Public policy reforms of Colorado's workers' compensation system

Material Information

Public policy reforms of Colorado's workers' compensation system a study of privatization, increased competition and healthcare cost controls
Stacey, Rulon F
Publication Date:
Physical Description:
xii, 177 leaves : illustrations ; 28 cm


Subjects / Keywords:
Workers' compensation -- Colorado ( lcsh )
Medical care, Cost of -- Colorado ( lcsh )
Privatization -- Colorado ( lcsh )
Medical care, Cost of ( fast )
Privatization ( fast )
Workers' compensation ( fast )
Colorado ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 168-177).
Statement of Responsibility:
by Rulon F. Stacey.

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:
44075753 ( OCLC )
LD1190.P86 1999d .S73 ( lcc )

Full Text
Rulon F. Stacey
B.S., Brigham Young University, 1984
M.H.A., Brigham Young University, 1986
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 Rulon F. Stacey
AH rights reserved.

This thesis for the Doctor of Philosophy
degree by
Rulon F. Stacey
Peter Vellman

Stacey, Rulon F. (Ph.D., Public Administration)
Public Policy Reforms of Colorado's Workers' Compensation: A Study of
Privatization, Increased Competition and Healthcare Cost Controls
Thesis directed by Franklin James, Ph.D., Chair.
In the late 1980s, Colorado legislators realized that the excessive costs of
workers' compensation insurance were restraining job creation in Colorado. The
legislature implemented a comprehensive public strategy to cut costs. The strategy
included the privatization of the state-owned workers' compensation carrier,
Colorado Compensation Insurance Authority (CCIA), and the passage of Senate
Bill 91-218. Among other things, this bill fostered price competition among
workers' compensation insurance carriers, including CCIA; froze the medical fee
structure for workers' compensation insurers; and mandated the use of managed
care techniques.
This study evaluates the impacts of this strategy. Prior to the reforms, the
literature showed that CCIA was the most expensive insurer in the state, and self-
insured carriers were the most cost effective. Changes in the relative costs of
CCIA provide a useful indicator of the impacts of the privatization and related
reforms. Relationships between the managed care techniques used by insurers and
the costs of medical care to clients provide clues about the effects of the managed
care reforms. The medical price freeze provided an unusual opportunity to

measure the effects of managed care methods, because all carriers faced the same
fee structure.
Data for the study describe a sample of workers' compensation claims to
commercial carriers, self-insured plans, and CCIA. Claims were examined for
three specific diagnoses which were common and which involved fairly definitive
Results and Conclusions:
This study was unable to demonstrate that the managed care tools were
effective in this setting. The carriers used a variety of managed care techniques,
but there were few significant relationships between managed care techniques and
medical costs. The study was able to demonstrate that after privatization, CCIA
supplanted self-insured carriers as the least costly provider of medical care and
non-medical care to workers' compensation patients. The results suggest that the
combination of privatization of CCIA and enhancement of competition within the
market provided dramatic incentives and opportunities for greater efficiency by
This abstract accurately represents the contei
recommend its publication.
Franklin James, Ph.D.

I dedicate this thesis to my beautiful wife, Linda, for her unfaltering
support and patience, and to my daughters, Laura, Maria, Jennifer, and Catherine,
for loving me even having never known me as anything but a student.

My thanks go to all those who have helped me through this process.
Specific thanks go to my father for his support in research and the entire process;
to Yvonne for tireless proofreading; to Jane and Margo for wisdom and support
when they didn't need to; and, to Dr. James who demonstrated the patience of Job.

The Workers Compensation Prior to Reform ................ 4
Self-Funding Criteria.................................. 5
Commercial Insurance .................................. 6
Self-Funded Insurance ................................. 7
Colorado Compensation Insurance Authority .............10
Colorado Workers' Compensation Benefits................12
Workers' Compensation Cost Crisis.........................14
Cost of Disability Settlements ........................18
Setting Rates in Colorado: Pre-1991 ...................22
Employers Leaving Colorado.............................27
Colorado Reforms of the Workers' Compensation System .... 27
Costs and Self-Insured Carriers .......................28
Privatization of CCIA .................................36
Senate Bill 91-218.....................................38
Price Competition......................................38
Medical Cost Containment Changes in SB 91-218 .........41
Summary Public Policy Changes...........................43
Managing Care ............................................44
Managed Care Tools ....................................45
Criteria Summary.......................................54
Hypothesis HI .........................................57
Hypothesis H2 .........................................58
Hypothesis H3 .........................................60


2. LITERATURE REVIEW..............................................61
Costs in the Healthcare Industry ........................62
Impact of Provider Choice on Cost.......................62
Cost v. Quality.........................................67
Costs in the Workers' Compensation Industry ...............68
Factors Contributing to Cost............................68
Reasons for Increasing Workers' Compensation Costs .... 72
Incentive of the Workers' Compensation System...........78
System Abuses...........................................80
Cost Cutting Strategies of the Recent Past ...........82
Quality in the Workers' Compensation Industry..............90
Return to Work Programs.................................92
Managed Care and the Impact on Workers' Compensation .... 95
Managing Care...........................................97
The Delay in Workers' Compensation Managed Care.........97
Reforms in Workers' Compensation Managed Care.........101
Managed Care Strategies................................102
Other Workers' Compensation Managed Care Initiatives 107
Future of Workers' Compensation Managed Care ..........108
3. METHODOLO.GY..................................................110
Study Design..............................................Ill
Data Gathering ...........................................114
Collection Criteria.......................................114
Selection Criteria........................................116
Diagnosis/Injuries Sampled ...............................117
Data Gathering-Using Managed Care Tools ...............120
Analysis of Variance...................................121
Quality of Care........................................122
Regression Analysis....................................123
Anticipated Findings......................................125
4. FINDINGS......................................................128
Findings: Medical Costs ..................................128
Hypothesis Testing ....................................128
Analysis of a Tear of the ACL ............................129
Medical Costs-Summary..................................131

Findings: Non-Medical Costs.............................132
Total Costs.............................................136
Managed Care Utilization ...............................139
Multiple Regression Analysis and Results.............141
Summary of Results......................................146
5. CONCLUSION ................................................148
Review .................................................148
Analysis-Medical Costs (HI) ............................148
Commercial Carriers..................................156
Analysis-Non-Medical Costs (H2).........................156
Analysis-Managed Care (H3)..............................158
WORKS CITED..........................................168

1.1 Colorado Workers' Compensation Closed Claim Study

1.1 Per Capital Workers'Compensation Claims ...........................15
1.2 Premium Increases Between 1980 and 1990 .......................... 17
1.3 Comparison of Premiums in 1990 by Job Classification and State .... 19
1.4 Average Amount to Settle Permanent Total Claims....................19
1.5 Average Amount to Settle Permanent Partial Claims..................20
1.6 Average Medical Cost Permanent Partial Disability by State.........21
1.7 Permanent Partial Disability Averages by Colorado Carrier..........29
1.8 Average Medical Payments by Colorado Carrier ......................30
1.9 Medical Claims over $50,000 and $100,000 by Carrier................31
1.10 Non-Medical Claims over $50,000 and $100,000 ..................... 31
1.11 Percentage of Claims Closed Within Time Frame......................32
2.1 Health Coverage Differences .......................................80
2.2 Employers' Objections to Managed. Care............................100
2.3 Effectiveness of Managed Care Initiatives.........................108
3.1 Responses to 1995 Survey by Carrier...............................120
4.1 Hypothesis Conclusions ...........................................128
4.2 ANOVA of Medical Costs by Diagnoses and Carrier ..................129
4.3 ANOVA of Non-Medical Costs by Diagnosis and Carrier ..............133
4.4 ANOVA of Total Costs by Diagnosis and Carrier.....................137
4.5 Percent of Claims on Which Each Tool is Used......................140
4.6 Medical Regression Results by Diagnosis and Carrier...............142
5.1 Most Cost Effective Managed Care Providers .......................150

During the latter part of the 1980s, the amount paid by Colorado employers
for workers compensation insurable premiums, settlements, and other expenses
rose so quickly that Colorado legislators became increasingly concerned. The
average premiums paid by employers for workers compensation insurance in
Colorado had increased an average of 26 percent annually during the four years
from 1986 through 1989 (Lewis 1988). Not only was this increased cost to
employers impacting their ability to stay profitable but also many companies
seeking to expand their businesses increasingly chose nearby states instead of
Colorado (National Council 1989). Several of these companies had even publicly
announced their intended relocation to Colorado, only to later select another
western state rather than assume the burdensome workers' compensation expense
(McCardle, telephone interview, 9 Apr. 1997).1 There was no indication that this
trend would cease as the decade came to a close.
1 Information came from the telephone interview with Mike McCardle who
quoted from the 1988 and 1989 annual reports.

In January of 1991, the National Council on Compensation Insurance
(NCCI)2 requested a 38.3 percent increase in the premium that workers'
compensation insurance companies in Colorado could charge their customers
(employers) (Berry 1993). The NCCI contended that the high price the workers
compensation insurance carriers were having to pay for medical care and disability
settlements in Colorado left them no other choice. As evidence, NCCI cited the
fact that in 1990 Colorado ranked eighth in the nation in the average cost-per-case
for permanent partial disability, at $43,114. NCCI pointed out that this figure was
the highest of any state surrounding Colorado (including New Mexico, Arizona,
Nebraska, Utah, Oklahoma, and Kansas), and $21,948 higher than the regional
average (National Council 1990).
Colorados legislature commissioned several studies to analyze the
situation and provide an opportunity to correct the problem. The Colorado
legislature used information from these studies to address the workers'
compensation issue. The legislature addressed this issue by passing Senate Bill
91-218 and by authorizing the privatization of the state's workers' compensation
insurance carrier. These two pieces of legislation constituted the most dramatic
reform of the workers' compensation industry ever attempted in Colorado and
2 NCCI is a national agency that represents workers' compensation insurance
carriers in Colorado. The agencys purpose and background are explained in
greater detail below.

culminated more than a decade of frustration by the legislature, organized labor,
employers and insurance carriers alike.
By collecting data from workers' compensation carriers in the state, this
dissertation will pursue two objectives: 1) to examine the impacts of the reforms
mentioned above; and, 2) to use collected data to test the impact of various
managed care tools on the utilization of healthcare within the Colorado workers'
compensation system.
The reason for the two different objectives of this paper is as follows: The
initial purpose of this paper is to evaluate the public policy created by the state
legislature to address the dramatic increase in premiums for workers
compensation insurance in the late 1980s. Whereas fees paid to workers'
compensation providers were frozen, the only way to impact costs was to manage
the utilization of care. In collecting data for this evaluation, it will prove no
harder to ask a few more questions about which managed care tools are being used
by the workers compensation carriers in Colorado, and then to determine which
of those tools are best at controlling medical costs in the Colorado workers
compensation industry.
To begin this study, this chapter will analyze the problems surrounding the
Colorado workers' compensation system which lead to these public policy
reforms. The analysis in this chapter will proceed as follows:

1. Discussion of the workers' compensation industry in Colorado and
of the workers' compensation cost crisis of the 1980s.
2. Discussion of the public policy reforms created to address the
workers' compensation cost crisis.
3. Identification and discussion of managed care techniques used to
address healthcare utilization in the Colorado workers' compensation
4. Presentation of conclusions from information presented in the
introduction and present three hypotheses proposed for this paper.
The Workers' Compensation System in Colorado
Prior to Reform
Although each state is required to provide a workers' compensation
system, the U.S. Government grants states latitude as to how best to accomplish
that task (U.S. Government Accounting April 1996, 1). Colorado has elected to
require that each of its employers purchase such insurance from one of three types
of carriers: commercial insurance carriers; self-funded carriers or the Colorado
Compensation Insurance Authority.
Workers' compensation insurance purchased from a commercial carrier
closely resembles coverage private citizens buy to insure their homes, cars, or
lives. Normally each individual employer considers several commercial insurance
carriers, determines the prices and benefits that each offers, then selects the one
best suited.

