boards, encouraging a
gentlemans agreement not to raise unpleasant issues (Mace, 1971, p. 89). All in all,
corporate boards have historically been a closed club that exercised little governance
or leadership.
More recently, however, for-profit board experts like Ralph Ward, have
begun to write of the changing role of for-profit boards. In fact, there has actually
been a exodus from for-profit boards. The need to demonstrate social responsibility
and the rigors of late twentieth century American liability have caused many to fear
for their fortunes when they serve on boards (1997, p. 148). Director liability is the
norm in for-profit boards and it is no wonder qualified people will not become
directors without the coverage of directors insurance. The legal costs of defending a
suit can quickly run into substantial six figures (Chorafas, 1988, p. 3).
Modem boards have become much more independent of management. The
roles of investors, owners, directors and management are much more delineated.
Directors had come to realize that a director must be prepared to pass judgment on
men who may be his personal friends (McSweeney, 1978, p. 12). In the 1990s,
such important corporate consulting and support entities as Booz, Allen and

Hamilton, the Conference Board and others now publish how-to books for
corporations, their managers and directors. These works emphasize potential
conflicts of interests, the value of a high proportion of outside directors, an
independent nominating process and the value of diversity on the board (Conference
Board, 1993, p. 3-4). Interestingly, for-profit boards are beginning to look and act
more like the boards of nonprofit organizations.
The Leadership Role in Nonprofit
Organization Management
The success of any organizationnonprofit, governmental or commercial
depends in large part on the quality of its management and leadership. While
management is practiced differently in the various sectors because of the
different goals of their organizations (e.g. profit in commercial organizations), it
should be accomplished as effectively and efficiently as possible. This study
considers leadership practices in one type of nonprofit organization in the hope
that it will lead to a better understanding of leadership in every type.
In his framework of management roles, Henry Mintzberg states that the
manager as leader defines the atmosphere in which the organization will work,
giving the organization direction and purposeclearly among the most
significant of all roles in the organization (Mintzberg, 1973, p. 60-61).

Leadership is clearly worth studying for all who wish to understand what allows
an organization to achieve its purposes.
While Mintzberg identifies leadership as one of the roles of management,
other writers (Bennis, 1989; Bennis and Nanus, 1985; De Ville, 1984; Jones,
1988) express the opinion that leadership is different from management.
Leadership, they state, is involvement in guiding direction, course, action and
opinion; while management means to bring about, to accomplish, to have charge
of or responsibility for, to conduct (Bennis and Nanus, 1985, p. 21). The
difference, these authors appear to believe, is that leaders work with vision while
managers work with administration (Bennis and Nanus, 1985, p. 92 and De Ville,
1984, p. 156). Bruce Jones notes that leadership is seen as a function of
management in the corporate world, but that the terms leader and manager
should not be used interchangeably because while good managers may often be
good leaders and vice versa, it is not always the case (1984, p. 40-41).
Much of the prescriptive literature on the nonprofit boards role and its
relationship to the chief executive and other staff adopts the hierarchical model
(Alexander, 1980; Bower, 1980; Conrad and Glenn, 1980; OConnell, 1976,
1985, 1988; Swanson, 1978). Indeed, as argued by Herman (1989), the
prescriptive literature advances a heroic model of the role of the nonprofit

board. The heroic ideal places substantial and ultimate responsibility for the
conduct of organizational affairs on the board of directors.
Brian OConnell believes that the leader must inspire, govern, market,
monitor and evaluate, motivate and secure necessary resources (OConnell 1976,
1985,1987,1988). OConnells extensive writings on nonprofit leadership will
be cited in detail later (cf. section on page 26, The Division of Leadership Roles
in Nonprofit Organizations).
Yet practitioners often comment on the fact that most boards do not
assume such superordinate responsibility for leading the organization. As a
consequence of the disparity between what the descriptive literature says the
chair should do versus what actually happens, there can be considerable tension
between the chair and the executive director. Herman and Heimovics note, in
spite of the wide-spread popularity of the prescriptive standards, the actual
performance of boards often seems to fall well short of the ideal (1990b, p. 168).
Middletons (1987) thorough review of the empirical literature demonstrates that
boards seldom completely fulfill their assigned roles and, consequently, the
relation between boards and chief executives is neither one of unambiguous
boss/employee nor of partnership. Rather, Middleton notes, board-executive
relations are more accurately described as strange loops and tangled hierarchies
(1987, p. 149). Board members are, in general, highly dependent on executive

directors for information and expertise, and executive directors frequently have a
greater stake inthat is, a careerand identification with the organization
(Herman and Heimovics, 1990b, p. 168).
While this thesis is concerned with the division of the leadership roles
between the chair and the executive, the work of Melissa Middleton sited above
is valuable to the central theme of this study even though it compares the role of
the entire board, rather than only the chair, to the executive. It is valuable but,
unfortunately, it is limited in value to the central theme of this research. While it
does help in the understanding of the division of roles, it is not well focused to
our needs. To enrich this vein a little further, however, it is worthwhile to note a
few additional examples of the board/executive relationship where an effort has
been made to consider the chair/executive relationship, as well.
The National Center for Nonprofit Boards has issued a pamphlet on the
Role of the Board Chairperson. It identifies the chair as the chief volunteer,
responsible for guiding the setting of policy, which is the boards role (Dorsey,
1992, p. 2). It states that the executive advises the board on policy options but
the board retains final authority for determining policy. This publication goes
further to note that the chair and the executive have a significant potential for
friction but that effective people in these roles subjugate their personal needs out
of devotion to the organization. It lists the duties of the executive as similar to

that of the CEO in industry in that the nonprofit executive director leads,
organizes, plans and manages daily operations. The chair, on the other hand,
builds participation, acquires and communicates information, evaluates
performance and delegates to the executive and others (Dorsey, 1992, p. 3-6).
Interestingly, all of the above builds on the belief that the chair operates for the
board in virtually all matters with the executive.
Dividing Leadership Roles between
the Chair and the Executive Director
Brian OConnell has been the most detailed in his description of the
division of leadership roles between the chair and the executive director. It is
very clear that in his view the chair serves the leader role in a nonprofit and
the executive director serves in support of the chair. This is an example of the
hierarchical model found in the literature. As noted again on page 21,
OConnells leadership during the 1980s of the powerful organization,
Independent Sector, put him in the position to have great influence on managers
and other thinkers in the nonprofit field.
Empirical studies of for-profit boards suggest a different practice.
Armand Stalnaker, in a study of the relationship between business (for-profit)
boards and the chief staff officers of the companies (referred to in the article by
the corporate title of chief executive officer), for the Center for the Study of

American Business, discovered some interesting attitudes about the relationship.
Stalnaker interviewed twelve chief executive officers of large firms who also
serve on the boards of other New York Stock Exchange or Fortune 500
companies. He learned from these leaders that they saw the relationship between
the roles as not one in which the board guides the CEO, but the CEO guides the
board (even in view of the fact that the board has the power to hire and fire the
chief executive) (Stalnaker, 1986, p. 4). Further, Myles Mace of Harvard
University discovered after extensive inquiries that for-profit boards do not
typically establish objectives and policies, audit performance thoroughly, ask
discerning questions, or even choose a chief executive officer except in a crisis
(Mace, 1971, p. 78).
The findings of Stalnaker and Mace about for-profit management differ
greatly from the view expressed by OConnell and other prescriptive writers in
the field of nonprofit management. By and large, the literature describing the
optimum relationship within a for-profit board reflects the view that the chief
staff officer (often called the chief executive officer or the CEO) is the leader,
while the chair of the board is the same person or someone who is led by the
CEO; conversely, the prescriptive literature on the same relationships in the
nonprofit field reflects the view that the chair is the leader, while the executive
director exists to support the chair.

Another view, expressed by Herman and Heimovics (1990a), is more
supportive of the executive director as the day-to-day leader of the nonprofit
organization. They report on the results of their empirical study by noting that
nonprofit executive directors are perceived as centrally responsible for the
successful outcomes of their organizations (p. 167). They further note that
nonprofit organization board chairs see themselves as affecting outcomes little
(p. 171). In fact, when an event or activity was successful, both chairs and
executive directors reported that the skills, abilities, hard work and effort of the
executive directors most contributed to the outcome (p. 170). Herman and
Heimovics conclude that the skill that differentiates executives who are regarded
as especially effective in nonprofit organizations is leadership (p. 167). In
another study, the same authors conclude that effective executives work with and
through their boards in order to affect the constraints and dependencies in the
nonprofit organizations environment (1990b, p. 107). Herman and Heimovics
studies suggest that the prescriptive, hierarchical model favored by the bulk of
the prescriptive writers in the fieldis seldom likely to
be fully achievable (1990b, p. 171).
Finally, Umbdenstock and Hageman, writing about corporate leadership in
nonprofit hospitals, point out that the chair and the executive should enjoy a good
relationship because it is critical to the blending of the policy and management

