Federal Employees' Compensation Act: Percentages of Take-Home Pay
Replaced by Compensation Benefits (Letter Report, 08/17/98,
GAO/GGD-98-174).

Pursuant to a congressional request, GAO provided information on
workers' compensation benefits for lost wages provided to workers with
job-related injuries under the Federal Employees' Compensation Act
(FECA), focusing on: (1) the percentages of take-home pay that FECA
benefits replaced for beneficiaries on the long-term rolls who were
receiving full benefits; (2) career patterns of workers in selected
occupations that were the same as the occupations of FECA beneficiaries;
and (3) beneficiaries' characteristics such as current age, age when
injured, compensation benefits paid in 1997, and pay at the time of
injury adjusted to 1997 pay levels.

GAO noted that: (1) for the more than 23,250 beneficiaries on the
long-term rolls for whom GAO developed replacement rates, GAO estimated
that FECA benefits replaced, on average, over 95 percent of the
take-home pay beneficiaries would have received had they not been
injured; (2) estimated replacement rates ranged between about 76 and 136
percent; (3) compensation benefits equaled between an estimated 80 and
99 percent of take-home pay for about 70 percent of these beneficiaries
and amounted to 100 percent or more in 29 percent of the cases; (4)
under assumptions GAO needed to make to compute beneficiaries' income
taxes and retirement contributions, replacement rates tended to be
higher for beneficiaries who: (a) received higher amounts of pay before
their injury; (b) were injured before 1980; (c) received the FECA
dependent benefit; and (d) lived in states with an income tax; (5) using
different assumptions to show their effect on replacement rates,
beneficiaries with more exemptions or deductions for income tax purposes
would have had lower replacement rates because these rates generally
decrease as taxable income decreases; (6) beneficiaries with a spouse
who had taxable income would have higher replacement rates because
replacement rates generally increase as spousal income increases; (7)
single and married beneficiaries who had no income subject to income
taxes while working--generally those with low incomes--would have
replacement rates of about 73 and 82 percent, respectively; (8) GAO's
analyses showed that about 70 percent of all beneficiaries were over 40
years old when they were injured, and the average adjusted pay of
beneficiaries in the selected occupations approximated the average pay
of active workers in the same occupations; (9) GAO was unable to
determine the extent to which beneficiaries' career prospects were
diminished by their on-the-job injuries because GAO's analyses were
limited to readily available data; (10) the career patterns of
individuals depended on a multitude of personal employment factors as
well as the specific jobs in which individuals are employed, according
to agency officials familiar with career patterns of workers; (11) about
65 percent of the 30,000 beneficiaries identified by GAO were over 55
years old, and the average age of beneficiaries was 61, as of June 1997;
and (12) in June 1997, their annual compensation averaged $26,220, and
their average gross pay at the time of injury adjusted to 1997 pay
levels was $34,833.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-98-174
     TITLE:  Federal Employees' Compensation Act: Percentages of 
             Take-Home Pay Replaced by Compensation Benefits
      DATE:  08/17/98
   SUBJECT:  Workers compensation
             Disability benefits
             Beneficiaries
             Federal employees
             Statistical data
             Employee medical benefits
             Federal employee disability programs
IDENTIFIER:  Federal Employees Compensation Act Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Chairman, Subcommittee on Civil Service, Committee on
Government Reform and Oversight, House of Representatives

August 1998

FEDERAL EMPLOYEES' COMPENSATION
ACT - PERCENTAGES OF TAKE-HOME PAY
REPLACED BY COMPENSATION BENEFITS

GAO/GGD-98-174

Workers' Compensation Benefits

(410199)


Abbreviations
=============================================================== ABBREV

  ATC - Air traffic control
  CSRS - Civil Service Retirement System
  FAA - Federal Aviation Administration
  FECA - Federal Employees' Compensation Act
  FERS - Federal Employees Retirement System
  GS - general schedule
  IRS - Internal Revenue Service
  OPM - Office of Personnel Management
  OWCP - Office of Workers' Compensation Programs
  VA - Department of Veterans Affairs
  WEC - Wage-earning capacity