A company who selects the self-insurance option must meet stringent
criteria created by statute.
Self- Funding Criteria
Employer must have 300 employees regularly employed in
Employer must have been in business for at least five years.
Employer must show a solid financial position with strong financial
Employer must demonstrate a successful safety and loss control
Employer must maintain competent claims adjusting.
Self-insurance permits are reviewed no less often than every year.
Source: Colorado Division of Workers' Compensation, report to the
legislature, 30 June 1994, p. 13.
A company that is too small to insure itself may elect to gain the
advantages of self-insurance by joining a cooperative with other similar-sized
businesses. The group then forms a trust to provide self-insurance for all the
member companies. This alternative is utilized often in Colorado particularly by
school districts, hospitals, etc.
Some companies, for one reason or another, cannot or do not wish to use
either a self-funded plan or a commercial insurance plan. In the recent past, these

companies have been either too small to qualify for self-funding or for some
reason (such as excessive utilization of workers compensation benefits) unable to
engage a commercial carrier. For these companies, the state of Colorado has
created a separate insurance company called the Colorado Compensation Insurance
Authority (CCIA). Technically, CCIA is a commercial carrier that provides
insurance for a premium. CCIA by law must accept any business which desires to
be insured by CCIA although the premiums charged can be high. In Colorado, the
major users of CCIA are construction companies (Berry 1993, 13).
Commercial Insurance
For commercial workers' compensation insurance, a company (employer)
buys the amount of coverage that provides for its workers' actuary-computed,
projected claims. This system has both advantages and disadvantages for the
The main advantage of commercial insurance is that the employers liability
is limited to the insurance premium and deductible. Unless an employer utilizing a
commercial plan has a track record bad enough to warrant cancellation of its
policy, the worst consequence the employer can expect from large claims is an
increase in its premium. This economic consistency provides the employer with a

predictable budget and ease of accounting. If the insurance premium is
reasonable, this consistency and predictability are, of course, very desirable.
The disadvantage of commercial insurance is that the premiums will
always be there and rarely decrease. There is certainly no chance of eliminating
the premium or keeping the surplus, as with the self-funded plans discussed below.
One of the major criticisms of commercial insurance coverage is that the
employer has little financial incentive to improve working conditions. When a
commercially insured company suffers an increase in workers' compensation
injuries, those increased costs are immediately born by the insurance carrier not
the employer. Because of this "buffering" effect the employer who uses a
commercial carrier is relatively unaffected (financially) by either short or long-
term increases in worker injuries. The insurance premiums charged by
commercial carriers reflect the risk record of the employer, and employers with a
poor track record or an unsafe working environment do pay a penalty eventually.
Self-Funded Insurance
When a company creates self-funded workers' compensation insurance,
that company opens a separate bank account into which it deposits monthly
premiums to pay workers injury claims. Advantages of this program are that the
company maintains control over its assets and, if the company is effective in

diminishing claims, it can keep the savings. The financial incentive for a self-
insured employer is to foster work place safety. When self-insured carriers
effectively improve safety, the costs of their workers' compensation program are
kept down, premiums decrease, and profits increase.
To establish a fund of this nature, employers must meet the state of
Colorados specific guidelines on how much money is required to open the
account, how much must be deposited monthly, and how much the account must
contain to justify discontinuing deposits. As a result there are advantages and
disadvantages to having an account of this nature.
Under Colorado law, the workers' compensation account can grow to such
a level that no further deposits are required. For a company to reach this point in
its self-funded plan, the financial controls and safety program must be optimum,
both of which are great advantages for employers and employees alike.
One disadvantage of a self-funded plan becomes evident when the outflow
from such an account grows too fast. Should this happen, the company is left to
pay the entire amount of all claims which could literally bankrupt a company if
proper precautions are not enacted. To protect against this happening, all self-
insured plans carry stop-loss insurance coverage. This coverage is designed to pay
for any claims that go above a predetermined limit. This coverage reduces the
chance of a self-insured carrier going bankrupt because of one large claim, but

does not protect the self-insured company from the financial losses incurred prior
to the stop-loss coverage engaging (Tillinghast 1993, 34).
Another disadvantage to self-insurance plans has to do with a twist on
financial incentives. Organized labor has long argued that the reason why self-
insured plans keep the costs down is not so much that the self-insured plans are
effective at controlling the over-utilization of services, but that they simply do not
provide necessary services to their employees. This argument has been a topic of
discussion between labor and management for some time.
One of the benefits of a self-insured plan, as pointed out by organized
labor, is its ability to keep the costs of paying workers' compensation benefits to a
minimum. For example, because most self-insured companies tend by their very
nature to be large, these companies are in a better position to offer restricted duty
assignments to injured workers, or to give other full-time jobs to workers no
longer able to continue in their respective positions. These self-funded employers
can better keep their costs down by having many different options.
Nevertheless, it is important to point out that any such advantages available
to self-funded employers during the late 1980s and early 1990s were still available
to them in the mid 1990s. Nothing in any public policy reform that will be studied
in this paper altered the ability of self-funded employers to utilize their workers in
this way.

Colorado Compensation Insurance Authority
Colorado Compensation Insurance Authority (CCIA) is a "catch-all" that
the state of Colorado created for employers who cannot or chose not to, for one
reason or another, find insurance elsewhere. CCIA is required by law to accept
any company that asks to be covered, regardless of that company's past record
with workers' compensation utilization. CCIA does not, however, have to offer
the coverage at a reduced price. For a company that cannot compensate injured
workers through either a self-funded or commercial plan because it has over-
utilized workers' compensation payouts (called the company's experience
modification factor)3, CCIA is not required to offer low premiums. CCIA can be
extremely expensive because of its risk adjusted premiums. For companies with
good experience modification factors, CCIA is more competitive in cost.
However, the resulting high average rates have created a poor image for CCIA
with many companies in Colorado. It has been alleged that CCIA staffing was
inadequate when it was a public company and that CCIA representatives were too
3 Each employer has a different rate of injury. Some auto mechanic shops
have injuries 10 percent of the time, and some have injuries 5 percent of the time,
even though they do basically the same work. The rate at which employees of a
given employer have injuries is called the experience modification factor (EMF). If
an employer has injuries of its work force at a rate equal to others in the same job
classification, then the EMF is 1.00. If an employer has more injuries, then the
EMF is greater than 1, and conversely, if the employer has fewer injuries, then the
EMF is less than 1.

busy to adequately defend claims or keep the costs down.4 Additionally, as the
cost crisis in the Colorado workers' compensation market (which will be reviewed
in detail below) began to become more profound, more companies began to seek
coverage from CC1A, which was required to accept their business. This only
served to increase the workload on an already overburdened staff, and to create
further problems to rapidly respond to claims and meet an ever more challenging
budget. The result was more and larger-than-needed disability settlements and
slower service which served to increase the company's experience modification
Nevertheless, simply because CCIA may have higher premiums than its
counterparts that does not imply any inefficiency. For example, because of its
nature, CCIA has, in the past, been the carrier of choice for construction
companies (Berry 1993, 13). Since construction is inherently a more risky
business than many other occupations, the average cost of premiums for CCIA
could make these appear higher than other insurers.
CCIA was a public agency of the State of Colorado until 1992. As such, it
was subject to Colorado civil service policies regarding pay, hiring and firing, and
promotion. It was also subject to normal executive and legislative budget
4 Data came from interviews by author with companies who are insuring the
CCIA and with other companies in Colorado. Companies included in the survey
were from the healthcare industry, manufacturing and construction.

processes. As will be discussed below, it is alleged that being in the state system
constrained the capacity of the agency to adapt to the rapidly changing market for
insurance during the late 1980s. Privatization in 1992 loosened the constraints at
the agency.
Colorado Workers Compensation Benefits
The workers' compensation system in Colorado is structured so that every
worker in the state, regardless of insurance carrier, is guaranteed by law similar
benefits for a work-related injury. Theoretically, this means that all employees
injured on-the-job will receive the same care whether their employers are covered
by a commercial insurer, CCIA, or a self-funded plan.5
When an injured worker seeks care under the Colorado workers
compensation system, the healthcare received differs in several ways from
treatment for a non-work related injury. A synopsis of workers' compensation
5 Although it is true that the law guarantees the same coverage regardless of
the plan of the employer, it is important to remember that self-funded plans have
historically managed to pay less on average for workers' compensation claims than
have other carriers. The reason is widely thought to be because the self-funded
plans have the greatest motivation to control costs and therefore more effective
utilization review procedures as described below.

An injured worker is not responsible for any part of the payment for
medical services. There is no deductible for the worker, no co-
payment, no associated payment of any kind.
Any time away from work to treat the injury is paid to the worker.
Additionally, any associated recuperation time is also paid time off.
The employer may require that an injured worker work at another
job that does not aggravate the injury. However, if that job pays
less than the workers' original job, then the difference must be paid
to the employee.
An injured worker may receive a monetary settlement for damages
incurred. A worker who is unable to continue in any job is said to
be permanently totally disabled. A worker who can work, but in a
lesser job, is said to be permanently partially disabled. Other
categories include temporary total and temporary partial. Since
each category differs in severity, settlements differ as well.6
Employees who are injured at home may select their own
physicians. In Colorado, an employer may designate which
physician its employees will see following a work related injury.
While this is a fairly recent development in Colorado, the results
seem to be significant. In Colorado, should an employer mandate
that an injured worker see a specific physician, then the costs of that
case decrease by nearly half (Tillinghast 1993, 22). Additionally,
chiropractors have been eliminated from the choice alternative by
the State because the amount that was charged to treat an injury by
chiropractors in general was considered excessive (Tillinghast 1993,
6 A major concern with the workers' compensation system in Colorado was
the lack of correlation between dollar amounts paid and severity of injury. For
example, it was not unusual to see a lower settlement amount awarded by an
administrative law judge for a severed limb than the settlement awarded by a
different judge for a broken finger. This trend concerned both the labor side and
the management side, since one could argue that the attorney involved played a
greater role in determining the settlement than the injury.

The Workers Compensation Cost Crisis
Between 1980 and 1990, the average cost for workers' compensation
insurance premiums charged employers increased 228 percent in Colorado. To
put this figure in perspective, during the same period, payments for workers'
compensation injuries by insurers throughout the rest of the country increased a
far more modest 92 percent (Schmulowitz 1995, 51). Premiums in Colorado
increased more than in any other surrounding state (see Table 1.2).
Figure 1.1 Colorado Workers' Compensation Closed Claim Study
Colorado Workers Compensation Closed Claim Study
Claim Count Distribution by Accident Year
350 I-----------------------------------------------------------------
1977 1979 1981 1983 1985 1987 1989 1991
Accident Year
Commercial gZS Self-Insured
One reason for Colorados increase in workers' compensation costs was
simply the larger number of claims. Figure 1.1 shows the number of claims by

type of carrier for each accident year from 1977 to 1990 (Tillinghast 1993, 26).
The dramatic increase in claims is better realized when taken in consideration with
the population of Colorado. This is summarized in the Table 1.1 below.
Table 1.1 Per Capita Workers' Compensation Claims
Year Colorado* Population Workers Compensation Claims No. of Colorado Residents per W.C. Claim
1980 2,890,000 15,000 193
1986 3,237,457 35,000 93
1990 3,294,000 350,000 9
* Statistical Abstract of the United States, 1998.
(http: //www. census .gov/statab/ff eq/98s0026. txt)
** Tillinghast Report to General Assembly, 1993.
This increase in the number of claims was caused by legislation unique to
Colorado which authorized workers' access to care which was not available in
most surrounding states (McCardle, telephone interview, 9 Apr. 1997). Some
examples of this unique legislation follow:
1. During the period in question, Colorado did not put a time limit on
the amount of time an injured employee could receive benefits. New
Mexico, Utah and Oklahoma all limited the number of weeks a person
could receive benefits to either 100 or 500 weeks.
2. During the period in question, Colorado did not have any type of
schedule to use by which to determine the amount to be paid for the loss of

a body part. Arizona, Kansas, Nebraska, New Mexico, Oklahoma and
Utah all had a defined schedule to use which allowed them to calculate the
cost of the loss of a specific body part.
3. During this period, Colorado law did not require vocational
rehabilitation. Every state surrounding Colorado did require vocational
rehabilitation (Schlackman 1995).
The result of these and other legislative differences did not impact the
Colorado workers compensation industry until the late 1980s. The Division of
Workers' Compensation surmises that intense financial pressures on employees,
the legal system, and the general economy (Colorado was in a recession during the
late 80s) of the time combined with the restrictive nature of the laws in
surrounding states led to the rapid increase in the number of claims in Colorado
beginning in 1986. The result, as can be seen in Figure 1.1, was a drastic increase
in the number of claims submitted. This is to say nothing of the amount of those
claims which was also increasing.
Of particular note is the fact that in the early 1990s the number of claims
being submitted began to decrease. Interestingly, the number of claims decreased
after the implementation of the public policy reforms being evaluated in this paper.
A portion of the Findings Chapter of this paper will discuss the correlation of
the drop of claims and the implementation of the public policies discussed.
Not only did the number of claims being filed increase in the late 1980s,
but also the amount of those claims was growing faster in Colorado than in other

states. A summary of the rate of growth in workers' compensation premiums in
Colorado and surrounding states follows:
Table 1.2 Total Workers' Compensation Insurance Premium Increases Between
1980 and 1990, by State
Premium Increase
New Mexico
228 percent
123 percent
119 percent
66 percent
61 percent
28 percent
4.8 percent
Source: National Commission on Compensation Insurance Annual Report, 1991.
It is important to note that a comparison such as this is risky because
statutory regulations differ in various states. These premium increases left
Colorado at or near the top of the region for premium rates of nearly every job
classification. However, increases in premiums do not of themselves indicate a
problem. It is quite possible that Colorado carriers before this period were
disproportionately low in the premiums that they were charging. A comparison of
costs of similar claims shows that, in fact, Colorado was more expensive in
providing workers compensation care.