activities into true corporate (nonprofit) leadership. These authors also stress the
need for the chair and executive to be availabile and accessible to each other and to
schedule regular meetings to allow each to serve as counsel, not only about current
matters but also to blue sky with each other. They point out the need for the two
to develop a complementary set of annual objectives for the two positions, each in
support of the institutions goals. Because they are the two most visible leaders of
the organization, it is critical that they be in sync (1984, p. 19-22).
The Division of Leadership Roles in
Nonprofit Organizations
With regard to nonprofit organizations, all of the major prescriptive and
empirical authors share the view that the role of leadership is assumed by both
the chair and the executive director. The disagreement is about which of these
two players has the more important role. Oleck, writing as a generally
recognized authority on nonprofit organization legal questions, notes:
In nonprofit corporations, at least from a legal, if not functional,
perspective, the chairman of the board (chair) has less financial
power than his for-profit corporation counterpart. For example,
he cannot borrow money or issue notes for the corporation (1980,
p. 661).
But Oleck points out that the nonprofit chairman does have the power
to exercise his responsibilities as the organizations general management power
(p. 661). Later, Oleck notes that the nonprofit chairman may, however, delegate

purely administrative powers, though not discretionary powers, to the executive
director, to run the daily, routine business of the organization (p. 671). Again,
the legal view of the division of management in nonprofit organizations assumes
leadership responsibilities for both the chair and the executive director but agrees
with the prescriptive writers by attributing greater responsibility to the chair.
In his writings, OConnell notes that he believes that the leadership roles
between the chair and the executive director should be divided so that the
organization is set up along the heroic, hierarchical model, with the elected chair
as the head. OConnell (1985) says that the elected chair:
Holds basic leadership responsibilities (p. 34)
Leadsdoesnt delegate this function to staff (p. 40)
Plans (p. 34)
Recruits board membersdoesnt delegate to staff (p. 35)
Orients board (p. 38)
Motivates, encourages and prods (p. 39)
Represents the organization to the community (p. 51)
Heads board in the fulfillment of its accountability functions (p. 53)
In the same publication, OConnell lays out the duties of the executive director.
The executive director:
Helps the volunteers do the work of the organization (p. 44)

Serves as expert (acts in an advisory capacity to the interested
volunteers) and as a detail person (carries out the clerical details of
the agencys functions) at the same time (p. 44)
Assists staff so as to leave the big people [sic] free to think
about big things and then gives them [the big people] sufficient
assistance to translate their thoughts into big results (p. 45)
Assists the chair (p. 46)
Helps in recruitment of board members (p. 46)
Has authority over other staff members (p. 50)
As organization grows, adds other delegated functions per job
descriptions (chief program officer, chief spokesperson, chief
fundraiser or planner, or other) (p. 53)
While nTany believe that this diversity helps the sector to be stronger, it
should also be noted that different types of organizations may have differences in
their leadership patterns, and that different types of organizations may require
different types of leadership patterns to achieve success. Conversely, one might
find that certain divisions of the leadership role might be effective in any
nonprofit organization no matter what its size, structure, financial wherewithal, or
mission. Either way, the present study was undertaken in hopes of revealing
which particular divisions of the leadership role are most closely associated with
organizational success, among at least one type of nonprofit organization
sampled for this study.
As previously noted, Brian OConnell has written extensively on leadership

in nonprofit organizations. Many consider him to be an expert on the roles which a
leader of a nonprofit must perform, as well as on the ways in which these functions
should be divided between the elected, volunteer headthe chair of the
organizationand the hired, paid headthe executive director of the organization.
Writing in The Board Members Book (1985), OConnell identifies the
following leadership functions that must be accomplished in nonprofit
organizations. The leader(s) must:
Inspire and lead
Recruit a talented and diverse group of leaders to the board and .
then orient and train them
Dream, inspire and develop organizational mission and purpose
(they know it, review it periodically, work towards its
Plan, develop strategies and set goals
Allocate resources
Raise and allocate financial resources
Account and report
Budget (and control expenses)
Report to the community and all relevant overseers
Insure the completion of an objective audit

Conduct public relations
Interpret the organization to the community
Interpret the community to the organization
Monitor and evaluate operations and programs
Reward and motivate management personnel
Several other authors support this list of leadership duties in whole
or in part (Connors, 1980; Dayton, 1987; Dorsey, 1992; Ellis, 1986;
Gardner, 1990; Houle, 1989; Wolf, 1984).
Models Found in the Literature
At least three models for the distribution of leadership between the
volunteer chair and the paid executive director of a nonprofit organization are
delineated in the literature and in the discussions among boards and managers of
nonprofit organizations. The models are presented here (see also Table 2.1).
In this discussion, the words legalistic and prescriptive will be used.
The word legalistic will mean that which is written in the law, and the term
prescriptive will mean that which gives direction or tells what others should
Model T. The chair is more important to the leadership of a
nonprofit organization than the executive director.

The model in the prescriptive literature states that the chair is the
dominant leader and the executive director is subordinate. As most popularly
expressed in the writings of Brian O'Connell, this view says that the nonprofit
organization leader is the chair and the executive director exists to serve the
needs of the chair. Belief in this model is supported by the legalistic view as
expressed in the prescriptive literature that relates to nonprofit organization
management and the law (cf. Hopkins, 1979; Nason, 1990; Oleck, 1980).
Model 2: The executive director is more important to the
leadership of a nonprofit organization than the chair.
This model is formulated by Robert Herman and Richard Heimovics and
is based on their empirical research as reported in 1990 (Herman and Heimovics,
1990a, p. 167). This model states that the executive director is primarily
responsible for the success of the organization while the chair is seen as having
little impact on organizational success.
Model 3: The chair and the executive director divide
the leadership role roughly equally.
The third model delineates the position of several important prescriptive
writers (cf. Chait, 1992; Conrad and Glenn, 1980; Dayton, 1987). In this model,
the chair and the executive director work together. This position is supported by
Leduc and Block (1985, p. 67-76) who state that the chair provides long-term
leadership and the executive director provides day-to-day leadership. All of these

writers conclude that teamwork between the chair and the staff maximizes
nonprofit mission accomplishments.
The three models (Table 2.1 on the following page) demonstrate ways
that the writers and other commentators believe that leadership authority and
responsibility are distributed between the chair and the executive director as co-
holders of the leadership role in the organization.
One must note the divergence of opinion about which role, chair or
executive director, is more critical to the success of the organization or, to say it
a different way, which role is more important to the organization. It is this
divergence of opinion that will be addressed here.
Ultimately, the study attempts to determine which of the prevailing
models about the division of the leadership role between the chair and the
executive director of an organization is associated with the greatest success for
the organization.

Three Models for the Division of Leadership Role Functions
Between Chair and Executive Director
in Nonprofit Organizations
Chair Executive Director Comments
Superordinate Subordinate Principal writers are OConnell, Hopkins, Nason and Oleck. Prescriptive work with strong bias towards chair as dominant leader while executive director exists only to serve the chair
Principal writers are Herman and Heimovics
Subordinate Superordinate Empirical research that finds that the executive director is primarily responsible for organizational success while the chair is seen as having little impact
Equal Equal Principal writers are Chait, Conrad, Glenn and Dayton. Prescriptive work that states the chair and executive director share leadership responsibilities but the executive director has the time and ultimate responsibility for success

While there has been attention to the study of leadership in the
commercial sector, little attention has been paid to the study of leadership in the
nonprofit sector. Minimal empirical research has been accomplished beyond the
previously mentioned work of Herman and Heimovics concerning leadership in
nonprofit organizations, and no specific empirical research has been
accomplished concerning the topic of dividing leadership between the two
positions of chair and executive director. Empirical research refers to the kind
of quantitative study that, as Earl Babbie notes, is concerned with developing
and testing limited theoretical hypotheses (1979, p. 102).
The experiences of more than 25 years as an executive director (and
occasionally as chair) of nonprofit organizations, coupled with observations of
other managers and leaders in the field, and reading of the literature led this
researcher to recognize the discrepancies between prescription and practice.
Looking at those discrepancies, and looking at what works and does not work, in
the actual practice of dividing leadership between the elected leader and the paid

staff leader of the various nonprofit organizations servedand /or observedled
to the development of several hypotheses.
Research was conducted to collect more information. This thesis was
conducted as exploratory research in the interest of generating hypotheses, rather
than the testing of hypotheses. Non-probability sampling was best suited to this
study because, as Babbie states, .. .it is more timely, and it is less expensive..
(1979, p. 200). While this methodology lends itself to some limitations, it is, I
believe, the best way to accomplish the goals of this thesis study. This research
utilized both quantitative and qualitative research in order to maximize the
probability that this exploratory research would develop interesting hypotheses.
This research also tests three obvious hypotheses as follows, also in the interest of
developing additional hypotheses. The fact that the study started out testing two
hypotheses and found a third to test, is proof that this effort was successful.
Three hypotheses were developed for this study. The first two were based
on the literature review and the researchers observations and experiences. The
third was the result of information gathered as the result of the quantitative
research of this study.