Letter
=============================================================== LETTER


B-279234

August 17, 1998

The Honorable John L.  Mica
Chairman, Subcommittee on
 Civil Service
Committee on Government Reform
 and Oversight
House of Representatives

Dear Mr.  Chairman: 

This report responds to your request for information on workers'
compensation benefits for lost wages provided to workers with
job-related injuries under the Federal Employees' Compensation Act
(FECA) (5 U.S.C.  8101 et seq., as amended).  Specifically, you asked
for information on (1) percentages of take-home pay that FECA
benefits replaced\1 for beneficiaries on the long-term rolls\2

who were receiving full benefits; (2) career patterns of workers in
selected occupations that were the same as the occupations of FECA
beneficiaries, which might indicate the extent to which
beneficiaries' injuries affected their career progression prospects;
and (3) beneficiaries' characteristics such as current age, age when
injured, compensation benefits paid in 1997, and pay at the time of
injury adjusted to 1997 pay levels.  As agreed with your office, we
did not assess the fairness, adequacy, or equity of the benefits
provided nor did we compare or contrast the different methods--gross
pay or take-home pay--by which workers' compensation benefits are
calculated under federal or states' workers' compensation laws. 

In calculating percentages of take-home pay replaced by FECA
benefits, we estimated gross pay and net take-home pay for over
23,250 beneficiaries who were receiving full FECA wage-loss
compensation benefits of either 66-2/3 or 75 percent\3 of their gross
pay as of June 1997.  These beneficiaries lived in 19 states, 4 of
which did not have a state income tax.\4 We estimated injured
workers' gross pay at the time of their injury using information on
current FECA benefits, FECA cost-of-living increases, and average pay
comparability increases for active workers.  Like workers'
compensation organizations, we defined take-home pay as gross pay
less mandatory deductions for retirement and Medicare contributions
and federal and state income taxes.  To estimate take-home pay, we
had to make assumptions about spouses' income and beneficiaries'
dependents and deductions for income tax purposes.  FECA benefits are
not subject to federal or state income taxes. 

For workers in (1) letter carrier and postal distribution occupations
with the United States Postal Service, (2) nursing positions with the
Department of Veterans Affairs (VA), and (3) air traffic control
(ATC) positions with the Federal Aviation Administration (FAA), we
obtained information on their career patterns and potential merit
increases and promotions from agency officials familiar with their
career progression to compare with available profile information for
beneficiaries in the same occupations at the time of injury.  We
obtained profile and compensation benefit information on injured
workers on FECA's long-term rolls from the Department of Labor's
automated information systems used to manage FECA claims and pay FECA
beneficiaries.  Appendix I contains a detailed description of our
scope and methodology. 

We requested written comments on a draft of this report from the
Secretary of Labor.  Labor's comments are discussed at the end of
this letter and reprinted in appendix IV.  We performed our work from
October 1997 to July 1998 in accordance with generally accepted
government auditing standards. 


--------------------
\1 Unless otherwise noted, when we refer to the percentage of
take-home pay replaced by FECA benefits (i.e., replacement rate), we
are referring to the workers' gross pay at the time of their
injuries, adjusted to 1997 pay levels less deductions for retirement
contributions and federal and state income taxes, if applicable, and
compared with amounts of FECA benefits received in June 1997. 

\2 Beneficiaries on the long-term rolls are those with permanent
disabilities or with injuries that have lasted or are expected to
last for prolonged periods (over 1 year). 

\3 Beneficiaries without dependents receive benefits based on 66-2/3
percent of pay.  Beneficiaries with at least one dependent receive
benefits of 75 percent of pay. 