Table 1.3 presents several job classifications and their associated premiums
in Colorado and each of the states surrounding Colorado in effect in 1990. The
rates listed below are for every $100 of payroll in individual job classifications.
For example, if a company in Colorado had field crop employees to whom they
paid $100 per year, the cost of the insurance would be $11.29. The cost for those
same employees in New Mexico would be $10.39. As will be shown, rates in
Colorado by the end of the 1980s were significantly higher in most categories than
in nearby states. Colorado led the region in premiums much of the time, and was
never lower than fourth in any one category.
Cost of Disability Settlements
Costs of disability settlements also boosted premiums in Colorado relative
to other places. Colorado led the region in 1991 in the costs of settlements of
permanent total disability claims (National Council 1992). Table 1.4 presents a
comparison of the settlement for disability claims in the region.

Table 1.3 Comparison of Premiums in 1990 by Job Classification and by State.
Each classification highest rate in bold. All figures presented in dollars per $100
of payroll
Field Meat Computer Airline Retail
CroDs Packing Manufacture Carpentry Grounds-Crew Grocery Clerical
Colorado 11.297 31.68 1.10 20.95 4.73 8.76 0.49
New Mexico 10.39 15.81 2.04 18.99 4.16 8.82 0.67
Arizona 9.95 11.15 1.01 25.02 6.78 3.27 0.50
Oklahoma 6.97 11.33 2.28 11.93 1.20 3.79 0.41
Kansas 6.29 8.11 0.94 9.03 2.17 3.04 0.28
Nebraska 6.05 4.13 2.12 8.84 2.0 2.37 0.25
Utah 5.31 8.27 0.64 8.97 4.88 4.23 0.29
Source: National Council on Compensation Insurance Annual Report, 1992.
Figures reprint data from 1991.
Table 1.4 1991 Average Amount Needed to Settle A Permanent Total Disability
Claim, by State. (Figures include medical and non-medical costs of providing
care, plus any settlement amount)
Colorado $224,863
Arizona $168,109
Nebraska $ 67,216
Oklahoma $63,369
Kansas $54,049
New Mexico $48,561
Utah $47,614
Regional Avg. $96,254
Source: National Council on Compensation Insurance Annual Statistical Bulletin,
7 Note that premium amounts correlate directly with the risk of injury for
each specific job.

As Table 1.5 shows, Colorado also led the region in disability settlements
for permanent partial disability claims:
Table 1.5 1991 Average Amount Needed to Settle A Permanent Partial Disability
Claim, by State. (Figures include medical and non-medical costs of providing
care, plus any settlement amount.)
Colorado $43,114
New Mexico $28,269
Arizona $20,018
Nebraska $18,046
Utah $14,764
Oklahoma $12,407
Kansas $11,542
Regional Average $21,166
Source: National Council on Compensation Insurance Annual Statistical Bulletin,
Table 1.5 includes both medical and non-medical costs associated with the
disability claim. However, when considering just medical costs, Colorado also
ranked close to the top. The summary of those numbers is presented in Table 1.6.
One could argue with this information that the cost crisis in the workers'
compensation industry in Colorado resulted from several factors, including the
cost of claims, the cost of disability settlements, the rapidly increasing cost of
claims, and, finally, the increasing number of claims being filed. Thus, correcting
the escalating costs of Colorados workers compensation program was a multi-

faceted problem that required public policy changes designed to address each of
the separate issues.
Table 1.6 1991 Average Medical Cost per Permanent Partial
Disability Case, by State
New Mexico $1,991
Colorado $1,372
Arizona $1,350
Oklahoma $1,166
Utah $939
Nebraska $781
Kansas $717
Source: NCCI Annual Statistical Bulletin, 1992.
Even with this analysis of the workers' compensation cost crisis, there
remains the issue of whether or not there is a correlation between the economy of
Colorado and the number and/or the amount of workers' compensation claims.
Specifically, since Colorado was experiencing an economic recession during the
latter part of the 1980s, did that play any role into whether or not employees
would file for a workers compensation claim?
It is clear after a review of the literature on the subject that no one has
drawn any such relationship between economic conditions and the number of
workers' compensation claims. Since such a study is beyond the scope of this
paper, it was assumed for the purposes of this research that any such correlation

was negligible. An in-depth study regarding any such correlation, however, is
certainly worth the effort.
Setting Rates in Colorado: Pre-1991
Prior to 1991, the Colorado State Commissioner of Insurance established
the rates used to calculate insurance premiums charged by insurance companies in
Colorado.8 How those rates were set, however, was quite a complex process.
Even though Colorados insurance commissioner actually set the rates for
workers compensation insurance, the commissioner did so only after receiving
and considering (as mandated by statute) (Strauss 1998) data from the National
Council on Compensation Insurance (NCCI). NCCI supplied this information to
the insurance commissioner because the state legislators selected them to represent
the insurance carriers for workers' compensation matters.
NCCI, a national, nonprofit corporation based in Tampa, Florida, offers its
services (i.e., rate review, access to information, negotiation services, etc.) to
states that need representation for their insurance carriers. Although a state or the
many carriers in a state may select whoever they wish to represent them, most
8 Information collected during a phone call to the Colorado State Insurance
Commissioners Office, May 1998.

states choose NCCI to provide this service. NCCI has served Colorado in this
capacity since 1972 (Strauss, telephone interview, 22 June 1998).
Before 1991, NCCI collected policy and claim information from all carriers
doing business in Colorado. Using this information, NCCI reviewed specific
trends for each employer and each job classification. NCCI also reviewed
financial data from employers including amounts paid for premiums, losses
associated with workers' compensation, financial productivity, losses for various
job classifications, etc. Similar information from competing carriers in Colorado
was examined to set a "base" rate for each job classification.9
From descriptions of specific jobs and the associated risks, NCCI
calculated each base rate and added what it considered to be appropriate expenses
for doing business in the workers' compensation insurance industry in Colorado.
Expenses that NCCI routinely added were profit (typically zero to 2 percent
(Strauss 1998)), taxes, advertising, etc.
Subsequently, NCCI was prepared to make a "fully-loaded" rate request to
the office of the insurance commissioner. Although the insurance commissioner
was required to consider the information provided by NCCI, the commissioner
was not mandated to follow that recommendation and, in practice, rarely did.
9 A job classification describes each occupation so an insurance rate may be
assigned to it (e.g., farm worker, auto mechanic, school teacher, etc.). Each job
classification received a rate assignment according to the related risk of injury.

Rather, the insurance commissioner gave the information provided by NCCI to an
actuarial firm hired by the commissioner's office to review the numbers in great
detail. Generally, the rate finally authorized was somewhat lower than the fully-
loaded rate requested by NCCI. Once the insurance commissioner accepted the
rate, it became known as the approved "manual rate."
With the manual rate approved by the insurance commissioner, the private
sector then took over to a small degree. Each workers' compensation insurance
carrier in the state used the manual rate as a base to create the premium that it
offered to employers throughout the state. It is important to clarify that the
manual rate approved by the insurance commissioner differed from the premium
which was eventually offered to employers throughout the state.
Insurance companies in Colorado used the manual rate to calculate
premiums by using several sets of information. The manual rate was multiplied by
the number of units of exposure for a given employer. The number of units of
exposure was calculated by taking the number of employees and combining that
with the experience modification factor of the employer.10 The number of units of
exposure was then multiplied by the manual rate to determine the premium.
Seldom did the private market allow for actually setting the premium
calculated by multiplying the manual rate and units of exposure. Individual
10 See foomote 3 above for definition.

insurance carriers had their own marketing tools for creating the final price quote
to employers. For example, some of the marketing programs included variable
deductibles, premium discounts, discounts for being a member of a national or
local professional association, schedule rating (marketing discounts or surcharges),
discounts for having effective loss control programs in place, and discounts for
having your own medical provider (McCardle, telephone interview, 9 Apr. 1997).
The evidence suggests that this system limited effective price competition
among commercial insurance carriers, and that the state used its premium setting
powers to attempt to hold employer premiums below the costs of benefits provided
by the industry. Perhaps the best evidence of this came in a review of insurance
carrier losses. These losses (or gains) are calculated by the insurance
commissioners office in each state and is a ratio of premiums paid to claims paid.
When an insurance company has a loss ratio of .98 that is the same as saying that
it has a profit margin of 2 percent. Anything greater than one signifies losses to
the insurance carrier.
In 1990, this loss ratio in Colorado for non-CCIA business was 1.25, by
far the highest in the region, and the third highest nationally (behind California
and Oregon). The ratio for CCIA was 1.375 (National Council 1991, 2). These
numbers indicated two facts:

1. First, Commercial and self-insured carriers were losing money and
had been for several years prior to 1990. They had little choice but to
discontinue writing insurance in the state or simply go out of business.
2. Second, CCIA was incurring huge losses. CCIA was required to
insure any company in the state that asked to be insured and because its
ratio was so unfavorable, CCIA had no alternative to losing money.
During the latter part of the last decade, no fewer than 50 insurance
companies elected to discontinue writing workers compensation insurance in
Colorado (Firm, telephone interview, 9 April 1997). Those that stayed rarely
accepted new business. When a carrier chose to cease its services in the state,
employers left without insurance had to find some way of complying with the law
mandating coverage. Since few companies were accepting new business, CCIA
had the burden of servicing many more employers. And, according to Gary Pon,
Director of CCIA, less than 5 percent of the new business that CCIA received at
that time was from employers with an experience modification factor less than
To address this situation, the state legislature commissioned several
studies.11 As the state legislature gathered information from these studies, the
general feeling was that this problem (increasing costs in the workers'
11 These commissioned studies included John Lewis & Associates, 1987:
John Burger, Inc., 1990; John Berry and Associates, 1991; and Tillinghast, Inc.,

compensation system) would have to be resolved immediately.12 The legislatures
concerns centered on two areas: 1) companies (employers) leaving Colorado or
refusing to relocate to Colorado; and, 2) workers' compensation insurance carriers
deciding to discontinue insuring workers in the state.
Employers Leaving Colorado
During the early part of the 1980s, the economy of Colorado grew about 3
percent per year (Colo. Dept, of Labor 1999, 17). That lasted until 1987 when the
economy virtually stopped growing. There were no new jobs or other indicators
of economic growth during the last three years of the 1980s or in 1990.
Colorado Reforms of the Workers' Compensation System
The state legislatures response to the rapid increase of workers'
compensation premiums throughout the 80s focused on two different areas starting
in 1987. In 1987, the legislature authorized CCIA to begin a five-year transition
to becoming private, effective January 1, 1992. As mentioned above, the
legislature commissioned several studies to help it understand how to better
address the rest of the industry's problems. In view of the information received as
12 The feelings of the Colorado legislature are presented after interviews with
several leaders of the group during the time. Most notably, the review came from
Ken Chlouber, House Majority Whip at the time of these events.

a result of these actions, the state legislature determined that the workers'
compensation system in Colorado needed an overhaul. An impetus was NCCIs
request for a rate increase of 28.3 percent in January 1989. Given the large
increases of the previous years, the insurance commissioner did not approve that
rate. Instead, the commissioner granted only a 14.3 percent increase effective
June 1, 1990. NCCI accepted that decision rather than engage in a long and
drawn out court battle; however, later in 1990, NCCI announced a proposed rate
increase of 38.3 percent for 1991 (Berry 1993, 17).
Much of the focus on reform was to increase the efficiency of insurance
provided by CCIA and commercial insurance carriers. The evidence was
compelling and strongly suggested that CCIA benefits were much more costly than
were those of the commercial insurance carriers, and that commercial insurance
carriers were more expensive than were self-funded carriers.
Costs and Self-Insured Carriers
Having previously outlined the workers' compensation market and how it
was structured prior to 1991, we can discuss the perception of many in the
industry at the time when self-insured carriers kept their costs lower than either of
the other two carriers. Prior to 1991, it was the general belief in the state that
self-insured carriers were simply better at keeping the costs to a minimum