Hypothesis 1: The role of the executive director, or the chair, is the most
important (most critical to the success of the organization) leader within
nonprofit organizations.
Hypothesis 2: The executive director is the most important (most critical
to the success of the organization) day-to-day leader within nonprofit
organizations but supports the chairs legal and long-term leadership
responsibilities. And the chair supports the executive director in day-to-day
leadership but retains the legal and long-term leadership responsibilities.
Hypothesis 3: The responses to the mail survey and the interviews by the
chair and the executive director will be more consistent in successful
nonprofit organizations and less consistent in less-successful nonprofit
organizations. That is to say that the chair and the executive director will
more often agree on their respective roles in successful organizations than
will the chairs and executive directors in less-successful organizations.
Population and Sample
This research focuses on a sample of organizations selected from the
larger population of nonprofit human service and arts organizations in the Los
Angeles/Long Beach community. The population is one in which many members
of the subset are easily identified, but the enumeration of all would be nearly
impossibleand very expensive of money and time. (Babbie, 1979, p. 195-196).
As was previously noted, nonprofit organizations vary greatly in size, scope,
structure, and mission. Methodological and practical considerations made it

unfeasible to undertake a study of the leadership structure across the entire range of
nonprofits. It was decided to focus on arts and human services, but not health,
educational, religious or other types of nonprofit organizations.
A Taxonomy of Nonprofit Organizations
Moyers notes:
The nonprofit sector is composed of approximately
1.140.000 organizations, and each year for more than the
last decade, approximately 29,000 organizations apply for
and receive tax-exempt status from the Internal Revenue
Service. During this century, all segments of the nonprofit
sector (except churches) have increased their numbers at a
rate that outpaces the rest of the economy (1994, p. 8).
The National Taxonomy of Exempt Entities (NTEE), a classification
system for tax-exempt organizations, includes 10 major categories of public-
serving, charitable organizations (see Table 3.1 on the following page).
Moyers comments on this table:
The percentages indicate the distribution of specific
types of organizations within a group of more than
133.000 nonprofit organizations, that account for an
estimated 98 percent of the assets and revenues of all
charitable organizations (1994, p. 9).

Table 3.1 National Taxonomy of Exempt Entities
Arts, culture and humanities 11.4%
Education 13.6%
Environmental 2.8 %
Health 20.4 %
Human services 36.6 %
International, foreign affairs 1.0%
Public and societal benefit 8.7 %
Religion-related 3.3 %
Mutual/membership benefit 1.8%
Unknown/unclassified 1.8%
SOURCE: Richard L. Moyers, ed, Board Member(Washington, D.C.: National Center for Nonprofit Boards, 1994): 1-14
Concerning the entire Third Sector, Jon Van Til has pointed out that it is host
to a bewildering variety of organizations with differing degrees of visibility, power,
and activeness (1988, p. 3). He further states that although they vary greatly in
scope and specific purposes, their general purposes are broadly similarto do things
business and government are either not doing, not doing well, or not doing often
enough. (p. 5)
The organizations studied in the current research were selected from a group
of nonprofits sharing several characteristics. All of the nonprofit organizations are
similar in size, geography and community support. All are human services or arts

organizations. Human services organizations were selected because they are the
largest class of nonprofit organizations, comprising an estimated 36.6 percent of
the entire domain of 501(c)(3) organizations in the U.S. (Moyers, 1994). Arts
organizations are also numerous, representing 11.4 percent of all nonprofit
organizations. Together, these organizations represent 48 percent of the field.
Human services and arts organizations also share virtually the same pool of board
members as evidenced by the fact that most of the chairs of the nonprofit
organizations included in this study also serve on the boards of arts groups (as
determined in the interviews discussed later in this study). Also, these nonprofit
organizations share many of the same sources of philanthropy as evidenced by their
inclusion in the Southern California Association of Grantors (SCAG) list of Los
Angeles/Long Beach nonprofit organizations. Further, as noted on page 43, this
list was used to refresh the memories of the judges as they made their selections of
successful and less-successful nonprofit organizations for inclusion in this research.
Finally, human service and arts nonprofit organizations were both used because the
executive directors from each of the groups tend to be educated in program (social
work, psychology, arts, etc.), rather than administrative sciences.
Only medium-sized nonprofit organizations were included in the sample.
Medium size is defined here as having an annual budget in the range of $ 1,000,000
to $20,000,000. Organizations within this size, compared to small nonprofit

organizations, generally have sufficient staff, board members, and clients served to
be used in assessment of the leadership patterns exercised by the board chair and the
executive director. Medium-sized organizations may also be expected to have
greater longevity and stability than smaller organizations, such that the relationships
between the executive director and the board chair are formal enough to be assessed
in a meaningful way.
All nonprofits included in the sample have a common geographical
characteristic: they are located within the greater Los Angeles/Long Beach area
and all serve clients based in that same area. Although convenience of access was
an important consideration in this decision, the primary reason for this geographic
restriction was to assure greater reliability and validity of the judgments made by
the panel of five expert judges who nominated certain organizations as exemplars
of highly successful and less-successful organizations. This panel of expert judges,
selecting such exemplars from their own geographic area, was much more likely to
have high inter-judge reliability in providing ratings of the degree of success of the
organizations they nominated.
Restriction of the research sample to these particular types of organizations is
based on the following methodological considerations: (1) to permit effective and
meaningful purposive sampling, (2) to increase the ability of the research to detect
variations related to leadership division and (3) to limit the possibilities that serious

confounding variables could arise if target organizations were selected from the
overall population of all types and sizes of nonprofit organizations.
The organizations included in the research constitute a purposive sample,
chosen from the previously described population. Purposive sampling involves
a selection of elements for the sample based on the judgment of the researcher.
Babbie has suggested that purposive sampling may yield the most comprehensive
understanding of the subject matter (1979, p. 215). Because the proposed purposive
sampling method yields a non-probability sample, one that is not necessarily
representative of the entire population, the method does not allow for the precise
testing of statistical relationships and may limit the generalizability of the findings;
but it serves as a beginning point in the exploration for preliminary indications that
certain hypothesized relationships exist. By limiting the sample to medium-sized
nonprofit organizations from the same community, and including only social service
and arts organizations, this research controls for at least some extraneous variables.
Identification of Successful and
Less-Successful Organizations
A three-step process was used to identify and gather information on
organizations to be studied: an initial identification process by a panel of expert
judges; a mail survey, using a survey instrument; and a series of personal

interviews based on questions developed from the literature review and from the
information gained from the mail survey.
Procedures were followed to protect the research subjects from any
adverse effects of study participation. Among those safeguards are the fact that
all respondents were identified by code numbers only in the second and third
steps, only the researcher ever saw the coded list of respondents, returned surveys
have been kept under lock and key, returned surveys and interview notes have
and will be kept in separate locations from the code number list and the data has
been reported only in the aggregate.
In the first step, two groups of nonprofit organizations (i.e. successful and
less-successful nonprofit organizations) served as the eventual sample. A panel
of expert judges, who are knowledgeable about the target population, its
elements, and the general aims of the present research selected the groups. The
judges identities are being kept confidential at their request.
These expert judges were chosen based on the following criteria:
1. A strong awareness of the population of human service and arts
organizations in the Los Angeles/Long Beach area.
2. Significant experience in the management and leadership of
nonprofit organizations as a volunteer chair and/or paid executive.
3. A willingness to participate in this study and to help in the
selection of successful and less-successful human service and arts

This panel of judges who helped to identify the successful and less-
successful organizations were, as noted by Sudman, participating in a form of
purposive sampling, a term used to describe a non-probability sampling method
in which initial respondents are selected by their formal roles (1976, p. 210).
After the five judges were chosen, they were asked to identify 10-15
successful nonprofit organizations and 10-15 less-successful nonprofit
organizations from within the Los Angeles/Long Beach arts and human service
agency community. The number of judges was set at five because this seemed to
be sufficient to achieve significant inter-judge consistency, and the number of
recommendations was set at 10-15, because this was what the judges were
willing to do. The judges were sent a Contact Letter (Appendix A), which
provided the list of criteria to be used for choosing participating organizations.
The judges were asked to select successful and less-successful agencies by
applying similar criteria to evaluate each agencys:
Reputation within the community as a successful organization
Success at fundraising
Success at board development
Success at client satisfaction
Success in accomplishing its mission
The judges were supplied a list of the grantees from the member
organizations of the Southern California Association of Grantors (SCAG) to

refresh their memories about the nonprofit organizations in the Los Angeles/Long
Beach community. After allowing several days to make their choices, the judges
gave their independent nominations to this researcher, who then compiled the final
list. Frequency of nominations generated by these expert judges were ultimately
used to construct an ordinally ranked list of 20 of the most successful (most highly
rated) nonprofits for the identified population, and another ranked list of 16 of the
less-successful (lowest-rated) organizations. This approach was convenient for the
experts and gave greater assurance that the researcher could reap the benefits and
strengths of the five different expert judges individually generated ideas.
While each of the judges was comfortable identifying the successful
nonprofit organizations in the community, they were much less comfortable
identifying the less-successful nonprofit organizations. The judges needed to be
reassured that their involvement was going to be kept confidential and that their
suggestions about the less-successful nonprofit organizations would not be taken to
mean that these organizations were unsuccessful or that the members of their
boards and staffs were not wise and hard working people.
To some degree, it was necessary to allow the judges to define the term they
would use for the relatively worse nonprofit organizations. They eventually agreed
to refer to these nonprofit organizations as less-successful rather than
unsuccessful. It was further agreed that the judges would designate as less-