\4 One state with an income tax did not tax income from salaries and
wages.  We included this state as one of the four without an income
tax. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

For the more than 23,250 beneficiaries on the long-term rolls for
whom we developed replacement rates, we estimated that FECA benefits
replaced, on average, over 95 percent of the take-home pay
beneficiaries would have received had they not been injured. 
Estimated replacement rates ranged between about 76 and 136 percent. 
Compensation benefits equaled between an estimated 80 and 99 percent
of take-home pay for about 70 percent of these beneficiaries and
amounted to 100 percent or more in 29 percent of the cases. 

Under assumptions\5 we needed to make to compute beneficiaries'
income taxes and retirement contributions, replacement rates tended
to be higher for beneficiaries who (1) received higher amounts of pay
before their injury, (2) were injured before 1980, (3) received the
FECA dependent benefit, and (4) lived in states with an income tax. 
Using different assumptions to show their effect on replacement
rates, beneficiaries with more exemptions or deductions for income
tax purposes would have had lower replacement rates because these
rates generally decrease as taxable income decreases.  Beneficiaries
with a spouse who had taxable income would have higher replacement
rates because replacement rates generally increase as spousal income
increases.  Single and married beneficiaries who had no income
subject to income taxes while working--generally those with low
incomes--would have replacement rates of about 73 and 82 percent,
respectively. 

Our analyses showed that about 70 percent of all beneficiaries were
over 40 years old when they were injured, and the average adjusted
pay of beneficiaries in the selected occupations approximated the
average pay of active workers in the same occupations.  These
characteristics might suggest that the beneficiaries were not in the
early stages of their careers at the time of their injuries. 
However, we were unable to determine the extent to which
beneficiaries' career prospects were diminished by their on-the-job
injuries because our analyses were limited to readily available data. 
Occupational data were available for only about one-third of the
beneficiaries we analyzed, and data were not readily available on
beneficiaries' career progression up to the time of their injuries. 

Further, career patterns of individuals depended on a multitude of
personal and employment factors as well as the specific jobs in which
individuals are employed, according to agency officials familiar with
career patterns of workers in selected Postal Service, FAA, and VA
occupations.  Profile information on FECA beneficiaries in these and
other occupations is given beginning on page 9 and in appendix III. 

Characteristics of the more than 23,250 beneficiaries for whom we
developed replacement rate information and for approximately 6,800 of
the other 11,460 FECA beneficiaries\6 on the long-term rolls who
received FECA wage-loss compensation benefits of either 66-2/3 or 75
percent of their gross pay were

  -- About 65 percent of these 30,000 beneficiaries were over 55
     years old, and the average age of these 30,000 beneficiaries was
     61, as of June 1997. 

  -- At their date of injury, almost 70 percent of them were over 40
     years old, and their average age was over 45 years old. 

  -- In June 1997, their annual compensation benefits averaged
     $26,220, and their average gross pay at the time of injury
     adjusted to 1997 pay levels was $34,833. 


--------------------
\5 For our principal analyses, we assumed that beneficiaries (1)
claimed standard deductions for income tax purposes; (2) receiving
the FECA dependent benefit, had a spouse with no taxable income; and
(3) participated in the Civil Service Retirement System (CSRS). 

\6 See appendix I for details on which beneficiaries were included in
our analyses. 


   BACKGROUND
------------------------------------------------------------ Letter :2

The relationship between nontaxable workers' compensation benefits
received by workers who were injured on the job and their income at
the time of injury has been the subject of discussion among workers'
compensation analysts for a long time.  According to the 1972 Report
of the National Commission on State Workmen's Compensation Laws,

     "A basic objective of a modern workmen's compensation program is
     to provide protection to workers against loss of income from
     work-related injuries and diseases.  To achieve this goal, the
     program must carefully weigh the worker's interest in
     substantial income benefits against factors such as the loss of
     incentive for rehabilitation, which some believe may occur if
     income benefits are too high."