(Tillinghast 1993, 33). Evidence which was used to create this perception is
presented below:
Table 1.7 Closed Claim13 Permanent-Partial Disability Average Costs by
Colorado Workers Compensation Carrier (1990).
Source: Colorado Workers' Compensation Closed Claim Study.
Tillinghast/Towers-Perrin, Denver, Colorado. March 1993, p. 33.
One of the most frequently used measures of relative cost control is a
summary of permanent partial disability (PPD). Such a summary is presented in
Table 1.7. This information shows that in 1990 self-insured carriers were able to
keep the costs for PPD closed claims lower than either of the other two carriers.
This same trend was observed for average payments for medical expenses
only, as listed below:14
13 A closed claim is a workers' compensation claim for one patient in which
the patient has reached maximum medical improvement (MMI), and is therefore not
expected to get any better. A closed claim represents all medical and non-medical
costs associated with the patient's injury through MMI.
14 The average medical payment information shown is calculated for all
workers' compensation injuries. Since a large portion of those injuries can be
1990 Average

Table 1.8 Average Medical Payments for all Closed Claims by Colorado
Workers' Compensation Carrier15 (1990)
1990 Average
Source: Colorado Workers' Compensation Closed Claim Study.
Tillinghast/Towers-Perrin, Denver, Colorado. March 1993, p 33.
Other information also indicates self-insured carriers were relatively
efficient in containing medical costs. For the years leading up to the public policy
reforms of 1991, self-insured programs generally had fewer claims involving very
large medical costs. The claims information is as follows:
treated with little medical intervention, the averages are relatively small. In this
study we will break out the charges by diagnosis to give a better understanding of
what can be done to control costs.
15 The average medical payment for each carrier is calculated as the sum of
all medical expenses paid by the carrier for a year, divided by the number of claims

Table 1.9 Percentage of Colorado Workers' Compensation Industry Closed
Claims with Medical Costs Over $50,000 and $100,000 prior to 1992, by Carrier
Carrier Claims > $50,000 Claims > $100,000
Self-Insured 17.8% 2.8%
Commercial 30.2% 7.5%
CCIA 37.7% 9.3%
All Groups 30.1% 7.1%
Source: Tillinghast Report to General Assembly, 1993.
This trend is the same with respect to the total costs of claims (medical and
Table 1.10 Percentage of Colorado Workers' Compensation Industry Closed
Claims with Total Costs Over $50,000 and $100,000 prior to 1992, by Carrier
Carrier Claims > $50.000 Claims > $100.000
Source: Tillinghast Report to General Assembly, 1993.
Self-Insured 52.8%
Commercial 61.5%
CCIA 71.4%
All Groups 64.9%
Self-insured carriers were also better at closing their claims quickly. The
following table lists the average number of claims closed by each carrier within
one and two years.

Table 1.11 Percentage of Colorado Workers' Compensation Industry Claims
Closed within One Year and Two Years Prior to 1992, by Carrier
Claims Closed Within:
Source: Tillinghast Report to The General Assembly, 1993.
Qualitative data collected from individuals familiar with the Colorado
system also supports the conclusion that self-funded plans were more efficient than
either CCIA or commercial carriers.
For example, in a conversation conducted in preparation for this study,
Ernest L. Dunn, Colorado Department of Labor and Employment, Division of
Workers' Compensation, asserted that he knew of no one who did not believe that
self-insured carriers were better at managing the care of their patients, and he
offered a few reasons why this would have been the case in the 80s and early 90s:
1. Self-funded plans are managed by the companies that pay the
premiums. This financial arrangement produced strong incentives to
reduce costs.
2. The staffs of self-funded plans were typically larger than those of
the other plans. This results in an increased administrative expense, but it
also serves to make the staffs significantly more efficient at pursuing claims
and increasing overall efficiency.
1 Year
2 Years

3. Finally, the system of education for personnel in self-funded plans
was better than that in other insurance plans. This improved education
included employee safety training (and routine inspections), managerial
training on back-to-work programs, and job retraining for employees who
are injured but still capable of light duty.
Additionally, the 1993 Tillinghast study which was commissioned to
review the workers' compensation industry before and after enactment of Senate
Bill 91-218 suggested two main reasons why self-insureds provide less expensive
healthcare than other compensation plans:
1. Self-insureds have a higher incidence of using their own medical
2. Self-insureds have a lower incidence of attorney involvement in
workers' compensation claims.
Other reports surrounding the workers' compensation environment
suggested a similar conclusion. For example, the 1993 John Berry (sec. 1) study
noted that its self-insured plan, because of the companys financial interest in
controlling costs, spent more time and effort in bringing cases to a successful
conclusion than either of its counterparts. The result, the study claimed, was that
self-insured plans were simply "better at keeping their costs down." The studies
completed by Grant Thornton (1991, 26) and Tillinghast (1993) came to similar

Like the conclusions of these independent studies, the Division of Workers'
Compensation agrees that self-insured carriers have historically been better at
controlling costs. In the divisions 1994 annual report, it said:
"Self-insurance is widely viewed within the workers'
compensation community as providing better
management for workers' compensation claims.
Some have argued that self-insurers' closer ties to
their employees and to their claims costs provide an
incentive that is lacking in other methods of claims
processing. Although there is no complete study on
this question,16 some recent data has been reported .
. in New Mexico [that would substantiate this
claim.] "(Colo. Div. of Worker's Comp. 1994, 14)
Using the information gained by the various commissioned studies (Lewis
1988, 18; Berry 1993, 54), the legislatures General Assembly passed Senate Bill
91-218 in their 1991 session, effective July 1, 1991. Because workers and
employers had different incentives with respect to enacting this legislation,
negotiations between organized labor and business representatives delayed its
passage. The negotiators broke off talks because they disagreed about the
definition of permanent disability; thereafter, organized labor opposed every
aspect of Senate Bill 91-218, even though the bill eventually became law. During
the debates, Eldon Cooper, the Colorado President of the AFL-CIO, walked out of
16 An important issue here is that one goal of this paper is to provide a
complete study of this question by evaluating the abilities of all carriers to manage
their compensation claims.

the senate sub-committee addressing the issue. Later, Mr. Cooper uninvited
Governor Romer, who had been scheduled to speak at their annual meeting,
because he signed the law.
This hotly contested issue did not end with enactment of the new law
(Miller, B., telephone interview, 22 March 1997). Business interests and labor
organizers continued to do the battle until the labor unions filed suit to invalidate
Senate Bill 91-218 (Berry 1993, 13). However, after more than two years in the
courts, the objections of organized labor were summarily dismissed by a district
court judge in Denver in July of 1993 (Perrault 1993).
To increase the relative efficiency of CCIA, the strategy of the legislature
was to:
1. Privatize CCIA. In January of 1992, CCIA's organizational
structure was changed to alter it from a department of the state to a private
2. Competition. The public policy changes created a better
opportunity for the introduction of competitive forces into the states
workers' compensation system, especially competition among commercial
3. Freeze on Medical Costs. In an effort to let the public policy
initiatives be fully reviewed and implemented, a freeze on all workers'
compensation payments to providers was enacted.
These public policy reforms are outlined in detail below.

Privatization of CCIA
On July 1, 1987, CCIA ceased to be a division of the Department of Labor
and Employment of the State of Colorado. Instead, CCIA became an Authority of
the State, a privatized status that authorized it to have its own seven-member board
of directors and to institute its own personnel policies, rules and regulations. As a
division of the state, CCIA employees had been subject to the state's personnel
policies and its rules and regulations; additionally, as state employees, CCIA
employees were eligible for certified status17. Although all employees of CCIA
had a "grace" period of five years to remain state employees, all new employees
were hired in an "at-will" status and were subject to the new CCIA personnel
polices which were implemented on January 1, 1988.
Because of this five-year "grace" period, CCIA did not achieve complete
autonomy until July 1, 1992 (as it turned out, this date is significant because it is
only one year removed from the implementation of Senate Bill 91-218 on July 1,
1991). Initially, only Gary Pon, the Executive Director of CCIA, changed status
to become an employee of CCIA as the board required him to do. No other CCIA
17 Certified status is the right of state employees to keep their jobs. Once
this status has been achieved, after three to five years of employment, the employee
is considered to have a "right" to a job, and in a lay-off or other change, the
certified worker can claim the job of a lesser employee, even if the certified worker
has not been trained for that position.

employee changed status for more than six months, and a substantial portion of
state employees still worked there well into 1992.
One of the objectives of the privatization of CCIA was to allow the agency
the opportunity to be more responsive to the forces of the private market. In
allowing this change, the legislature had hoped to create an entirely new private
As a new entity with the autonomy to create a budget and hire employees,
CCIA moved out of its state-owned building in 1990, further distancing itself from
its former status. Although the transformation would not be completed for another
two years, CCIA18 soon became more effective at changing its structure and
planning its future (Pon, telephone interview, 28 June 1997).
The significance of the timing cannot be discounted since this privatization
of CCIA occurred (effective 1992) in the height of the workers' compensation cost
crisis. Additional references to and details of CCIAs privatization will appear as
this study progresses.
18 The information relative to the change of CCIA came from the author's
interview with Gary Pon, Executive Director, CCIA. Pon was the executive
director before the change to an authority and was responsible for all subsequent
changes in CCIA after the change to being an authority.

Senate Bill 91-218
The legislature clearly tried to increase competition between carriers in the
creation of Senate Bill 91-218. Given that the legislature had already authorized
the privatization of CCIA, it was clear that its desire was to allow greater free
market forces to have an impact on the workers compensation industry in
Colorado. The bill also moved to impose stronger controls on medical care costs
in the workers' compensation system.
Given the significance of the changes, each is described in some detail
below. The administrative changes made in Senate Bill 91-218 will be listed first.
Price Competition19
Lost-Cost Rating. Senate Bill 91-218 mandated changes in how rates in
Colorado were set. Instead of NCCI submitting a "fully-loaded" fee schedule20 to
the insurance commissioner, Senate Bill 91-218 required it to submit a "lost-cost"
rating. The lost-cost rating creates a manual rate in exactly the opposite fashion
from the fully-loaded concept. In this scenario, NCCI is required to determine
how much each job classification costs each year and submit that before including
19 Information in this section is taken from Colorado Senate Bill 91-218, and
the NCCI Annual Statistical Bulletins from 1991 and 1992.
20 See discussion on setting rates in the workers' compensation market on
pages 10 and 11 above.

expenses. The insurance commissioner then reviews the lost-cost rating submitted
by NCCI. To this lost-cost rating, each carrier adds its own expenses and
overhead to yield an individual manual rate, as in this sample equation:
Lost Cost + Expenses = Manual Rate
Where: Lost Cost is submitted by NCCI
Expenses are calculated by each carrier
Manual rate is unique to each carrier
The percentage by which the lost cost is increased to yield the manual rate
is called the lost-cost multiplier (LCM). Each company determines its own LCM.
As discussed before in Senate Bill 91-218, each carrier had the same manual rate
and any discounts, training programs, etc., provided the only competitive force.
Given new found freedom, insurance carriers could attract employers by offering
lower prices in the form of greater discounts, outcomes, and other price sensitive
Incentive for Safer Work Place. Senate Bill 91-218 provided incentives
that encouraged increased competition between carriers. The legislature, however,
realizing that this incentive may not adequately protect the worker, also included
financial incentives to provide for safety in the work place. The most notable of
these incentives was the Risk Certification Plan. This plan offered discounts to
employers for implementing state certified risk-reduction programs. After
creating a risk reduction plan, Senate Bill 91-218 mandated that employers

received either a 10 percent discount with a desirable experience modification
factor (EMF)21 or a 5 percent discount without.
After enactment of Senate Bill 91-218, only businesses with more than
$4,000 in yearly workers' compensation insurance premiums were required to
have an EMF. Smaller companies that were certified still received a 10 percent
discount, and larger companies received 5 percent.
Deductible. Further providing an incentive for worker safety, Senate Bill
91-218 mandated that the first $5,000 of each claim not count in the calculation of
EMF. Thus, if a company had several claims, none of which passed the
deductible of $5,000, the company's EMF would list no injuries and enjoy an
associated decrease in premiums. A company whose claim totaled $25,000 would
have only $20,000 credited against the calculation of its EMF. Colorado was the
only state in the nation with such a law at the time of its passage (National Council
Medical Provider. Finally, the law mandated a discount of 2.5 percent for
employers who designated their own medical providers and used them according
to published guidelines.
21 See Definition, page 10 above.