successful those nonprofit organizations which were notorious for problems (many
of the less-successful nonprofit organizations were in the local newspapers for
problems of management) or which were otherwise known in the community for
their problems with funding (or for the fact that their boards of directors or
management were failing to secure the necessary funding). Further, all the judges
were concerned to make sure that their choices of less-successful nonprofit
organizations would be congruent with the choices that other seasoned veterans in
the field would make.
After working with the researcher in person and by phone for a total of
40 to 105 minutes apiece, each judge identified at least six to ten successful and.
four to ten less-successful nonprofit organizations in the Los Angeles/Long Beach
area. A compilation of their choices led to a list of 30 successful and 26 less-
successful nonprofit organizations, of which 20 of the successful and 16 of the
less-successful nonprofit organizations were mentioned by at least two of the
expert judges. This process secured the goal of significant inter-judge reliability.
Importantly, there was not a single instance of an organization being chosen as
successful by one judge and less-successful by another.
The overall selection process achieved the methodological considerations
identified in the sections on Population and Sample and Identification of
Successful and Less-Successful Organizations. These considerations were: (1) to

permit effective and meaningful purposive sampling; (2) to increase the ability of
the research to detect variations related to leadership division; and (3) to limit the
possibilities for serious confounding of variables that could arise if target
organizations were selected from the overall population of all types and sizes of
nonprofit organizations.
The expert judges recommended an overlapping list of organizations. That
is to say, many of the nonprofit organizations were identified as successful or less-
successful by two or more judges. Recognizing and recording this overlap provided
greater confidence that these successful and less-successful organizations were truly
so. The judges were very consistent in their nominations. Of the 20 successful
organizations chosen by the judges: 10 organizations were chosen by 2 judges, 4
organizations by 3 judges, 4 organizations by 4 judges and 2 organizations were
chosen by all five judges. Of the 16 less-successful organizations: 12 organizations
were chosen by 2 judges, 2 organizations by 3 judges and 2 organizations by 4
The next two steps are discussed in the following section.

Table 3.2 Consistency of Judges' Choices
Number of Judges Successful Organizations Less-successful Organizations
2 10 12
3 4 2
4 4 2
5 2 0
Total 20 16
Data-Collection Instruments
After the judges identified the successful and less-successful organizations
and this researcher developed a final list of successful and less-successful groups,
two types of survey techniques were employed. For each organization identified,
this researcher conducted: (1) a mail survey to solicit the responses of the two key
people of each leadership dyad (the executive director and the chair of the board)
and other select members of the organization leadership in the persons of key board
and staff members, followed by (2) intensive interviews with members of the
nonprofit organizations leadership dyads (other organizational leaders were not
interviewed). The mail survey informed the quantitative aspects of this research,
while the interviews provided important qualitative data. Pre-testing of the mail
survey and interview instruments was accomplished by administering it to
individuals not otherwise involved in this research project who serve on boards
and as chairs and executive directors. This will be discussed in more detail later
in the paper.

After the researcher determined the final selections for the more and less-
successful organizations, an initial contact letter, followed by an information-
gathering instrument, was sent to the 36 sets of nonprofit organization leadership
dyads (20 successful and 16 less-successful) and to other leaders to assess their
perceptions of the success of their own organization and to gauge the relative
influence of the chair and the executive director in the leadership of each
organization. Follow-up phone calls were made to encourage full participation
by those contacted by mail.
(1) Contact letter. Leadership dyads of the 36 organizations selected for
the research received a contact letter that was developed to open the relationship and
give the executive directors and chairs ownership of this project within their
organizations (Appendix B). The letter was intended to introduce the research, foster
a positive and cooperative attitude regarding the research, attempt to persuade the
organizational leaders to participate in the research, and let the organizational leaders
know that they were to be contacted in person and asked for their input. Care was
taken to make sure that this letter did not introduce bias. Subsequently, the dyads
of nine of the successful and eight of the less-successful organizations participated in
the research.
(2) Mail-survey instrument. Accompanying the contact letter was the

mail-survey instrument. The instrument consists of a fact sheet and a list of 15
questions about leadership as it is practiced in the particular nonprofit organization.
The first 14 leadership questions in the mail survey instrument were scored on
Likert-type scales with values of one to seven. One question was in four parts
so that 17 separate questions were asked in total. These questions were adapted
from OConnell (1988c) and Burgess (1975). Respondents were asked to choose
a number on the scale to indicate whether a function was practiced more by the
chair (1 designated that a function was strongly associated with the role of the
chair) or by the executive director (7 designated that a function was strongly
associated with the role of the executive director). If the respondent circled
number four, s/he was indicating that the role was evenly divided between the
chair and the executive director. If the respondent circled number two, the
function in question was being identified as mostly the responsibility of the chair,
but with some input from the executive director. Similarly, designating number
six indicated that the function in question was primarily the responsibility of the
executive director, but with some input ffom the chair. Number three indicated
the respondent felt that the function in question was primarily the role of the
chair but with significant influence ffom the executive director. Similarly,
number five indicated the opposite (Appendix C). The questionnaire closed
with an open-ended question that asked, What else should have been asked?

(3) Intensive interview guide. The third and final step in the data-gathering
process was to conduct intensive structured interviews with the dyads. The
interviews focused on the qualitative aspects of the relationship between the
chair and the executive director. The researcher attempted to determine how this
relationship works, building on the questions of the mail survey but attempting to
get respondents to expand on that data to share the specifics of their personal
attempts to lead the organizations. In this way, the researcher hoped to accomplish
with intensive interviewing what could not be done with the survey instrument.
(Interview Guides are presented in Appendices D and E).
Separate interview guides were developed for interviewing the chairpersons
and the executive directors. Both the guide for interviewing the chairs and the guide
for interviewing the executive directors contained the same four general areas of
topics to be covered. Each area included carefully worded questions. The guides
served to keep the interview on topic, pace the interviews relative to the amount
of time each respondent was willing to provide for the interview, and phase the
questions from one topic area to the next.
Each respondent was asked at the beginning of the interview to estimate the
amount of time they were willing to provide. Virtually all gave more time than they
initially committed. Every effort was made to keep the interview process as
spontaneous as possible in order to increase the likelihood that lively and unexpected

answers could be obtained from the interviewees. An equally strong effort was made
to maintain equality of time spent with each interviewee and to determine that certain
core questions were asked of each interviewee. Each interview lasted 35 to 85
minutes. During the interview as many questions as possible were covered. The
longer interviews resulted from the fact that all of the interviewees were generous
with their time and some had a lot to say about these topics.
The purpose of the interviews was to collect qualitative data that would
enrich the research. Interviews are more likely to provide nontraditional opinions
and observations than do mail surveys. The interviews were conducted in a semi-
structured way that would allow for an analysis by comparison. No statistical
analysis was attempted (cf. earlier discussions of purposive sampling identified
strengths for initial, exploratory study of an area and limitations for statistical
Pre-Testing of the Questionnaire
and the Interview Tools
In the interest of strengthening the quantitative survey and the qualitative
interview tools, and other issues, this researcher sought the advice and counsel of 10
nonprofit organization leaders not otherwise involved in this research. The pre-test
group was composed of chairs and executive directors from organizations similar to
the organizations ultimately chosen as the subject organizations of this study. Several

of these pre-test members were, in fact, executive directors of organizations
recommended by the judges who were unable to secure the participation of their own
chairs, thereby negating the possibility that they could serve as dyads for this research.
Others in the pre-test group were people who had served as past chairs of the target
organizations, or of other major nonprofit organizations within and outside of the Los
Angeles and Long Beach community. Still others were people who worked at local
foundations and the United Way, who fund organizations like the ones ultimately
secured for this study.
This pre-testing was accomplished in order to identify at least two potential
problems of (1) interviewee confusion about specific vocabulary, and (2) anything
else in the questionnaire and/or the interview guides that could be problematic.
Pre-testing was also done to associate those activities identified in the questionnaire,
with the day-to-day management duties of an executive director, as required to test
Hypothesis 2 of this research.
Two problems with the research tools were identified. One was that most of
these pre-testers felt that it would be very difficult to involve enough dyads from
those chosen by the judges to make a large enough sample to accomplish worthwhile
quantitative, and to a lesser degree, qualitative research. The pre-test group
recognized that to make this research valuable, there needed to be involvement by
two groups of very busy and important people, the executive directors and the chairs

of the select group of Los Angeles/Long Beach nonprofit organizations, chosen by
the judges. These comments lead this researcher to believe that it would be relatively
easier to involve the executive directors into this research but much more difficult to
involve the volunteer chair people. The chairs, they felt, would be too busy, too self-
important, and otherwise unimpressed by the value of the research to include this
activity in their busy schedules.
As feared by the pre-test group, these chairs turned out to be very difficult
to involve in this research because of reasons noted above. The chairs are leaders
from the major for-profit organizations of their community, such as large
corporations, banks, accounting and legal firms, and utilities. It not only proved
difficult to involve these chairs in this study, but the ones who did agree to
participate proved very difficult to keep involved. The ones who participated in
this study proved to be very difficult to keep involved, even after they agreed to
participate. This was even true in several cases where the executive director, to
whom this researcher first appealed for involvement, had volunteered their
personal time for this research but failed to gain the acquiescence of their chairs.
The pre-test group noted that since the researcher was asking so much of
the respondentstime to fill out the questionnaire and time for personal
interviewsthe researcher should be realistic about how many dyads could be
involved. If possible, compromises should be made between the goals of the