The 1972 National Commission's Report recommended that workers'
weekly benefits should replace at least 80 percent of their spendable
weekly earnings, subject to a state's maximum weekly benefit.  As
states increased workers' compensation benefits following the
National Commission's report, an issue arose as to whether benefits
were so high that incentives for injured employees to return to work
might be impaired.  Workers' compensation program analysts are
reluctant to take a position on what the "correct" level of workers'
compensation benefits should be, leaving that matter to the judgment
of legislators.  According to a 1985 Workers Compensation Research
Institute\7 report, legislatures in many states must walk a fine line
between benefits that are high enough to provide adequate income, but
not so high as to discourage an employee's return to work when he or
she is no longer disabled. 

In addition to discussions about the appropriateness of workers'
compensation programs' benefit levels, some observers have made the
point that beneficiaries with long-term or permanent disabilities who
were injured early in their careers may have lost promotions or other
opportunities to increase their pay relative to the compensation
benefits they may be currently receiving. 

Under FECA, workers' compensation benefits for those who are totally
disabled are 66-2/3 percent of wages for workers without dependents
and 75 percent of wages for workers with one or more dependents. 
These benefits are not subject to federal or state income taxes. 
Most states' workers' compensation programs provide benefits ranging
from 60 to 72 percent of gross wages.\8 Six states use a percentage
of spendable earnings\9 (ranging from 75 to 80 percent) rather than
wages as the basis for computing compensation benefits. 

The Department of Labor's Office of Workers' Compensation Programs
(OWCP) is responsible for administering FECA and adjudicating claims
submitted on behalf of injured workers.  For the year ending June
1997, FECA costs totaled about $1.9 billion--$1.3 billion for
compensation benefits, $444 million for medical benefits, and $125
million for death benefits.  For this period, OWCP paid medical
benefits in about 238,450 cases, death benefits in over 6,260 cases,
and compensation benefits in over 78,060 cases.  Of these 78,060
cases, 51,265 were on the long-term rolls, as of June 1997.  In these
51,265 cases, about 34,700 totally disabled individuals were
receiving FECA wage-loss benefits at either the 66-2/3 or 75 percent
rate. 


--------------------
\7 The Workers Compensation Research Institute is a not-for-profit
research organization whose mission is to provide objective
information about public policy issues involving workers'
compensation systems. 

\8 Our report entitled Workers' Compensation:  Selected Comparisons
of Federal and State Laws (GAO/GGD-96-76, Apr.  3, 1996) discusses
and compares federal and state workers' compensation laws in more
detail. 

\9 Spendable earnings (take-home pay) for working employees are
computed by taking an employee's before-tax earnings at the time of
injury and subtracting Social Security taxes and federal and state
income taxes.  The taxes are taken from published withholding tables
that are based on average tax rates, given an employee's actual
exemptions and a standard deduction. 


   ANALYSES OF TAKE-HOME PAY
   REPLACEMENT RATES
------------------------------------------------------------ Letter :3

For the more than 23,250 beneficiaries included in our analyses, we
estimated that FECA benefits replaced, on average, over 95 percent of
the take-home pay they would have received had they not been injured. 
Figure 1 shows percentages of beneficiaries whose FECA benefits
resulted in various ranges of take-home pay replacement rates. 

   Figure 1:  Percentages of
   Beneficiaries and Their
   Take-Home Pay Replaced by FECA
   Benefits as of June 1997

   (See figure in printed
   edition.)

Note:  In calculating these replacement rates, we assumed that
beneficiaries (1) claimed standard deductions for income tax
purposes; (2) receiving the FECA dependent benefit, had a spouse with
no taxable income; and (3) participated in CSRS. 

Source:  GAO analysis of OWCP data. 

Beneficiaries' estimated take-home pay replacement rates ranged from
a low of about 76 percent to a high of 136 percent depending on when
they were injured, their pay when injured, and whether they had
dependents or lived in a state with an income tax.  To calculate
federal and state income taxes to use in computing beneficiaries'
take-home pay, we had to make assumptions regarding the amount of
taxable income earned by a beneficiary's spouse and the number of
exemptions and amounts of deductions claimed for income tax purposes. 
Although OWCP's automated databases identified beneficiaries
receiving FECA dependent's benefits, they did not contain information
on spouses' income, additional exemptions, or additional deductions. 