Medical Cost Containment Changes in Senate Bill 91-218
With the passage of Senate Bill 91-218 in 1991, payments to healthcare
providers for medical services given to workers' compensation patients were
frozen at then-current rates. In the early part of this decade, medical expenses
associated with workers' compensation claims accounted for 43 percent of the
overall costs of injuries in the workers compensation industry in Colorado
(National Council 1992). This places Colorado second in this region in amounts
spent on the medical portion of workers' compensation claims (behind New
Mexico) (Berry 1993, 25).
The state-commissioned Lewis (1988) study contended that the main reason
for the increased medical cost of the workers' compensation industry other than
public policy problems was the overall rise in the cost of medical supplies. Lewis
argued that during the 1980s the percentage of the totals for medical expenses
related to workers' compensation claims in the state more than doubled.
As a result of Lewis' findings, Senate Bill-91-218 specifically listed as one
of its goals the reduction of medical expenditures associated with the workers
compensation system of Colorado (Berry 1993, 1). Of primary importance was
permanent total disability.
Reductions in Disability Settlements. Previous sections stated that
Colorado's cost of a permanent total disability claim was significantly higher than

in adjacent states. Not mentioned, however, was that in Colorado 8.5 percent of
all injuries resulted in permanent total disability claims; meaning, of course, that
the patient could never work again. For comparison, the national average is 1.89
percent of injuries resulting in permanent total disability (National Council 1998).
Thus, an injured worker in Colorado was more than 400 percent more likely to be
declared permanently disabled than in other states. The definition of permanent
total disability changed in Senate Bill 91-218. Prior to Senate Bill 91-218, the
definition of permanent disability centered on whether or not an injured worker
could continue to work in his or her current occupation. Even if that worker could
be retrained to earn a living in another occupation, the law considered the worker
permanently disabled. After Senate Bill 91-218, only workers who were incapable
of working in any job were considered as permanently totally disabled.
Permanent-Partial Disability. Ninety-two percent of the permanent partial
disability settlements paid before 1991 were the maximum allowed by law,
$32,650 (Colo. Div. of Worker's Comp. 1991, 27). This was true if a worker lost
the right leg or the tip of the right little finger. Not surprisingly, the legislature
believed that this cap under compensated the more severely injured workers and
overcompensated the less severely injured.
Senate Bill 91-218 standardized disability ratings using the American
Medical Association Guide. 3rd edition, and established a ceiling. Every

physician in the state is now required to follow these guidelines. Ratings provided
from this guide translate into benefits (after multiplying by time off for the
worker). A worker who is less than 25 percent disabled receives a maximum
benefit of $65,000; more than a 25 percent disability can yield a maximum benefit
of $125,000.
Summary Public Policy Changes
By focusing workers' compensation reform on increasing the influence of
private market forces, the state legislature had signaled significant changes in
public policy. These changes provide an excellent opportunity to evaluate the
success of these policies by measuring the impact on costs of the workers'
compensation industry.
Unfortunately, Senate Bill 91-218 did not address what impact the
utilization of managed care techniques might have on the industry. Managed care
is so central to healthcare costs on a national basis that it seemed prudent to
include such a review in this paper. A review of managed care and its
implications to this study follows.

Managing Care
With the added emphasis of competition, workers' compensation carriers in
the state needed to begin to pay more attention to the way in which they managed
the utilization of care given to patients using their insurance. Most of the many
programs which were available to the carriers for this purpose found their
beginnings in managed-care environments (i.e., HMOs, PPOs, etc.), yet the
processes themselves have proven very advantageous to other groups in the
healthcare industry who were interested in minimizing their expenditures on
Other states, most notably Hawaii, have found such workers' compensation
programs effective. Some states have even considered mandating a program
requiring the utilization of managed care tools to keep the medical costs of
workers' compensation to a minimum (Kertesz 1995, 40). In Hawaii, the
workers' compensation fees have been frozen at 10 percent above the Medicare
Index22 while attempting to create a managed care plan that enforces adequate
controls on the rapidly accelerating workers' compensation costs (Kertesz 1995,
22 The Medicare Index is a price structure that the Health Care Finance
Administration (HCFA) sets annually of the rates Medicare pays for any given
medical procedure.

Colorado's Senate Bill 91-218 did not mandate utilization of managed care
tools to restrict the costs of the workers' compensation system. Workers'
compensation insurance carriers were left to their own problem-solving devices.
The insurance carriers reviewed the utilization of workers' compensation patients,
by and large, with the same tools used throughout the healthcare industry.
Managed Care Tools
Insurance companies and others use managed care tools to decrease the cost
of providing care, including any program or other device that exacts pressure on
the care giver to lower costs. These same tools used to alter the utilization of care
were available to the Colorado workers' compensation carriers. A review of
managed care tools most widely used identified the following six items.23
Monetary Incentives
Preauthorization Criteria
Concurrent Review
Prospective Review
Retrospective Review
Care Pathways
23 See Chapter 2 for literature review on this topic.

Monetary Incentives and Capitation. The physicians compensation was
formerly tied to the amount of care provided. In a capitated market, the physician
receives a fixed amount of money to cover a certain number of patients. Each
month the physician is given a check for the care of the group of patients under his
or her control. Whatever care is provided is then paid for by the physician (Cave
1994). Obviously, this scenario offers a financial incentive for the physician to
provide as little care as possible, since supplying any care causes the physician to
lose income.24 A patients stay in the hospital or a referral to another physician
becomes a very expensive proposition for the physician since the money to pay for
these services must come out of the physician's fixed payment (Koles 1995).
Other financial incentives have to do with bonuses or penalties that are paid
directly to the care provider as a result of the efficient practices. For example,
Neil Schlackman, MD, has argued that when organizing any type of managed care
plan or any device with which to manage care, performance bonuses for the
physicians and other care givers are essential. In his experience, Dr. Schlackman
(1995) notes that such incentives often have such a dramatic effect that no other
24 The best discussion on the pressures on the providers of healthcare in a
capitated market come from a variety of publications by the Advisory Board
Company, a Washington, D.C. based healthcare think-tank. See Strategies and
Vision of American's Leading Insurers an October 1995 publication by Lynn West
of the Advisory Board Company. The Advisory Board is located at 600 New'
Hampshire Avenue, Washington, D.C. 20037.

effort is warranted. Additionally, such bonuses are becoming the norm in leading
managed care states, such as Minnesota and California (Torchia 1994, 32).
With respect to the Colorado workers' compensation program, such
financial incentives did not work as they might because the rates paid for medical
services were frozen. In that context, some would argue that the incentive is to
provide more care since the return on just a few cases is so minimal. Clearly,
incentives to provide more care are not in the long-term best interest of the
workers' compensation system in Colorado (Lipsitz 1993, 82).
Because the implementation of a price freeze in Colorado has distorted the
ability to analyze monetary incentives as a tool for managing care that criteria
(monetary incentives) were not used in this study to evaluate how Colorados
companies manage healthcare utilization for injured workers. However, because
financial incentives are so important to every other aspect of healthcare utilization
review, that subject is included here.
Preauthorization Criteria. Preauthorization criteria were designed
specifically to reduce the utilization of inpatient care and to focus the plan of
treatment toward the less expensive outpatient environment (Miller, L. 1994, 44).
This is accomplished by requiring that a patient using this system receive
permission from the insurance company before treatment. Unlike the physicians
monetary incentives listed above, the financial penalty for misusing the

preauthorization criteria is assessed against patients in the form of a higher co-pay
for unauthorized care.
This means that a patient who is treated in a designated setting (i.e.,
inpatient, emergency room, surgery, etc.) is responsible for contacting their
insurance company and obtaining permission BEFORE the service is rendered.
Failure to do so costs the patient all or a part of the covered expense (Spitzer
An insurance company using this method might create a set of criteria that
must be met for the company to pay for an inpatient stay. Without this permission
in advance of the admission or procedure, reimbursement for the stay is denied.
Although the patient is responsible for obtaining this permission, the physician's
office personnel often calls the insurance office.
Lee N. Newcomer, M.D., Vice President at United Healthcare
Corporation, one of the nation's largest managed care companies, has said that,
upon learning (from preauthorization review) what is planned for a patient,
intervention becomes valuable. Dr. Newcomer (1994, 4) states, "Suppose we're
concerned because a physician is about to do abdominal gallbladder surgery
instead of a laparoscopic cholecystectomy. We'd fax him a copy of the practice
guideline we use and accompanying scientific references."

Preauthorization thus can be utilized as an effective tool with which to alter
or prohibit the care that is given a patient. With proper implementation, the tool is
very effective.
Concurrent Review. Whereas pre-authorization review precedes treatment,
concurrent review occurs during treatment in either an inpatient or an outpatient
setting. This review consists of frequent contact with the physician or other
provider to determine the appropriateness, intensity, frequency, or other factors of
the care being provided. The reviewer in addition often asks the provider about
future plans, then holds the provider to those plans to the greatest degree possible
(Zablocki 1994, 4).
Concurrent review is often considered by physicians and other providers as
the most unpleasant of the various managed care tools. The reason is that the
physician may be subjected to frequent calls inquiring about the patient's status,
the services provided to the patient in the recent past, and what plans are being
made by the physician for the termination of care. Physicians find the persistent
calls annoying which often as not are meant to accomplish that very response
(Capitation 1994).
Dr. Newcomer, who was referenced above, argued that concurrent review
can often change the financial outcome of the care of a patient without altering the
clinical outcome. Specifically for workers' compensation patients, Dr. Newcomer

suggested that much medicine (today) is practiced more out of habit than out of
medical judgment. As evidence, he remarked on his companys success in
reducing the number of MRI studies for workers compensation patients simply by
asking the physician how the results of the test would alter the care. In 40 percent
of these cases, the answer was that the results would not impact the course of care
in the least, and upon realizing this, the physician did not order the MRI
(Newcomer 1994, 5).
Perhaps the most aggravating aspect to physicians is that such calls are
rarely, if ever, made by other physicians. Calls are sometimes made by technicians
with little or minimal medical education. Calls are also made frequently by
registered or licensed practical nurses with course work in managed care. Either
way the physicians are left frustrated at the mandated necessity of justifying their
care to a person with less than half their training. But if this means that the
physician terminates the process of treating the patient more quickly, then the
concurrent review criteria have worked well for insurance carriers in spite of the
physician's frustration.
Prospective Review. Prospective review is typified by an attempt to
manage care by reviewing claims after services are provided to the patient but
before authorizing payment. To provide for this type of rapid review, specific
systems have been developed.

Dr. Donald C. Harrington and a few colleagues developed criteria during
the mid 1950s to evaluate the appropriateness of care provided in a managed care
environment. They called this set of criteria "Patterns of Treatment." These
initial criteria defined levels of excessive use of ambulatory services in terms of
appropriateness and frequency. The theory behind this plan was that upon quick
review of the chart a physician's practice patterns could be gauged and quickly
compared to the patterns of treatment criteria in an effort to identify areas of
excess utilization (Miller, L. 1994, 45) .
The patterns of treatment criteria have established over the years that 90
percent of treatments given to patients are appropriate. For the remaining 10
percent, three main factors raise the costs of care above the norm:
1. Inaccurate or inappropriate coding errors.
2. Patient illnesses that are outside the norm.
3. Provider practice patterns that are less efficient than established
patterns and, therefore, constitute excess utilization.
As one might expect, of the 10 percent of patients who fall outside the
Patterns of Treatment criteria, 80 percent are a result of inappropriate utilization
on the part of the provider (number 3 above) (Miller, L. 1994, 45).
For prospective review to be as beneficial as possible, as many patients as
possible must be surveyed in a short period of time. Great emphasis is presently

being placed on developing computer software that enhances the ability of
managed care reviewers to identify and correct problems as quickly as possible
and on as large a scale as possible (Coile 1994, 6-7).
Retrospective Review. Because of the magnitude of claims presented to
any insurance company, all cannot be addressed by prospective review. A large
number must, of necessity, be reviewed after the claim is paid. Information is
gathered in an effort to form profiles on provider utilization characteristics and
identify areas of over utilization. The process is known as retrospective review.
After analysis, the results can be returned to physicians leading to changes in
practice by aberrant providers and support for efficient, high-quality providers
(Miller, L. 1994, 45).
With the data gathered by employers and insurance companies by
retrospective review, there is literally no limit to the restrictions, incentives and
requirements that can be devised for providers of healthcare. In Minnesota, one of
the most advanced managed care markets in the nation, companies are utilizing
retrospective data to ensure that cost variation among all medical clinics in a
specific group are within an acceptable range. Restrictions are imposed on the
offending clinics when variations are excessive (Torchia 1994, 32).
Blue Cross and Blue Shield of Nebraska used such information to evaluate
two competing psychiatric clinics. The outcome determined that both clinics saw