research and the demands of the dyad members. Therefore, and in the interest of
securing as many dyads as possible, the pre-test group advised this researcher to
keep it simple, stupid. Their feeling was that this research should get what it
can and not unduly confuse the dyads with the vocabulary of research. They
advised this researcher to try to determine general patterns of leadership,
importance and success rather than to seek specific and precise conceptual
underpinnings of these terms.
It was therefore decided that the researcher, while still building on the
work of Brian OConnell, should aggregate such concepts as leadership,
criticality, importance and success into a blended concept that the pre-testers
believed the chairs would recognize and understand. The pre-test group felt that
all of these concepts are positive and all are components of organizational
success. These terms could be used interchangeably, the pre-testers believed,
if not synonymously. Experience bore this out. Even after limiting the questions
to issues ofimportance and avoiding questions about similar concepts, three of
the chairs still spoke to their confusion about all of these concepts. Two even
spoke to their desire to just stick to the obvious relationships between many of
the questions and advised the researcher to just speak to the issue of who is
more important to the organization. The remaining 14 chairs, however,

responded positively to this researchers questions about their comfort level with
the questions in the questionnaire and the interviews.
The second use of the pre-test group was to determine which of the roles
specified in the questionnaire should be identified as day-to-day duties. The group
identified those day-to-day management duties as fundraising; finance and daily
administration; motivator of volunteers and staff and primary contact to media,
funders, clients and the general public. The need to identify the day-to-day functions
was necessary to consider the second hypothesis, which spoke to the division of the
day-to-day management duties.
Mail Survey
After the list of 36 nonprofit organizations was developed (20 successful
and 16 less-successful), a letter was sent to the executive directors and chairs of the
organizations. Every effort was made to involve each of the 36 dyads. Yet many
did not take part in the research study. As noted in the previous discussion about
the pre-testing of the research tools, there was considerable difficulty involving
the chairs of these organizations.
Among the 20 successful nonprofit organizations, two of the executive
directors decided not to participate, and in the eight additional cases, informed this
researcher that their chairs did not want to participate. One of the successful
organizations executive directors filled out the survey, but the chair did not return

a completed survey, even after several requests to do so. The fact that these
organizations did not choose to participate was not surprising, as noted by the
members of the pre-test group, these are among the most important and busy people
in the community, so the request for time for this research was too burdensome.
While those in the remaining nine successful organizations whose dyads
completed the mail survey were also busy, their chairs and executive directors felt
a commitment to this research and made time to provide information. These nine
represented 45 percent of the nominated successful nonprofit organizations.
Among the less-successful organizations, the researcher was unable to
contact anyone in two of the nominated organizations. In five of the cases,
the executive director or the chair chose not to participate. In another case, the
executive director had recently separated from the organization and the chair,
speaking for the organization, informed this researcher that the organization was
closing its doors. Eventually, eight of the nominated less-successful organizations
participated, a 50 percent participation level.
Thus, nine successful and eight less-successful organizational dyads were
ultimately involved in the surveys and the follow-up interviews. All 17 chairs and
all 17 executive directors completed and returned their questionnaires that had been
pre-tested for clarity of vocabulary. However, one of the executive directors and

four of the chairs required additional mailings, calls and faxes to secure their
participation. Table 3.2 shows potential and actual participation in the study.
Table 3.3 Percentages of Potential Participation to Actual Participation
Respondents Asked to Participate Participated
Successful nonprofit organizations 20 9 (45 %)
Less-successful nonprofit organizations 16 8 (50 %)
Executive Directors 17 17(100%)
Chairs 17 17(100%)
Frequencies, t-test, and discriminant analyses were conducted on the data. The
results of the survey are reported in the following section, Analysis of the Data.
Structured Interviews
The third and final step of the information-gathering process was to conduct
a structured interview with each member of the 17 dyads34 individuals.
Everyone was interviewed in person or by phone, using the interview guides
identified as Appendix DInterview Guide for the Chair and Appendix E
Interview Guide for the Executive Director. These interview guides had been pre-

tested to minimize confusion over the use of terms used. No attempt was made to
conduct follow-up interviews with the non-chair board members or staff members.
The level of cooperation among these 34 participants was very strong. The
chairs and executive directors made themselves available and were generous with
their time. Many of them offered ideas beyond formal questions, and all asked to
be notified about the results of the study. Several of the dyads asked this researcher
to meet with their boards upon completion of the research. The results of the
interviews are reported in the section of this chapter entitled Discussion of
Qualitative Data from the Interviews.
The interview guides for the chair and the executive director are similar in
form. Each is composed of four sections with carefully worded questions within
each section. The four sections are designed to (1) make the respondents
comfortable and put them in a frame of mind to talk about nonprofit organizations
and their personal experiences; (2) continue with a discussion of the division of
leadership in their particular nonprofit organization; (3) encourage respondents to
expand on their views about the function of leadership within the organization and
the people involved therein, as well as talk about their views on division of
leadership in nonprofit organizations in general; and, finally, 4) allow respondents
to discuss other issues of nonprofit organization management and governance they

might be wanting to discuss. In every case, the intent was to maximize the time
available for discussion on section two.
Analysis of the Data
Information gathered on the demographics and leadership questions
sections of the mail survey was analyzed in several ways. Discriminant analysis
was applied to the responses from the 17 chairs and 17 executive directors. A
discussion of this analysis, as well as the research conclusions and
recommendations, can be found in Chapter 5.
Analysis Techniques/Statistical Approaches
Responses were analyzed in several ways. The responses were analyzed
for frequencies of occurrence in each descriptive category. These responses were
then analyzed in groupsfrequencies of occurrence in successful and in less-
successful organizations. Finally, comparisons were made. Responses to the
first 14 leadership questions produced information on 17 leadership functions,
since question 14 contained four parts. Question 15 was an open-ended question.
Because items 1 through 14 of the leadership questions were
systematically developed to represent distinct variables of interest to the study
(success of sample organizations), items were aggregated into subscales. These
subscales can be identified as leader, motivator, marketer, financier and

figurehead/spokesperson. To test the hypotheses of differential leadership
distributions between the chair and the executive director, discriminant analysis
technique was used to test hypotheses related to the following: (a) the chair is
more important than the executive director and (b) the executive director is more
important than the chair. The term important is used here to mean that one
leader is reported to be more responsible for the ultimate success of the
organization than the other, with respect to the specific subscale.
For each subscale, average (mean) scores were computed and were used
in subsequent analyses. The reason for using average rather than total scores is
that the different subscales may have different numbers of items. Next, the
subscale scores were used to compare the levels of consistency/inconsistency
across the two groups of organizations (successful/less-successful).
Finally, and keeping in mind the multi-level nature of this study, a
discriminant analysis was used to compare the successful and less-successful
groups of organizations. For an excellent discussion of discriminant analysis, see
Multivariate Analysis: Techniques for Educational and Psychological Research,
second edition, Tatsuoka and Lohnes, 1988, pages 210- 265. In the discriminant
analysis, the levels of success of organizations (successful/ less-successful) were
used as grouping variables and subscale scores, (leader, motivator, marketer,
financier and figurehead/ spokesperson) as well as the differences scores, were

used as discriminating variables. The results of discriminant functions, as well as
descriptive statistics, helped in testing the hypotheses used in this study.
Limitations of This Study
It is important to recognize that this thesis is exploratory in nature and
designed primarily for hypotheses generation rather than hypotheses testing.
Also, because only a relatively small purposive sample of medium-sized,
501(c)(3), arts and human service organizations serving the Los Angeles/Long
Beach area were studied, the results, while suggestive and helpful, cannot be
generalized to all nonprofit organizations. Again, this is an exploratory study;
and it is expected that subsequent studies will need to be accomplished surveying
larger samples in order to determine the generalizability of the propositions
developed here. It is the expectation of this researcher, however, that similar
results would be found. Finally, it should be emphasized again that while it is
possible to do statistical analysis with purposive sampling, the level of
generalizability is less.
The model in the prescriptive literature, called the hierarchical model in
this paper, says that the chair is the leader and that the executive director exists to
serve the needs of the chair. Given the fact that a marked discrepancy exists
between the prescriptive literatures statements on the division of leadership and

responsibility between the chair and the executive director and the observed
balance of that power and responsibility in actual practice, this research sought to
determine which of the modelsthe hierarchical model from the prescriptive
literature, or the observed actions of boards and executive directors from the
empirical literatureled to the greatest effectiveness of a nonprofit organization.
What experience and practice indicate concerning the effective distribution of the
leadership functions between the executive director and the chair, should guide
what is being taught within the academy, as well as what training and technical
assistance is provided within the field.
Three hypotheses were developed. The first two were developed based
on the literature review and the researchers observations and experiences. As
discussed in the Terminology section of Chapter 1, the term important was
used in the questionnaire, the interviews and in the following hypotheses. Pre-
testing, that is to say that testing on people similar to those of the sample, of the
questionnaire and the interview guides lead to the use of this term over such other
options as valuable, efficient and others. As noted, the respondents chose
more important over their second choice of more critical to the success of.
Hypothesis 1: The executive director, or the chair, is the most important
(most critical to the success of the organization) leader within nonprofit