Under our assumptions, replacement rates were affected by (1)
beneficiaries' dates of injury, (2) pay levels and progressive income
tax rate structures, (3) benefit rates based on the absence or
presence of dependents, and (4) beneficiaries' states of residence. 
The effects of these variables on replacement rates are summarized
below and discussed in more detail in appendix II. 


      LENGTH OF TIME ON LONG-TERM
      ROLLS INCREASES REPLACEMENT
      RATE
---------------------------------------------------------- Letter :3.1

In general, the older the date of injury, the higher the replacement
rate.  The older dates result in higher replacement rates because
over long periods of time, FECA cost-of-living increases\10 exceeded
general schedule (GS) pay increases that individuals would have
received had they not been injured.  To illustrate in one case, a
worker with an injury date just before March 1, 1996, would have
received the March 1, 1997, FECA cost-of-living increase of 3.3
percent.  Workers who had not been injured would have received a
general schedule pay increase averaging 3 percent in January 1997. 
In another case, a worker injured in January 1970 would have received
FECA cost-of-living increases through March 1997 and, in absolute
numbers, these increases would have totaled 139.5 percent of
compensation.  General schedule pay increases for workers who had not
been injured would have averaged 118.7 percent of pay over the same
period.  The replacement rate for a single person receiving FECA
benefits of $20,000 in June 1997 in the first case would be 83.6
percent of take-home pay, whereas, in the second, older case, it
would be 101.3 percent of take-home pay. 


--------------------
\10 Starting in 1966, FECA provided for adjustments in compensation
benefits based on the Consumer Price Index.  These cost-of-living
increases are provided to injured employees who stopped work on
account of an injury more than 1 year prior to the effective date of
the increase.  In contrast, general schedule pay increases, although
specified by statute, generally emerge from budget negotiations each
year. 


      HIGHER PAY LEVELS AND
      PROGRESSIVE INCOME TAX RATES
      INCREASE REPLACEMENT RATES
---------------------------------------------------------- Letter :3.2

Because the federal government and many states have progressive
income tax rate structures, workers generally pay taxes at higher
rates as their taxable income increases.  In our analyses, applicable
federal income tax rates ranged from 15 to 31 percent of taxable
income and state income tax rates ranged from 0.5 to 9.3 percent of
taxable income.  For beneficiaries who earned higher pay, nontaxable
FECA benefits replaced pay that would have been subject to higher tax
rates.  FECA benefits replaced an estimated 91 percent of take-home
pay for beneficiaries whose pay before the injury, adjusted to 1997
pay levels, was under $20,000.  For beneficiaries with pay over
$60,000, FECA benefits replaced over 105 percent of take-home pay. 


      FECA DEPENDENT BENEFIT
      INCREASES REPLACEMENT RATES
---------------------------------------------------------- Letter :3.3

Replacement rates for FECA beneficiaries receiving the dependent
benefit averaged an estimated 97 percent compared with 92 percent for
beneficiaries who did not receive this benefit.  FECA authorizes an
additional 8-1/3 percent in benefits for beneficiaries with
dependents.  If these additional benefits were not provided, some
beneficiaries' replacement rates would be lower because their
take-home pay would be compared with a compensation benefit of 66-2/3
percent rather than 75 percent of gross pay. 


      STATE INCOME TAXES INCREASE
      REPLACEMENT RATES
---------------------------------------------------------- Letter :3.4

Replacement rates for beneficiaries who lived in states that taxed
income were, on average, an estimated 96 percent compared with about
94 percent for those living in states with no income tax.  Like
federal income taxes, income taxes that workers paid to states before
they were injured would serve to further reduce their take-home pay,
thereby increasing the portion of take-home pay replaced by
nontaxable FECA benefits. 