patients with similar problems. One clinic saw a lower average number of patients
but had higher costs. The other clinic saw more patients but spent far less per
patient. As a result, changes in the high cost clinic were mandated before more
referrals to that clinic were authorized (Spitzer 1994, 82).
The data from retrospective review can significantly alter healthcare.
Although more time is needed to collect the data in a retrospective review because
of the volume of information available, it is often more valuable than the
prospective review.
Care Pathways. Care paths are lists of optimal performance indicators
associated with a particular disease or injury. For example, a care path associated
with a tear of the ACL (which is studied in this paper) might list the appropriate
stay in surgery, in recovery, and in the hospital post-surgically. It may also
indicate appropriate times for the patient to achieve "milestones" in the recovery
process. Physicians who consistently fall out of the accepted path are watched
carefully and helped when necessary (Cave 1994, 6).
A good example of the care pathway concept is used at St. Francis Hospital
in Blue Island, Illinois, where this system has been used for some time to ensure
that cardiac surgery patients are treated as efficiently as possible while
simultaneously minimizing patients quality concerns. The hospital's care path for
coronary artery bypass graft surgery specifies what should happen to a patient and

when. Periodically a patient "falls off the path and requires special care, but the
general result has been to decrease the average length of a stay by almost half the
national average, as published by the Healthcare Finance Administration
(HCFA).25 This ability to provide cost-effective care while still maintaining high
quality outcomes has saved millions of dollars for insurance companies and is,
therefore, very advantageous in limiting the overall costs in any healthcare setting.
Criteria Summary
These criteria previously discussed are the major processes by which
insurance companies manage the care of patients, without actually providing the
care themselves. With the exception of monetary incentives, the remaining criteria
apply to the Colorado workers' compensation environment. Fifteen self-insured
carriers, 15 commercially insured carriers, all located in Colorado, and a
representative from CCIA were randomly contacted to determine whether any of
these tools have been used. The result documented that while not all companies
use each criterion, each of those criteria listed above are used in varying degrees
25 Healthcare Finance Administration (HCFA) data published Fall 1995 in
the Federal Register listed national CABG LOS data. St. Francis data is available
from this author.

in Colorado.26 Only those criteria listed above (with the exception of monetary
incentives) will be considered in this study to evaluate the types of managed care
criteria used by Colorado workers' compensation insurance companies.
This chapter describes the rise in costs for the workers' compensation
industry in Colorado. Considerable evidence indicates that the Colorado market in
the late 1980s and early 1990s experienced cost increases greater than other
surrounding states. Colorado state legislators were collectively convinced that the
industry was in crisis, and they felt compelled to enact the far-reaching changes
embodied in the public policy reforms discussed above. As a result of these
changes, workers' compensation carriers in the state of Colorado were allowed
more latitude in implementing competitive approaches to gain business.
The insurance carriers became responsible for justifying their expenses and
were encouraged through the private market to lower the cost of their services,
including better management of the patients costs, both medical and non-medical.
The purpose of this paper is to create a study that will adequately review the effect
of public policy changes enacted in these reforms.
26 Full capitation is not yet legal in the Colorado workers' compensation
industry, although the 1996 legislature debated the issue and authorized a smdy on
the topic using the experience of Hawaii as a rough guide.

The only data available for this study appeared after the implementation of
these reforms leaving no choice but to complete a simple post, post-comparison.
Comparison of the relative costs of the carriers involved should nevertheless be
possible to determine whether or not the public policy in question worked. The
relative costs of the carriers should be different if these reforms were effective.
Commercially insured carriers and CCIA should have lower costs if price
competition worked. CCIA should have decreased costs in comparison to the
others if privatization worked. Either way, the information generated will be
Not only will such a study prove valuable to the state legislature in
determining the success of its policy, but also the workers' compensation industry
will gain critical information relative to how successful it has been in
implementing cost containment strategies. Because profits are directly tied to
costs, the results of this study will also be of great value to the workers'
compensation industry.
There is reason to believe that before enactment of these reforms the self-
insured carriers in Colorado were better at managing the costs of patients than
other types of carriers. Although no formal study has yet evaluated this claim, the
Colorado Division of Workers' Compensation considered this information so
important that its recent annual report called for just such a study.

The outcome of the reforms outlined above has been investigated and for
purposes of this study, three hypotheses are offered:
Hypothesis 1 (HI)
H1: Self-insured carriers are better than other carriers in keeping the cost
of medical care down, and all other carriers are better at limiting these
costs than is CCIA.
H1 is phrased in this way because it is the general consensus of workers'
compensation industry experts that prior to 1991 self-insured carriers were the
most efficient insurers. Likewise, these experts agree that during this time frame
CCIA was not as good as others at keeping specific costs down. Therefore, H1
addresses this issue for the first time.
Much debate before and during the public policy debate focused on the
control of medical and non-medical expenses alike. The legislature believed that
the privatization of CCIA would allow greater latitude in competing in the
workers' compensation market.
Some businesses expressed concern, however, that Senate Bill 91-218 was
too focused on medical costs and neglected non-medical costs. Employers worried
that the incentives created with Senate Bill 91-218 would provide short-term gains
for the insurance carrier by emphasizing reduction of medical costs over non-

medical costs. The reasoning was that medical costs are paid directly by the
insurance company, but most non-medical costs including some disability
settlements are paid jointly by the insurance carrier and the employer. For
example, while an injured worker stays home and is not working, the employer
and the insurance company each pay a portion of the employees wages. The same
is true for light duty employees who are working at other than their permanent
jobs. If an insurance carrier were to emphasize limiting medical costs while
ignoring high non-medical expenses, the employer would eventually have to look
elsewhere for coverage. The insurance company could possibly gain from the
situation in a short-term arrangement.
Hypothesis 2 (H2)
H2: Self-insured companies are better than other carriers in limiting
non-medical costs. All other carriers are better at limiting non-medical
costs than is CCIA.
The results from this study will assess the success of the reforms in
managing healthcare. Additionally, H2 answers the Division of Workers'
Compensation's call for specific review of each insurance carriers performance.
The structure of the public policy changes in question offers an opportunity
to review the effects of managed care on the workers compensation insurance

market, an assessment that has never been presented before. These insurance
carriers made no concerted effort until 1991 to manage patient care in Colorado.
No acute need existed since premiums continued to increase (until 1990).
After 1991, the changes gave great incentive to insurance carriers to better
manage healthcare, and gave the academic world an excellent opportunity to
measure the programs success. Insurers now had to create their own manual rate
before calculating premiums for their customers. The net effect of this change was
pressure to reduce costs, the largest of which was the payment of medical and non-
medical workers' compensation claims.
Simultaneous to this change in incentives prices for medical care were
frozen. This meant that the only way to reduce costs was to improve the
management of healthcare. We now have a one-time opportunity to evaluate
which managed care tools are the most effective at producing desired outcomes,
and which insurance carriers are the best at implementing those tools and
managing a workers' compensation market that heretofore had no utilization
review at all.

Hypothesis 3 CH3)
H3: Implementation of one or more of the managed care strategies
herein discussed by a workers' compensation carrier in Colorado will
reduce the overall cost of related healthcare services.
This information will prove useful in determining which is the preferred
strategy or set of strategies in a market that has never managed healthcare
Each hypothesis addresses a subject of great interest to public policy
makers in Colorado, subjects never before investigated. The results clarify the
process followed, the outcomes that succeeded, and the alterations needed in
future policy-making decisions.

This literature review provides an in-depth background for understanding
the healthcare industry in general and related workers' compensation issues in
particular. Subjects of special emphasis are the provisions of healthcare to
workers' compensation patients, and the costs and quality of that care.
Specifically, this chapter cites literature pertaining to the following areas:
1. Costs in the healthcare industry. The costs associated with
providing healthcare often stems from the motives behind utilization and supply of
medical services. How these motivations impact the overall healthcare picture
including workers' compensation care is this sections theme.
2. Costs in the workers' compensation industry. This portion of the
review details how costs associated with healthcare in general affect the workers
compensation industry in particular. An additional emphasis is placed on how
costs associated with the workers' compensation industry have developed
differently than in the healthcare industry.
3. Quality in the workers' compensation industry. One of the
challenges of the workers compensation industry is to maintain quality healthcare

while attempting to control costs effectively. The literature presented here outlines
what measures are being taken to ensure that quality healthcare is still the product
of an industry that directs so much effort to cost reduction.
4. Managed care and the impact on workers' compensation. Often the
result of cost reduction in the healthcare industry has been the creation of managed
care companies and plans. The workers' compensation industry is no different.
During recent years there have been multiple attempts to manage and implement
the care of workers compensation patients in Colorado and throughout the nation.
This section summarizes the impact of such efforts on the Colorado market.
Costs in The Healthcare Industry
Impact of Provider Choice on Cost
One must understand the importance of the choices made by patients and
providers alike when evaluating the costs of healthcare. The literature shows that
many alterations in costs associated with the healthcare industry are in as much a
result of patient and provider decision patterns regarding such areas as supply
costs, litigation, personnel, equipment issues, etc.
For example, early in the debate over utilization of healthcare resources,
James Buchanan, a well known economist from the University of Virginia, argued

that healthcare is a relatively elastic good, and that people who are required to pay
for their own healthcare would use less of it (White 1989, 552).
Buchanan gave an example to explain further-the British "experiment" with
the National Health Service. Buchanan offers three possible reasons why the
National Health Service is a failure:
1. The undesirable features observed in the British system are
due to mistakes in administering the health services mistakes
which "wiser" people could have avoided.
2. The difficulties are due to the very structure of the
instimtions through which health services are provided, and which
could, therefore, be removed only after major reforms in this
institutional structure.
3. The difficulties are inherent in the nature of the health
services themselves (Buchanan, J. 1972, 28).
Buchanan contended that the second of these possible reasons was in fact
the correct choice. His argument was that no structure which attempted to provide
for free a relatively elastic good could ever survive (1972, 28).
Buchanan's reasoning concerning impacts on healthcare utilization is
certainly not unique. In fact, the literature would suggest that utilization of
healthcare resources is impacted by many other constraints in addition to its elastic
Several uncontrolled studies in 1988 showed that a prepaid, group medical
practice hospitalized its patients from 15 to 40 percent less than patients who were

covered by traditional fee-for-service insurance programs (Luft 1978, 1336-1343).
A similar study in 1994, this time controlled, found that people participating in a
managed care program (in this case an HMO) were 40 percent less likely to enter
the hospital as patients than those in a fee-for-service arrangement (Willard 1984,
Other studies support this outcome (Newhouse et al. 1981, 1501-1507);
however, the only quality evaluated were variances in healthcare usage between a
fee-for-service setting and a prepaid group practice. One might wonder, therefore
if other areas within the healthcare environment are subject to such wide
All these studies interestingly conclude that although more healthcare is
provided to those who use the system more nothing factual indicates that the
greater quantity offers better quality of care. Patients in the survey groups were
equally healthy and had nearly identical outcomes regardless of the amount of
healthcare that was used. The quality issue, however, remains central and is
addressed further at the end of this section.
The previous paragraphs deal mainly with patients who require healthcare
services in an inpatient setting. Outpatient care is equally important and
constitutes a substantial portion of total healthcare expenditure. Knowledge of

these expenditures helps us understand the related usage patterns prevalent in the
workers' compensation setting soon to be examined.
As with the inpatient setting, outpatients who are responsible for financing
their own medical care use less care than do those who do not self-pay. This is
true in hospital outpatient arenas, other immediate care facilities, and, also,
medical offices (Eisen et al. 1990, 6).
The most telling facts about healthcare usage are that different amounts can
be spent in treating the same numbers of people as clearly exemplified by the
differential state spending in medicaid. In just one instance, New York recently
spent $9 billion during a fiscal year for medicaid coverage; in comparison, Texas
spent $1.9 billion to cover approximately the same number of people for the same
time period27.
This difference in spending from state to state28 has been attributed to
multiple variables. Holahan and Cohen found in a separate study that higher
income states tend to spend more on their healthcare decisions than lower income
states (1986, 2). Medicaid allows broad flexibility in program design from state to
1 See U.S. Department of Health and Human Services, Healthcare Financing
Administration, Division of Medicaid Statistics, HCFA 2028 Data Tables. FY
1988. See Table 1 for total recipients and Table 9 for expenditures.
2 During the same period mentioned above, California spent $5.2 billion to
cover a much larger population. See HCFA 2028 Data Tables. FY1988. Table 9.

state. States with more resources may choose to provide more services, pay
higher provider fees, allow more consumer choice, have broader eligibility criteria
or a combination of all of these. Subsequently, Robert J. Buchanan identified
political factors as important in the expenditure of healthcare dollars. Buchanan's
research showed that in states more politically liberal costs to provide for the
medical indigent were greater than in states with a somewhat more conservative
political climate (1987, 18). Saundra K. Schneider (1991, 765-763), in what has
come to be known as a landmark study on this subject, documented that as
participation in medicaid programs grows (Holahan and Cohen 1986, 321-346)
the dollars spent increases by a factor greater than one. This is to suggest that as
medicaid enrollment increases and as the government's ability to monitor each
enrollee decreases individual expenditures for the program increase.
Buchanan, Cappelleri and Ohsfeldt also discovered that the cost of a
states medicaid program increased whenever it was managed by a local
government (Schneider 1991, 759-760; U.S. Federal Register 1989, 49, 358).
They identified three reasons for this increase. The third reason identified is that
"since locally administered programs are still funded by federal and state revenues,
local governments do not directly bear the financial burden of their administrative
decisions" ((Buchanan, Cappelleri and Ohsfeldt 1991, 69).