Hypothesis 2: The executive director is the most important (most critical
to the success of the organization) day-to-day leader within nonprofit
organizations but supports the chairs legal and long-term leadership
responsibilities. And the chair supports the executive director in day-to-day
leadership but retains the legal and long-term leadership responsibilities.
Information was collected and analyzed to determine if the role of the
chair is more important to the success of the organization; if the role of the
executive director is the more important to the success of the organization; or if
the chair, board members and others advise the executive director, who in turn
guides the board to the accomplishment of their legal duties and the
programmatic and administrative success of the organization.
In the course of information gathering, the researcher made an observation
that led to the development of the third hypothesis. This hypothesis will be
discussed in detail in Chapter 5. This hypothesis was not envisioned during the
development of this research but only became apparent upon analysis of the survey
materials. Hypothesis 3 was considered, however, in the interview portion of this
Hypothesis 3: The responses to the mail survey and the interviews by
the chair and the executive director will be more consistent in successful
nonprofit organizations and less consistent in less-successful nonprofit

The various data-gathering activities described in the Methodology chapter
yielded important information that was then analyzed in several ways. It is the
intent of this chapter to report the findings of this study in detail and begin to
consider the importance thereof. That consideration will be fully accomplished in
Chapter 5.
Demographic Information on the Respondents
to the Mail Survey
The 34 respondents to the mail survey represented 17 chairs and 17
executive directors from the 17 nonprofit organizations of the survey, nine that
had been judged successful and eight, which had been judged less-successful.
The demographic information gathered on respondents is shown in Appendix G.
The data failed to reveal any significant differences between the
demographics of the successful and less-successful nonprofits. The data compiled
in Appendix G in this survey notes the similarity of the respondents from the
successful and less-successful nonprofits with regard to age, gender, duration of
service in the nonprofit sector, service on nonprofit boards, level of education,
field of education, professional field and political views.

Responses to the Leadership Questions
in the Mail Survey
As previously noted, the mail survey was divided between demographic
questions about the respondents and questions concerning the leadership of
nonprofit organizations. As indicated in the section on Data Collection
Instruments, respondents marked their answers to the first 14 leadership
questions on a Likert-type scale of 1-7. Question 14 had four parts, so responses
were collected concerning 17 leadership functions in total. Scores greater then
4 meant that the respondents identified the function as that of the executive
directors, while scores of less than 4 mean that the respondents identified the
function as that of the chairs. The closer to 1 the more the function is identified
with the chair, and the closer to 7; the more the function is identified with the
executive director.
Comparisons between the responses for successful and less-successful
nonprofit organizations were made using an independent group t-test to statistically
test the differences between group means. Table 4.1 shows the results of that
analysis. Column two shows the means and standard deviations for the answers
that all of the 34 respondents gave to the questions on the 17 leadership functions.
The third and fourth columns show the means and standard deviations grouped by
successful and less-successful organizations. The fifth column provides the p and

column six provides the level of significance for the t statistic for the difference
between group means. To avoid inflation of the type I error rate, the within cell
sum of the squares was used in the denominator.
Table 4.1 Means, Standard Deviations, t-value, and p for All 17 Leadership Functions (n=34)
Function All Responses Responses from Successful NPOs Responses from Less- Successful NPOs t- value Signifi- cance
1 Inspiration Mean SD 4.53 (1.35) 4.89 (1.13) 4.13 (1.50) 1.69 .10
2 Primary Leader Mean SD 4.30 (1.61) 5.00 (1.28) 3.56 (1.63) 2.83 .01
3 Fundraiser Mean SD 4.38 (1.30) 4.33 (1.09) 4.43 (1.55) -.23 .82
4 Financial Manager Mean SD 5.68 (1.15) 5.50 (1.20) 5.88 (1.09) -.95 .35
5 Daily Manager Mean SD 6.65 (0.49) 6.67 (0.49) 6.63 (0.50) .25 .81
6 Board Motivator Mean SD 3.62 (1.26) 3.78 (1.11) 3.44 (1.41) .78 .44
7 Staff Motivator Mean SD 6.29 (1.03) 6.56 (0.51) 6.00 (1.37) 1.61 .12
8 Volunteer Motivator Mean SD 4.41 (1.52) 4.25 (1.29) ' 4.56 (1.75) -.57 .57
9 Public Relater Mean SD 5.32 (1.15) 5.50 (0.99) 5.13 (1.31) .95 .35

Table 4.1 (cont.) Means, Standard Deviations, t-value, and p for All 17 Leadership Functions (n=34)
Function All Responses Responses from Successful NPOs Responses from Less- Successful NPOs t- value Signifi- cance
10 Public Speaker Mean SD 4.71 (1.38) 4.83 (1.46) 4.56 (1.32) .56 .58
11 Most important to Org. Success Mean SD 4.74 (1.36) 5.00 (1.14) 4.44 (1.54) 1.22 .23
12 Long-term Planner Mean SD 4.53 (1.48) 5.00 (0.91) 4.00 (1.83) 2.06 .05
13 Legal Leader Mean SD 2.70 (2.14) 2.53 (2.00) 2.88 (2.34) -.46 .65
14 Media Contact Mean SD 5.28 (1.67) 5.50 (1.46) 5.06 (1.88) .74 .47
15 Funder Contact Mean SD 4.70 (1.57) 4.88 (1.36) 4.50 (1.79) .69 .49
16 Client Contact Mean SD 5.59 (1.43) 5.94 (0.85) 5.25 (1.81) 1.38 .18
17 Public Contact Mean SD 5.22 (1.50) 5.38 (1.59) 5.06 (1.44) .58 .56
Remember, again, this is explorative research. Further the sample is small and
purposive. The nonprofit sector is huge and the organizations that comprise it are not
alike in many important ways. However, Table 4.1 suggests a number of interesting
findings including the following: two of the leadership questions appear to be

significant, who is the primary leader and who should lead the long-term planning
1. With regard to the question about who is the primary leader in their nonprofit
organization, the chairs and executive directors from the successful nonprofit
organizations see the executive as accomplishing that role while the leaders of
the less-successful nonprofit organizations see the role as that of the chair. This
is the role about which there is the greatest disparity (5.00 3.56 = 1.44) of all
2. the leadership questions. This demonstrates the greatest difference of opinion
between the leaders of the successful and less-successful nonprofit organizations.
The level of confidence about this question is at the 99% level.
3. The question about who should serve as the lead planner for their nonprofit
organization resulted in the second largest disparity between the successful and
less-successful nonprofit organizations (5.00 4.00 = 1.00). Again, the leaders
from the successful nonprofit organizations appear to feel strongly that the
planning function is that of the chair while the leaders from the less-successful
nonprofit organizations saw the role as evenly divided between the chair and the
executive director. The level of confidence about this question is at the 95%
Analyzing the data from another perspective may be valuable. Since the
respondents were instructed to assign roles between the chair and the executive

director on a 7-point Likert-type scale, the assignments at the one and seven levels
would represent the strongest level of selection. Similarly, selections at the levels
of two and six would represent the next highest levels of confidence in the
appropriateness for either the chair or the executive. Scores at the levels of 3
and 5 would reflect weaker opinions for allocating a specific role and a score of
4 would reflect that the respondents saw the role as being equally divided between
the two players.
There were no allocations at either the 1 or 7 levels but the respondents
assigned scores of greater than 6 (i.e. that is to say they strongly saw the role as
that of the executive directors) to the roles of day-to-day manager and motivator of
the staff. High scores were also assigned to the executive director for the role of
financial manager (5.68), and for providing client contact (5.59), public relations
(5.32) and contact to the media (5.28).
Question number 13 had scores that were most strongly assigned to the chair,
implying that the respondents view the chair as the legal leader of their nonprofit
organization (2.70). The role of the motivator of the board was also assigned to the
chair, but at a much weaker level (3.62). The chairs and executive directors from the
less-successful nonprofit organizations also assigned the role of primary leader to the
chair, but at a very weak level (3.56). As noted above, the chairs and executive

directors of the successful nonprofit organizations assigned that role to the executive
director (5.00).
The rest of the roles, provider of inspiration, the primary leader, fundraiser,
motivator of the volunteers, public speaker, most important to the organization, long-
term planner and contact to funders were seen as divided between the chair and the
executive director, but closer the executive director.
Discriminant Analysis
As indicated earlier, the focus of this study was to compare the successful
and less-successful organizations and to find out in which area(s) they differed
significantly. The most interesting comparisons were in the area of leadership
functions. There were 17 leadership function questions and each, the successful
and less-successful organizations, could have been compared by using a t-test.
However, the main problem with a multiple t comparison is the inflation of the
type-I error rate (Moore, 1993, p. 484). Thus, using all the leadership questions in a
multivariate model seemed to be the choice and a discriminant analysis was deemed
more appropriate.
Initially, the researcher conducted discriminant analysis on five composite
scoresleader, motivator, marketer, financier, and spokesperson, using these as
discriminating variables, and using chair and executive director as grouping
variables. The results indicated that there were not any significant differences

between the chair and the executive director on any of these five composite
variables. The lack of significant results may have been caused by measurement
error due to creating composite scores from questions that are not highly correlated.
To overcome the problem of measurement error due to the creation of the
composite variables, all 17 items were included.
Thus, a discriminant analysis was conducted using successful and less-
successful as grouping variables and all 17 leadership functions as discriminating
variables. This analysis yielded one significant function (chi square = 8.787,
df = 2, sig = .0124). Table 4.2 provides the discriminant analysis structure matrix.
Two variables, Q. 12. (long-term planning by the executive director) and Q. 3.
(raise funds by the chair), had significant correlation with the discriminant function,
indicating that these questions have the power to discriminate successful and less-
successful organizations. Q. 1. (inspire organization) and to some extent Q. 11.
(more important), while not statistically significant, also are able to discriminate to
some extent between successful and less-successful groups.