      DIFFERENT ASSUMPTIONS WOULD
      CHANGE ESTIMATED REPLACEMENT
      RATES
---------------------------------------------------------- Letter :3.5

To calculate take-home pay replacement rates, we made certain
assumptions about beneficiaries based on data that were readily
available to us.  The effects of using different assumptions on
spouses' income, numbers of exemptions, and amounts of deductions are
summarized below and discussed here and in more detail in appendix
II. 

Spouse's income.  In estimating replacement rates for beneficiaries
with a spouse, we assumed that their spouses did not have taxable
income.  If spouses had income, replacement rates could be higher. 
The presence of a spouse's income results in a higher effective rate
of tax on the income earned by the beneficiary returning to work.  A
higher effective tax rate means that the returning worker's take-home
pay could be lower and, therefore, the ratio of FECA benefits to
take-home pay could be higher. 

Number of dependents (exemptions).  In computing federal income
taxes, we assumed that beneficiaries who received augmented FECA
benefits had one dependent and that the dependent was a spouse.  In
1997, each exemption claimed was worth $2,650 in computing taxable
income.  Replacement rates would have decreased by about 1.5 percent
for each additional exemption.  For example, the replacement rate for
a married worker with 1 child (3 exemptions), with income of $30,000,
would have been 89.3 percent compared with 90.7 percent for a couple
(2 exemptions).  We did not assume additional exemptions for age or
blindness. 

Tax deduction amounts.  In computing income taxes, we assumed that
beneficiaries would have claimed federal standard deduction amounts
of either $4,150 if single, or $6,900 if married.  If these
individuals had itemized deductions that were either double or triple
the standard deduction amounts, their take-home pay replacement rates
would have been lower than our estimates by percentages ranging from
about 2 to 7 percent depending on (1) whether they were single or
married and (2) their pay before being injured. 

For beneficiaries who did not have taxable income while working
either because they had low income, large deductions, or multiple
dependents, replacement rates would have been about 73 percent if
single, or about 82 percent if married.  For beneficiaries who did
not owe income taxes, their take-home pay would be gross pay minus
deductions of 8.45 percent for retirement and Medicare benefits.  The
relationship between FECA benefits (either 66-2/3 or 75 percent of
gross pay) and take-home pay would be the same (about 73 or 82
percent, respectively) at all pay levels. 


   CAREER PATTERNS FOR WORKERS IN
   SELECTED OCCUPATIONS
------------------------------------------------------------ Letter :4

We were unable to determine whether beneficiaries' career progression
patterns were affected by their on-the-job injuries.  Our analyses
showed that about 70 percent of all beneficiaries were over 40 years
old when they were injured, and the average adjusted pay of
beneficiaries in the selected occupations approximated the average
pay of active workers in the same occupations.  These characteristics
might suggest that the beneficiaries were not in the early stages of
their careers at the time of their injuries.  However, our analyses
were limited because occupational data were available for only about
one-third of the beneficiaries and because data were not readily
available on beneficiaries' career progression up to the time of
their injuries. 

Career pattern information we obtained from agency officials for
workers in selected occupations who were in the same occupations as
selected FECA beneficiaries included in our analysis--letter
carriers, postal distribution workers, registered nurses, practical
nurses, nursing assistants, and air traffic controllers--indicated
that career patterns can vary widely.  These occupations were
selected because they were either the occupations (1) that were the
most frequently identified in the OWCP information we analyzed or (2)
for which many beneficiaries were likely to be employed by the same
agency.  Career pattern information obtained from the above officials
and information on FECA beneficiaries from OWCP's records are
discussed in the following sections. 


      LETTER CARRIER AND POSTAL
      DISTRIBUTION OCCUPATIONS
---------------------------------------------------------- Letter :4.1

According to FECA data, at the time of injury, the average age for
the 1,897 letter carriers and postal distribution workers we could
identify was about 42 years old.  The pay of these workers at the
time of injury adjusted to 1997 pay levels averaged $35,054 and
$36,588, respectively. 