In an unrelated study, Kieffer, Alexander and Mor validated somewhat the
notion that people making the same healthcare decision choose differently based on
their financial objectives, not necessarily their physical ailment (Sloan 1992, 321-
346). Simply stated, people buy more healthcare with someone else's money.
Kieffer, Alexander and Mor found in a study of prenatal care use among
several ethnic populations that womens choices fluctuated significantly based on
the financial burden personally incurred. These researchers also learned that
women in the military used by far the most resources but, in contrast to opinions
voiced elsewhere, had the best outcomes during delivery (1992, 655-656).
Clearly, multiple factors are involved in making healthcare utilization
decisions, and to generalize that all decisions are financially driven is an
overstatement. Companies who are conscientious about the amount of money
spent on healthcare for their employees can find numerous ways to limit their
losses (Millman and Roberts 1987, 2).
Cost vs. Quality
As mentioned earlier in this chapter, regardless of the type of insurance,
the rationale behind utilization of the healthcare system or the amount of care
used, there is no appreciable difference in the quality of care provided to patients.

The literature also indicates that when the trade off between cost and
quality is translated into a care vs. a no care situation, there is a difference.
Holahan and Cohen (1986), in their discussion of medicaid, conclude that quality
may not vary according to how much care individual insured groups use, but the
quality does not exist when care is withheld.
It is true that people use more healthcare services, an elastic good, when
someone else pays all or a portion of the bill. The quality of care or outcome for
those persons fortunate enough to be insured remains optimum. Their situation
remains stable, i.e., unchanged. The quality of life and healthcare outcomes
change enormously for persons who are not insured or have no access to care
when they are provided the chance to spend resources not formerly available for
healthcare services. When the differences in outcomes are discussed, it is
important to distinguish between those groups using different amounts of
healthcare: those who have access to the required resources and can choose to use
healthcare and those who do not have access to resources or healthcare.
Costs in The Workers' Compensation Industry
Factors Contributing to Cost
The health benefit cost for employees was the single greatest concern to the
nation's business owners in a recent survey (Goldblatt 1996, 31). Just behind

health benefits in general, the next biggest worry to business owners was workers'
compensation cost and the strain it put on businesses in general (Goldblatt 1996,
This concern results from the fact that workers' compensation costs
(including premiums) have increased at a much more rapid pace than the costs of
the general healthcare industry. The problem with increasing workers'
compensation costs has nearly crippled businesses and industries in some
instances. Perhaps the best example was the recent near collapse of the metal
work industry, a collapse attributed in part to the rapid increase of workers'
compensation rates (Frahm 1996, 1). Between 1970 and 1990, medical expenses29
under workers' compensation insurance increased 1,335 percentalmost double
the 795 percent increases in the nation's medical expenses overall (Insurance, NY
Times 16 April 1993, 12).
Medical expenses were not the only problem cited as the reason for these
dramatic increases in workers' compensation costs over the past decade. Other
problems those corporate risk managers have suggested as significant contributors
are as follows:
3 Medical expenses are only one part of the workers' compensation cost
structure, albeit a major part. These expenses will be examined in greater detail

Employees receiving excess medical treatments resulting in excess
Litigation over compensation for claims.
Inconsistent state regulations.
Increases in claims described as "difficult to diagnose as work
Fraudulent claims.
Increases in mental health-related claims (Tabak 1995, 16).
Still more reasons for the sharp increase of the past decade were listed in a
recent report by the National Council on Compensation Insurance (NCCI) (1992,
Reimbursement for medical care is statutorily regulated and is still
based on fee-for-service in more than half the states.
Demographics define an aging workforce, more susceptible to
injury and illness.
Medical technology continues to offer more expensive equipment
and extensive diagnostic procedures.
Cost-shifting has been documented as the result of the shrinking of
provider reimbursement from employee benefit plans and medicare.
Litigation costs have dramatically increased (Glover 1991, 27).
Each of these items certainly had some impact on the increase of the
workers' compensation costs during the past decade. The literature shows many
reform attempts have been made. A noteworthy point, however, is that pan of the

problem of increased costs with workers' compensation has been the result of
failed attempts to address the costs themselves.
Some argue, for example, that instead of uncovering specific causes
responsible for driving up workers' compensation costs and implementing the
"correct" solutions, managers often rush to buy the "latest" solution.
Consequently, costs are often unintentionally forced upward. One example is the
recent rush to implement toll-free 800 numbers in an effort to streamline the
claims process (Bruce 1996, 33). This was done before the insurers realized that
the forms given the patient and the forms given the operators were different
resulting in an increased per claim time of nearly 30 minutes.
A second example is the current trend toward re-engineering, often without
appropriate forethought of the consequences. The ensuing results have been far
worse than the status quo (Bruce 1996, 33).
The literature is quite clear in recommending that careful study is needed to
define exactly how the provision of medical services impacts the workers
compensation system including a close evaluation of different treatment patterns
for the same diagnosis30 by individual providers (Johnson et al. 1993, 22). Closer
examination at this point of the reasons for increased costs of the workers'
compensation system is in order.
30 Note, this is one of the intended evaluations to be completed in this study.

Reasons for Increasing Workers' Compensation Costs
Cost-Shifting. The Minnesota Department of Labor and Industry
conducted a study in the early 1990s that computed the median cost for treating
back injuries under workers' compensation at $308 with a treatment duration of 21
days. However, under a Blue Cross medical plan in the same area, the median
cost was only $132 over a treatment period of 10 days (Minn. Dept, of Labor
1991, 14).
To understand the importance of this figure, one must also understand that
at the time of this study managed care was more prevalent in Minnesota than
almost any other state (Managed, Advisory Board 1991, 16). Consequently, as
managed care in the general healthcare market began to cut costs, providers still in
the fee-for-service environment began to make up the difference. The result? The
workers' compensation system was the victim of cost-shifting since its planners
were not as aggressive as other payers in using payment reducing tools (Workers'
Comp. Research 1994, 45). Workers' compensation costs would, in part, be
driven higher.
As the tendency to restrict costs on the part of general healthcare insurers
continued, the pressures to shift costs to the workers' compensation environment
could only increase. Many saw the workers' compensation industry as the "last
bastian of fee-for-service medicine" (Resnick 1992, 34) and, therefore, a target for

even greater increases in the future. The literature is clear in suggesting that
without changes the industry can expect providers to continue to charge workers'
compensation patients for the uncompensated care provided to under-insured or
non-insured patients (Greenwood 1992, 66).31
Safetv/Trainine. Another reason why workers' compensation costs rose is
simply because more people were getting hurt than ever before (Resnick 1992,
34). Employers have only to look at themselves. Often inadequate safety
measures lead to increased injury rates. Prevention of injuries is a prime objective
of any workers compensation program. When the appropriate emphasis is not
given to prevention, losses increase on not only the financial level, but, more
tragically, on the human level as well (Frahm 1996, 3).
A company must coordinate all its departments in the safety effort to ensure
that proper prevention is a part of any workers' compensation program. For
instance, employee health workers must know what the occupational physician or
claims adjuster is doing to avoid duplication of care, mistakes and recurrence of
injury (Spieler 1994).
31 It should be noted here that one reason given to justify the increase in cost
for care provided to workers compensation patients is because state regulations
require more than just an evaluation for these patients. For example, a physician
must not only treat a workers compensation patient, but that physician must also
determine when the patient can go back to work and under what conditions. It is
suggested that a portion of the increase could be a result of this.

There appears in an environment of downsizing to be an inverse correlation
between the number of workers in a plant and the number and severity of injuries.
This is a result of fewer people doing the work and working harder and faster than
ever before. The more hours an employee works, the higher the risk of being
injured (Church 1995, 9).
Such reductions of staff also affect the processing of claims. The efficiency
with which the increasing number of claims is processed decreases significantly
with fewer staff members (Church 1995, 10).
Prevention has been called the most important factor in developing a safety
program that effectively meets the needs of employees (ITT 1995). A
representative sample of other suggestions in the literature includes the following
Conduct your own accident investigation after every incident.
Integrate safety into your operations.
Make sure all machinery is equipped with appropriate safety
Provide safety incentives for managers and plant workers.
Ask your insurance company for help (Sellers 1994, 91).
The best safety programs also stress the training of employees as a top
priority (Caldwell 1996, 30). Training should also include a thorough background

on state regulations and focus specifically on mangers and supervisors (Knight
1995, 30). Such a safety program was recently (1993) implemented in Trane
Corporations operations. Trane emphasized training and prevention for new as
well as present employees, all of whom received extensive course work in safety
policies and management, safe work practices, accident and injury reporting, and
unsafe conditions.
Although this is just one example of the potential of such programs,
ultimately, Trane reduced total loss costs 43 percent in two years (Gillis 1995,
70). More importantly than the financial savings resulting from such a program,
the human benefits are justification enough (Fefer 1992, 82).
Medical Costs. To this point we have reviewed in detail some of the
factors which helped to cause the large increases in workers' compensation rates
during the past decade and more. However, we have not yet discussed what is
arguably the largest part of that increase-medical costs.
After reviewing the literature, we see that most authors attribute increases
in the medical cost component of workers' compensation to out-of-control
premium prices in Colorado and across the nation. The only exception is Michael
J. Shor (1995, 43) who argued that, "unlike group health, the bulk of workers
compensation expenditures is found not in medical care." Shor (1995, 43)
continues, "while the actual cost of medical care may be a relatively small pan of

the workers' compensation dollar, the influence of that care over the spending of
the rest of the dollar is overwhelming."
Even Shor agrees that the amount spent on the medical portion of a
workers' compensation claim is important to any final determination of settlement.
However, this literature review makes it fairly clear that no one knows for sure
exactly how much is spent on the healthcare portion of workers' compensation
claims. The published figures vary from 20 percent (Shor 1995, 43) all the way to
more than 50 percent of the total workers' compensation cost (Solomon 1993, 59).
No matter who confronts the issue, a scenario in which the medical cost of
workers' compensation is not a great overall drain is difficult to find.
Colorados Department of Labor recently established an all-out effort to
address the medical cost component of its workers' compensation industry.32 The
department established a three-pronged approach which it felt would address the
issues of increasing medical costs while still ensuring that the workers'
compensation patient received quality medical care (Colo. Dept, of Labor 1994,
15). The department's approach included:
1. Identifying Appropriate Treatment (Treatment Guidelines)
2. Enforcement of Treatment Guidelines (Utilization Standards)
6 Note that these efforts occurred after Senate Bill 91-218 and are in addition
to the price freeze and similar measures.