Table 4.2 Discriminant Analysis Structure Matrix for the 17 Leadership Functions
Q12 Long-term Planner .68408
Q3 Fund Raiser -.46871
Qi Inspiration .24608
Qll Most Important to Org Success .20890
Q5 Daily Manager .15035
Q8 Volunteer Motivator .14623
Q7 Staff Motivator .14215
Q14B Funder Contact -.13007
Q6 Board Motivator .11288
Q10 Public Speaker -.10285
Q13 Legal Leader .07296
Q9 Public Relator .07126
Q2 Primary Leader .06448
Q4 Financial Manager .02965
Q14A Media Contact .02324
Q14D Client Contact .01041
Q14C Public Contact -.00965

Securing Qualitative Data
from the Interviews
The qualitative data in this study was collected through structured
interviews with chairs and executive directors, conducted after the mail survey.
The researcher had already prepared structured interview guides (Appendix D.
and E.) for each position before the mail survey with the help of the pre-test group,
but adjusted them with the benefit of what was learned from the quantitative data.
These questions asked about actual practices in the nonprofit organization. Six of
the respondents to the mail survey answered the open-ended question 15 by saying
that the researcher should also ask about what ought to happen. Accordingly,
What ought to happen questions were added to form pairs with the What does
happen questions.
As noted, interviews were conducted with all 34 members of the 17 dyads.
By the time the interview stage of the process arrived, the respondents had
demonstrated their commitment to the study by filling out and returning the mail
surveys. These 17 dyads were, therefore, responsive and interested in the
interviews. Most were curious about what the researcher would do with the study
and how it could be put to use for their organizations. The researcher agreed to
make the thesis available to the respondents, to make oral reports to the respondents
and /or to their boards and to copy them on any articles that might be written from

this data. In every case, the researcher was able to involve the respondent in a
comfortable and productive partnership. The executives of the less-successful
organizations were not identified as such to them.
Thirty-one of the interviews were conducted by phone and 3 were
conducted in person. Each interview lasted 35 to 145 minutes. Four of the
interviews were conducted over two separate meetings in order to allow sufficient
time to complete the interview to the satisfaction of the interviewee and/or the
researcher. The interview guides, identified as appendix D. and appendix E. of
this study, were used. The instruments proved to be very helpful at keeping the
interviews on target and moving at a fast pace. In some of the interviews, the
questions were asked in order, in others the questions were varied due to the
flow of the conversation.
In every case the interviewees were confident in their responses, and most
were quite opinionated, having been encouraged by the researcher. A few were
even excited about the opportunity to share their views. One of the chairs told this
researcher that he had been waiting to be contacted because he had been thinking
about these kind of questions for years and was happy to be able to share his
insights. Every respondent took the interview seriously and even the busiest
people (the ones from whom it was most difficult to secure sufficient time)
eventually allowed as much time as needed to complete the interviews to the

researchers satisfaction. In every case, the respondents were thanked and assured
of their importance to the study.
Responses to Questions in
Interview Section One
Section one of the interviews was concerned with the goal of making the
respondents comfortable and putting them in the frame of mind to talk about
nonprofit organizations and themselves. Again, the counsel of the pre-test group
was very valuable in dealing with the limited time availability of the respondents,
especially the chairs. Because of the limited time available for these interviews, the
researcher was not able to ask every question of every respondent. Considerable
information, however, was collected and a number of important impressions were
made. This first part of the interviews lasted on average 8 to 15 minutes. As noted
in the section describing the demographic information on respondents, the sample
was very experienced in nonprofit management and governance. It was obvious
that the respondents were interested in and dedicated to their professional, and/or
volunteer work.
As previously noted, most of the chairs are corporate leaders or senior
partners from local corporations, banks and legal and accounting firms. They
understood that they were performing a role in support of their and their firms
community involvement. Many mentioned the limited term of their leadership

and that the executive director represented the continuity of leadership for the
organization. Many of the chairs were long-time board members of the specific
organization under discussion and many had served in other offices and committee
chair positions. Other chairs were doing a professional tour of duty in Los
Angeles/Long Beach and had only been with this specific nonprofit organization
a short time, but they had been involved in the same organization in another city
earlier in their career. This is all to say, every chair in this study was well
experienced in nonprofit organization governance.
The first surprise of the interview process was that only a small percentage
of executive directors served on other nonprofit organization boards of directors.
Many spoke to the fact that they often felt restricted in their ability to participate in
the leadership of other nonprofit organizations because they were unable to raise
funds for nonprofit organizations other than their own. Several executive
directors mentioned that they had all they could handle with their own nonprofit
organizations and a few spoke to their limited vision or contacts. Interestingly,
several seemed uncomfortable or even embarrassed that they were not more
involved in other nonprofit organizations. This was true for the executives of
each of the successful and less-successful nonprofit organizations.
Many executive directors made comments to this researcher like the
following, I need to save my voice (or presence) with funders for the sake of my

organization. It confuses them if I speak to the needs of other organizations. One
executive noted I ask my lead board people to speak to funders only about this
organization, so as to save their voice with funders for this organization.
Others made comments like I try to help other executives and nonprofits by
advising their leaders when I find someone with interest in funding their type of
organizations, as long as it doesnt take away from my funding. One executive
director commented, I have occasionally sent potential new, or retiring board
members to other executives where I thought the fit might be potentially
advantageous. And yet another spoke proudly about his pairing a very capable
financial staffer with a fellow executive director who needed that kind of help.
There were important differences in the answers given by the more-
successful and less-successful dyads. All but two of the chairs of the successful
nonprofit organizations spoke to the importance of the executive to their (the
chairs) success and only two of the chairs from the less-successful nonprofit
organizations made such comments. Some of the chairs of the less-successful
organizations were initially less comfortable that the role of the executive was most
critical to the success of the organization.
Secondly, the dyads from the successful nonprofit organizations spoke more
often to the primacy of the executive director to day-to-day operations than did
those from the less-successful nonprofit organizations. One chair put it best when

he said, we are each important members of the management team but my exec has
a life time of experience, contacts and success that we both lean on in times of
need. All of the chairs from the successful organizations spoke of that issue but
two of the chairs from the less-successful organizations spoke of their greater
criticality to the success of the organization. One of these chairs told the researcher
that he felt [he had to] lead this organization because of my legal responsibilities.
And finally, there was more conversation from the chairs of the successful
nonprofit organizations about the fact that their executives represented the long-
term strength of their nonprofit organizations. Again, all the chairs but two spoke
to the fact that the executives would likely still be leaders of their respective
organizations long after they had left.
It was in this first section of the interview process where this researcher
learned how much the chairs and executives cared for, respected and liked one
another. Two chairs made comments like someone needs to devote their lives to
this work and since he was man enough to do it, I am damn happy to be able to help
him in his work. Two executive directors made comments almost exactly like it
is an honor to work with my chair. He is a great man who could be helping anyone
and I am very pleased that he chooses to support my work. Finally, this researcher
was moved by the comment of one of the chairs when he said this man (an

executive director) could have been successful in any field of endeavor. It is great
to be able to help him do what I believe is Gods work.
Responses to Questions in
Interview Section Two
Section two of the interview process was designed to discuss the division of
leadership in nonprofit organizations, the central issue of this study. This section
included the same questions for the chairs and the executive directors. In response
to the most central question of the studyWho is the most important (most
critical to the success of the organization) leader of this nonprofit organization?
Thirty-one of the respondents selected the executive director role immediately.
Three of the respondentsall chairssaid that the role was shared equally between
the chair and the executive director. Yet, as these three chairs discussed their
situations further, they eventually noted that while they felt they should be, at least,
equal partners, in reality the executives were most critical to the success of the
A chair of one of the organizations deemed less-successful, offered the
comment I was elected to lead this organization and it is no small job to keep the
executive in his place. This same chair commented near the end of the interview
that, I eventually determined that the executive had so much more time to devote

to this effort that it is best for me to use my time to direct the director to ensure that
he is doing the boards mission.
The respondents told this researcher that the executive director is the most
important leader within their nonprofit organizations for the following reasons:
1) The executive directors have much more time to devote to the
organization (stated by 21 respondents)
2) The executive directors have much more professional experience and
many more contacts (14 respondents).
3) The executive directors stay in place for much longer, while the
volunteer chairs come and go (12 respondents).
4) The executive directors have much more at stake-e.g., their professional
reputations-than do the chairs (6 respondents).
Most of the respondents differentiated between policy setting activities and
daily management activities. The respondents felt that the chairs were ultimately
(legally) responsible for policy setting in the organization, although with
substantive input by the executive directors, while they felt that the executive
director was responsible for day-to-day operations. One chair put it best by noting,
As the leader of this board, I am responsible for the long-term vision of this
organization but it would be foolish to build a vision without the help of the
executive. I need her help with vision much more than she needs mine for the
day-to-day activities.