According to Postal Service officials, workers in letter carrier and
postal distribution crafts are covered under union contracts with
Postal Service management.  Entry-level pay in March 1997 for workers
in these crafts was $26,375 and $22,404, respectively.  Upon
completing contractual waiting periods, these workers would
automatically receive longevity-step increases.  Workers would
normally progress from entry-level pay to maximum pay within the same
grade in 12.4 years.  For letter carriers whose entry level is grade
5, maximum pay was $36,863 as of March 1997; for postal distribution
workers whose entry level is grade 4, maximum pay was $35,118.  In
addition to their basic pay, these workers may also receive premium
pay for night or Sunday work.\11

Postal Service officials told us that most letter carriers and postal
distribution workers remain in the same pay grade throughout their
careers.  They usually receive longevity-step pay increases and twice
yearly cost-of-living increases.  As of September 1997, almost 80
percent of about 52,650 postal distribution workers were at grade 4
and almost 50 percent of the 40,877 workers in this pay grade were in
the highest step.  For the approximately 201,500 letter carriers,
about 85 percent (172,590) were at grade 5 and of these, over 70
percent (123,250) were in the highest step. 


--------------------
\11 Because FECA compensation benefits are based on gross pay amounts
that may include premium pay for night or shift differentials and for
Sunday and holiday work, Postal Service workers' pay adjusted to 1997
pay levels may be somewhat inflated when compared with the maximum
basic pay of letter carriers and postal distribution workers. 


      NURSING OCCUPATIONS
---------------------------------------------------------- Letter :4.2

As table 1 shows, the average ages and adjusted pay of the 445
beneficiaries in nursing occupations approximated the average ages
and pay of both VA and non-VA nurses. 



                                Table 1
                
                 Pay and Age Information for Workers in
                          Nursing Occupations

                                                              OPM data
                                          FECA                for non-
Occupations                             data\a     VA data   VA nurses
----------------------------------  ----------  ----------  ----------
Registered nurses                          185      32,643       7,066
Average age                               48.6        46.7        46.5
Average pay                            $43,254     $47,530     $44,623
Practical nurses                           127       9,294       2,056
Average age                               47.4        44.8        47.4
Average pay                            $27,577     $27,514     $25,576
Nursing assistants                         133      10,095       1,346
Average age                               46.9        46.0        44.9
Average pay                            $24,937     $23,614     $22,023
----------------------------------------------------------------------
\a Information on beneficiaries' ages at time of injury and pay at
time of injury adjusted to 1997 pay levels. 

Source:  GAO analysis of FECA, VA, and Office of Personnel Management
(OPM) information. 

The entry level for most of VA's registered nurses in clinical
practice is generally somewhere between the equivalent of a GS-6 and
GS-8, according to a VA official familiar with the typical career
patterns of VA nurses.  Licensed practical nurses generally start at
the equivalent of a GS-4, and nursing assistants are generally hired
at the equivalent of a GS-3. 

According to the official, registered nurses with a bachelor of
science degree generally advance to the equivalent of a GS-11 in 3 to
5 years; nurses without a bachelor's degree generally advance to the
equivalent of a GS-9.  Nurses who reach the equivalent of a GS-12
would usually have a bachelor of science degree and function in
positions with responsibilities beyond the staff nurse.  These
additional responsibilities would include being a nurse manager, head
nurse, care manager, or instructor.  Furthermore, for nurses to
advance beyond the GS-12 level, they generally would have to have a
master's degree.  In addition to clinical practice, some VA
registered nurses become involved in education and training,
administration, or research activities for which they would generally
be paid at the GS-12 or GS-13 levels.  According to VA pay
information, about 700 of VA's 32,600 registered nurses serve in
executive, supervisory, or management positions with pay equivalents
in the GS-14/15 range. 