Payment by Fee Schedule (Colo. Div. Of Workers Comp. 1991,
As a part of this program, Colorado established accreditation for
physicians,33 utilization standards, enhanced fee schedules, etc. The most
important pan of the focus on medical costs was implementation of the
Independent Medical Examination program (Colo. Div. Of Workers' Comp. 1991,
17). This program established a list of physicians who would serve on a rotating
panel to hear complaints about inappropriate treatment given as a part of the
workers' compensation process. The theory behind this program was to increase
the efficiency of the claims process by accelerating review of these questions. The
results of this new program are not firm yet. It is the first of its kind in the nation,
and it will certainly be interesting to monitor the outcome and to see whether it is
adopted throughout the nation (1991, 17).
Nationally, concern over the medical cost component of workers
compensation continues "as the iceberg in front of the Titanic," (Resnick 1992,
35) as one author put it. Some confusion remains about how much medical costs
impact the workers' compensation industry (Frahm 1996, 3; Sellers 1994, 92;
Solomon 1993, 61; Johnson et al. 1993, 23; Hashimoto 1996, 101; Telles 1993,
33 This is the same accreditation for physicians that was cited in the previous

42). There is no misunderstanding about the fact that there is, indeed, an impact of
such costs.
With respect to the medical expenses of workers' compensation, two
factors seem to explain why medical costs increased so fast over the past decade
(aside, of course, from the indirect effect that cost shifting and prevention play, as
described above). Those factors are: 1) unique incentives for both; and, 2)
providers and patients in a workers compensation setting and the amount of abuse
in the current system.
Incentive Structure of the Workers' Compensation System.
The previous section on the general healthcare industry points out that
healthcare is an elastic good, meaning that a 1 percent reduction in the price of
healthcare results in a greater than 1 percent increase in the utilization of
healthcare. Other outside influences must also be addressed in addition to the
relatively elastic nature of healthcare in general.
Recent arguments have suggested that the absence of economic incentives
to limit care have disproportionately increased workers' compensation health costs
(Burton 1993, 1). That is, the annual average rate of increase for the medical care
portion of workers' compensation claims increased 15.2 percent while during the
same period national health expenditure increased at the lower annual rate of 9.5

percent (Burton 1993, 2). One cause is the fact that in a workers' compensation
claim so many parties are interested in the outcome: at the very least, the
employee/patient, the employer, the insurance company, the insurance company
adjuster, the plaintiff bar, organized labor, the regulatory authority, and others
(Shor 1995, 43). These conflicting interests create confusion and also increase
Furthermore, no incentives exist to reduce the costs of workplace accidents
when state laws permit an employee to seek a physician of his or her own
choosing. In the absence of deductibles or co-payments (which are not authorized
in a workers' compensation environment), there is no direct incentive or
opportunity for workers to act as prudent purchasers of healthcare (Ballen 1994,
The argument continues that, since the benefits associated with the
workers' compensation setting have been statutorily mandated and liberally
interpreted (Curtis 1993, 375), no incentives are in place to reduce what
"legislatures and the public view as the high costs of accidents" (Hashimoto 1996,
101). The comparison of characteristic features of individual health insurance and
workers' compensation insurance illustrates this (Table 2.1) (Mandated,
Occupational 1995, 215).

Although the information in this table does not fully explain the differences
of incentives in a general group medical indemnity plan and a workers' compensa-
tion plan, it does provide considerable insight into the expenses involved.
Table 2.1 Health Coverage Differences
Voluntary, Contractual System Mandatory, Statutory System
Cost shared by employees through deductibles, co-payments, etc. First-dollar coverage paid at 100 percent by employer.
Limits on some benefits. Unlimited benefits.
Restricts provider choice after limited initial period. Unrestricted provider choice.
Rates based on treatment expected to be received; rates quoted as monthly amount. Rates based on all treatment for injury or illness; rates quoted as percentage of payroll.
Numerous benefit designs. One state-mandated benefit design.
System Abuses
One would hope that outright fraud and other criminal abuses did not play
a significant role in the increase of workers compensation costs during the past
decade; nevertheless, it would seem somewhat naive to ignore the possibility. In
fact, some have suggested that litigation and fraud in the workers' compensation

system have caused a 10 percent increase in the associated medical costs (Frahm
1996, 3). Two types of fraud are likely: provider fraud and employee fraud.
Provider Fraud. As the effects of managed care throughout the nation
impact the providers of healthcare, they often look elsewhere to replace lost
income. The scam can be as simple as overcharging a patient for medical care
(Johnson et al. 1993, 23), or scheduling the patient for more exams than necessary
(Boden, Johnson and Smith 1992, 28). Sometimes the schemes turn into
something much more elaborate, such as the physician in San Diego who recruited
hundreds of workers to file false claims on the promise of returning a portion of
the money to them. This fraud cost the state of California more than $2 million
(Harmon 1995, 29).
Employee Fraud. Employee fraud sometimes originates from liberally
interpreted state statutes which are often weighted heavily in favor of the employee
(Curtis 1993, 380). The abuse usually comes in the employees effort to "milk"
the system for money and time off work. Large settlements for permanent
disability have generated hundreds of thousands of dollars for the recipients. One
person in Colorado was successful in receiving total permanent disability
settlements totaling more than $1 million from seven different employers before
being caught (Denver Post 12 April 1993, sec. B, 1).

Among the less-aggressive schemes is having a "weekend sports injury"
treated free of charge and on a "paid vacation" from work. As the system is
currently established, prevention of such a scheme is nearly impossible (Resnick
1992, 35).
Cost Cutting Strategies of the Recent Past
Many attempts to reduce the cost of workers' compensation has been made
with varying degrees of success in Colorado and across the nation. Few of the
strategies have incorporated managed care techniques, although such techniques
are gaining more favor. The major reason for not cutting costs via managed care
is that most state legislatures, including Colorado, have statutes that prohibit such
organizations or make it very difficult to use them (National Business Nov. 1991,
66). To date, at least 195 "anti" managed care bills have been approved by state
legislatures (Polakoff 1992, 58). Clearly this bias complicates the resolution of
cost overages. Nevertheless, Colorado currently uses several strategies that are
not prohibited by statute and are not necessarily managed care initiatives
(Workers Comp. Research 1989, 62). Some have met with considerable success
(Telles 1993, 43).

As might be expected, there are as many cost cutting strategies34 as there
are states and employers combined. One of the first strategies adopted by states,
including Colorado, was to simply freeze payments for workers' compensation
services (Doherty 1989, 42). This strategy, as pointed out in the introductory
chapter of this paper, was not without drawbacks of its own. Costs remain frozen
in many states. Interestingly, some have been frozen for so long that as pressure
to decrease the costs of the general healthcare system intensifies, the impetus to
use (and abuse) the workers compensation systems across the nation increases
(Hashimoto 1996, 106).
Exemplifying the employers problem, a study by Towers and Perrin
reported that 71 percent of businesses surveyed said workers compensation costs
threaten their ability to operate their business or stay in business (Hashimoto 1996,
106). The conclusion is then drawn that in some way workers compensation costs
to industry must be cut, by statute or not.
One must realize in reviewing potential cost cutting strategies that
programs used for healthcare services are not necessarily applicable to the
workers' compensation system (Telles 1993, 45). Each states statutes are unique
34 The cost cutting strategies outlined here have been used by workers'
compensation insurers and employers alike in the recent past. Despite the use of
managed care techniques, this section focuses on other cost cutting strategies.
However, the relationship of managed care to workers' compensation is examined
in some detail below.

and applicable. Incentives differ for each. Some cost containment strategies used
in various states, including Colorado, are as follows:
Limited Initial Provider Choice. Twenty-one states, including Colorado,
allow the employer to choose the provider an employee must see after an injury
(Workers' Comp. Research 1992, 33). Rules about provider selection are driven
by employers and insurers who insist that limiting an employees choice of a
physician, keeps costs to a minimum (Telles 1993, 45). Only one published study
disagrees35, and its data finds no correlation between who selects the physician and
the cost of care (Johnson et al. 1996, 23).
Limited Provider Change. Forty states now allow an employer to restrict
the number of times an employee may change physicians for treatment of the same
injury (Workers' Comp. Research 1992, 33). This cost cutting strategy like the
others presented here is not a pan of a managed care initiative, but it is important
because if managed care is to ever be an integral part of the workers'
compensation industry, limiting access to providers will be crucial (Telles 1993,
Regulation of Hospital Payment. Twenty-two states have authorized some
form of restriction on the amount that can be paid to a hospital for services
35 This was the only study this writer saw in which chiropractor care was not
deemed more expensive than other care.

rendered to a workers' compensation patient (Workers' Comp. Research 1992,
33). This restriction, however, has undergone considerable expansion because
almost every state is now negotiating for some form of rate discounts and other
fiscal breaks by healthcare providers, including hospitals (Tillinghast 1991).
Medical Fee Schedules. Medical fee schedules list most medical diagnoses
and the amount to be paid for each. Medical fee schedules restrict the amount paid
to medical interests other than the hospital, such as physicians, chiropractors,
physical therapists, and occupational therapists (Telles 1993, 45). As noted in
Chapter 1, Colorado has elected to restrict the utilization of chiropractors (Berry
1993, 15) along with freezing fees in the effort to contain costs. Other fee
schedule programs remain in effect, and all non-hospital providers are affected.
Utilization Review. Utilization review is simply an attempt by an
insurance company or some other payor to review medical care given and
minimize its scope. Utilization review is easily the most widely used tool to
contain costs. It is certainly an effective tool for the management of patient care,
although it is not a managed care system or program such as a PPO or HMO
(Resnick 1992, 35). Some states restrict the use of this tool to review for medical
reasons only and do not allow it to alter the care that is provided (Workers' Comp.
Research 1992, 18). Yet, altering care would be one motivation for using this tool
as an alternative to managed care. Utilization review combines with prospective,

concurrent, and retrospective review as important parts of the overall utilization
review cycle (Telles 1993, 45). The employer, thereby, hopes to ensure that only
authorized care is given. Even though these programs are statutorily authorized in
only a few states, they are nonetheless quite prevalent.
Case Manager. Another part of the utilization review cycle is the use of
case mangers who track the patients throughout the entire claim settlement
process. This allows the employer to ensure that the employee is working on
limited duty when possible and to supervise the entire case. Using case managers
represents a big commitment on the employers part because people must be hired
to do this work based in the belief that results will improve the bottom line
(Tillinghast 1991, 27).
Even with the availability of these (and other) proven cost containment
strategies, many states have not used them. It is safe to say that states could do
much more to stop runaway workers' compensation costs and to promote related
cost efficiencies (Durbin and Appel 1991, 39). Five factors have been suggested
(Bruce 1996, 32) to take full advantage of accepted cost containment strategies:
1. Corporate Structure. Develop and maintain a structure that
is conducive to flexibility and one in which a systematic approach to
costs is implemented.
2. Usable Data. Minimize data that comes to the head of a
company. Reams of paperwork are of no value to the CEO who
arguably needs only to know that his or her departments are ranked
from best to worst to make the decisions necessary.

3. Account Instructions. Minimize the authority of on-site
claims adjusters. The employer can then get as involved as
necessary to reduce workers' compensation costs.
4. File Reviews. The prudent manager routinely reviews all
files associated with open workers' compensation claims. This
minimizes the possibility that an employee can remain on workers'
compensation payments indefinitely.
5. Vendor Selection. Vendors in this context include insurance
carriers, independent administrators, insurance brokers, structured
settlement firms, investigators, attorneys, medical providers, etc.
Care should be taken in the selection of all such vendors to make
sure that they work together and provide the synergy needed to
enhance the company.
These cost containment strategies have been used successfully by
employers and insurance carriers to decrease the costs of workers' compensation
premiums. The carriers must now provide a competitive product to businesses.
However, only when combined with effective, implementation by employers of the
managed care initiatives reviewed below is maximum efficiency ensured.
Self-Insurance. Another strategy that has been exceptionally effective in
addition to the cost saving strategies listed above is self-insurance. As pointed out
in the previous chapter, companies that meet certain criteria may opt to ensure
themselves to decrease the costs associated with providing workers' compensation
insurance. Any thorough review of relevant literature must, therefore, cover self-

During the first five years of the 1990s, the number of companies in the
United States using self-insurance as a means of reducing costs has increased
nearly 20 percent (Drury 1996, 29). Also, whereas, the use of self-funded plans
was once primarily restricted to large companies, that trend has changed to
encompass smaller companies that pool their resources. The size of employers
who currently utilize some type of self-insurance in the United States is as follows
Clearly, company size correlates with the use of self-insurance as a cost
saving measure. Yet, nearly half the companies that employ only 200 to 499
workers also enjoy self-insurance. With this cost-minimizing advantage, the
utilization of self-insurance is rapidly growing into the most popular solution
(Comp, Transport Topics 1984, 13).
10 Source: Self Insurance Utilization Report, published by Foster Higgins
and Co., New York City, Summer 1995.
Self-insurance Company Size36
Less than 49 employees
50 to 199 employees
200 to 499 employees
500 to 999 employees
1.000 to 4,999 employees
5.000 to 9,999 employees
10.000 to 19,999 employees
more than 20,000 employees
13 percent
34 percent
44 percent
67 percent
78 percent
82 percent
84 percent
91 percent