Two chairs from the successful nonprofit organizations group referred to
policy setting as that role of the board that sets boundaries for the organization
about the who, what, where, and when for carrying out the organizations
mission and allocating its resources. They saw the daily responsibility of the
executive director as to focus on the how of accomplishing the activities of the
nonprofit organization.
Virtually all of the respondents spoke to the belief that the laws, and the
expectations of other board members, demand that the chair hold the ultimate
responsibility to represent the best interests of society. Many of the respondents
made a point to say that in the area of organizational policy, the chair has the last
word, although the executive director has significant influence. The respondents
further noted that the executive directors influence is so great because s/he often
anticipates potential questions of policy before such questions arise and helps lead
the chair and the board to make decisions based on the executive directors
perception of the issues. Implicit in these comments about policy is the belief by
chairs that they are responsible for overall policy setting and for overseeing the
implementation of policies, but that the executive director is responsible for the
day-to-day implementation of policy. One chair noted, By the time the board
focuses on an issue of policy, the executive has already begun to lead the board

to the right answer through the activities of the committees and by framing the
perceptions of the board leaders in that area.
While four of the executive directors from the less-successful organizations
spoke to the fact that they occasionally had problems with their chairs or even other
board members who try to get involved with day-to-day issues, they reported that
they had mostly fought and won those battles. One executive director, from a
less-successful nonprofit organization, said that this day-to-day involvement on the
part of two board members had been such a problem that he had to go to the chair
and ask him to settle down a couple of board members. Interestingly, the
executive directors of all of the successful organizations either noted that this kind
of board interference was not a problem or stated that it doesnt come up because
their chairs and board members are too sophisticated or too professional to
worry about such things.
Most of the executive directors, and all of the executives from the successful
nonprofit organizations, dealt with the policy and daily management issue with
similar comments that their particular chairs would not involve themselves in daily
management. As noted, four of the executives from the less-successful nonprofit
organizations had problems with their chairs or at least some board members with
this issue.

Many of the executive directors stated that they had spent very little time
with non-chair board members, especially those who were not chairing committees.
Remarks to this affect were often followed by the statement, others on the staff get
to spend more time with board members. Most of the executive directors used a
tone of voice that lead this researcher to believe they wished to have more, rather
than less, contact with their board members. One executive director commented,
I make sure I visit with every member of my board at least once a quarter. He
further stated, that everyone of my board members is my boss and they are too
important to the discussion to allow them to be uneducated to my opinions.
In response to the questions about how the members of a dyad (chair and
executive director) handle disagreements among themselves, all said that they try to
work out the problem between themselves and, if necessary, the two would involve
other members of the board. One chair noted that during two potentially divisive
conversations he had with the executive in the past, he and the executive involved
a couple of level-headed members of the board in the discussion to help get to the
right answer. The chair seemed pleased that his opinion prevailed in one case and
the executives in the other.
Thirty-two of the respondents (16 dyads) assured this researcher that there
were relatively few disagreements and the disagreements that occurred were solved
professionally and without personal rancor. In the remaining dyad, the chair and

executive director were obviously in the midst of a personal disagreement. It
eventually became clear that this disagreement was affecting their ability to work
together. This researcher subsequently learned from each of these respondents that
within 30 days of the initial interview, the executive director announced his
separation from the organization because the disagreement could not be resolved.
Interestingly, each of these respondents made a point of contacting this researcher
and informing him of the change in status, in case it affected this research.
Most of the respondents had opinions about what should be, and what they
expected from the other member of the dyad; conversely, most respondents had
opinions about what was owed to their dyad partner. Collectively, the respondents
spoke of the need to give and receive respect; time commitment; dedication to the
organizations mission, business acumen/management skills, trust, advice and
counsel; a willingness to share responsibilities and to serve as a sounding board.
No one person recited the entire list, but all identified two or more of these items.
Dyad members did express a need for some things they didnt always get (or
get enough). Several chairs spoke to the need for the executives to be available to
meet when the chairs could be available. Since the chairs are very busy people,
they believed that the executive directors should adjust to their schedules.
Executive directors expressed their need for the chairs to provide community
representation and to be willing to lead the volunteers to raise money. Importantly,

in 16 of the dyads at least one of the respondents stated that their dyad partner was
doing what could be reasonably expected, and more. In these 16 dyads, the
individuals were generally happy with the partnership that had developed.
To review, the important findings from the second section of the dyad
interviews are as follows:
1. The respondents, especially the pairs from the successful nonprofit
organizations, felt that the chairs were ultimately responsible for policy
management in the organization, although with substantive input by their
executive directors. Every respondent felt the executive director was
responsible for day-to-day operations.
2. The chair has the last word in the area of organizational policy, although the
executive director has significant influence.
3. Only a minority of the executive directors (all from the less-successful
organizations) had occasional problems with their chairs or other board
members trying to get involved with day-to-day issues. Examples of day-to-day
issues were given as dealing with staff, finances, media, volunteers, and client
contact. None of the successful nonprofit organization executives reported any
problems and all four of the executives from the less-successful organizations
who spoke to the matter, noted that this was a minor problem that had been
solved to their satisfaction. With one of these executives, it did become a
matter of some importance.

Responses to Questions in
Interview Section Three
In the third section of the interview, the intent was to get the respondents
to expand on their views about the function of leadership within the organization
and the people involved therein.
The amount of time that the chairs and their executive directors of both
the successful and less-successful nonprofit organizations spent working together
was somewhat less than might be expected. Most of the respondents stated that
they spend about an hour per week together, but that included time they spend
in meetings with other board and committee members and with present and
prospective funders. However, the dyads did not include as time together, time
spent on the phone. Several dyads noted that when the organization was
experiencing problems, or preparing for an important meeting with outsiders,
the time together increased, often dramatically. One executive reported he
only bothered the chair when he needed her and tried not to take advantage of
her position.
All of the respondents from the successful organizations had significant
experience working in other organizations, and with other chairs and executive
directors. Most of these executive directors had worked with three or more chairs
in their careers; many of the chairs had been chairs in other organizations where

they had worked with executive directors. One executive director reported that he
had worked for 16 different chairs during his career. He told this researcher he had
worked with many different chairs and they all treated me the same way, and I
feel every one of them deserved my respect.
The dyads of the less-successful organizations reported less experience with
other organizations, but they felt that they had sufficient experience to do well in
their roles. Again, their comments lead this researcher to believe there was a great
deal of respect between the members of the dyads, especially after one executive
noted that he had great respect for all the chairs for whom I have worked, even the
one I did not care for personally.
All of the respondents were asked about other sources of leadership within
their organizations. Most spoke of other board members, especially committee
chairs. Some of the respondents, especially the executive directors, included senior
staff members as other potential leaders. Among the staff members to have
leadership responsibilities were those involved in fundraising, finance and program
activities. Also, in two of the organizations in question, a respondent mentioned
that ex-board members still provide leadership in selected areas of the
organizations agenda. Two of the executives made comments to the effect I am
constantly looking for the next chair so I make sure to mine the senior board
members for the potential to serve this role in the future.

Several of the 17 organizations were structured so as to have a third person
in the management team that this study has referred to as a dyad, thereby making it
a triad. This third person is referred to as curator, music director, chief
counselor, medical director or a similar name. In situations where there is a third
member of the leadership team, that person often provides significant leadership,
but almost exclusively in the area of program. These third leaders do not appear to
have policy or resource level leadership responsibilities. Their influence is
restricted to the music, counseling or educational mission of the nonprofit
organization. Further, the interviews did not suggest that there was any difference
between successful and less-successful organizations with regard to the existence of
these third leaders. There were two third leaders in the less-successful group and
two among the successful. In two of the four cases, the executive told this
researcher that that the program leader served as an important safety valve that
allowed me to focus on the management of the organization.
When asked the question about what had to happen for these organizations
to be successful, several of the dyad members, especially from the successful
nonprofit organizations, stated that they believes that the quality of the chair and the
executive director and the strength of their partnership made the difference
between a successful and less-successful nonprofit organization. With respect to
this statement, there was strong consensus among the interviewees, causing this