The VA official told us that over a 3- to 5-year period, the highest
grade to which VA's nursing assistants would likely advance would be
the equivalent of a GS-5.  Most nursing assistants would be at the
GS-4 level.  Within about 2 years, VA's licensed practical nurses
could reach the equivalent of a GS-5 and within 4 to 5 years a GS-6. 
Most practical nurses would function at the GS-5 level.  To receive
higher pay, some nursing assistants would change career patterns and
work as radiological or medical technicians, or as physical
therapists.  Some practical nurses return to school to become
registered nurses or transfer to other VA departments. 


      AIR TRAFFIC CONTROL (ATC)
      OCCUPATIONS
---------------------------------------------------------- Letter :4.3

For the 74 beneficiaries we identified in ATC occupations, FECA
information showed that at the time of injury, their average age was
39.4 and their average pay adjusted to 1997 pay levels was $68,074. 
According to 1997 OPM information, individuals in ATC occupations
averaged almost 42 years of age with average pay of over $65,230. 
About 43 percent of these individuals were at the GS-14 level with
average pay of about $74,750. 

According to an FAA official, most individuals in ATC occupations
begin their careers as GS-7s.  About 75 percent of these individuals
have ATC responsibilities at either air route traffic control centers
or at FAA towers or terminals.  Other individuals in ATC occupations
serve as flight service station specialists and have responsibility
for providing pilots with weather briefings and receiving flight
plans filed by airlines and pilots.  The size and type of FAA
facility at which air traffic controllers serve generally determine
their typical career patterns.  According to the official,
controllers stationed at air route traffic control centers and the
busier airports would generally reach the GS-14 level.  Those serving
at smaller airports would generally reach the GS-12 or GS-13 level
depending on the amount of air traffic serviced by the facility. 

FECA and OPM information for individuals in additional occupations is
shown in appendix III. 


   PROFILES OF FECA BENEFICIARIES
------------------------------------------------------------ Letter :5

For the more than 30,000 beneficiaries we profiled, annual
compensation benefits averaged about $26,220, and the current value
of their gross pay before they were injured averaged $34,833.  About
70 percent of the beneficiaries were over 40 years old when injured. 
As of June 1997, about 65 percent were over 55 years old.  About 73
percent of the beneficiaries had a spouse or at least one dependent. 
For about 90 percent of the 30,000 beneficiaries, the current value
of their pay before they were injured was under $50,000 after
adjusting for pay comparability increases. 

Figure 2 contains profile information on the percentages of
beneficiaries with and without dependents, by age ranges when they
were injured and as of June 1997, by amounts of annualized workers'
compensation benefits, and by amounts of pay received at the time of
injury adjusted to 1997 pay levels. 

   Figure 2:  Profile Information
   on Selected FECA Beneficiaries
   on the Long-Term Compensation
   Rolls as of June 1997

   (See figure in printed
   edition.)

Source:  GAO analysis of OWCP data. 

In addition, over 18 percent (5,549) of the more than 30,000
beneficiaries lived in states that did not have an income tax.  As of
June 1997, about 74 percent of the beneficiaries lived in the same
state as the one where they were injured. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :6

We obtained written comments on a draft of this report from the
Department of Labor.  Labor commented that the report did a good job
of describing the various assumptions and methodology we used to
develop the replacement rate estimates and was very clear on how
changes in each individual assumption would generally affect the
replacement rates for classes of workers. 

Labor also suggested that our analysis might have been better
informed, if instead of assuming that all beneficiaries receiving
augmented benefits had a nonworking spouse, we could have used
readily available data and statistical sampling techniques to develop
replacement rate estimates that took into consideration the incidence
of dual earners, the amounts of income earned by these couples, and
estimates of the number and distribution of additional dependents by
household.  Labor added that in general it might have been more
useful if we had offered some estimates based on likely combinations
of assumptions and that varying assumptions one by one, while it
illustrated an impact or tendency, was probably misleading when
applied universally to all cases. 

In view of the time constraints we faced when we started this
assignment, we chose to develop take-home pay replacement rates based
on a methodology that was similar to those that had been used in
other workers' compensation studies, such as those conducted by the
Workers Compensation Research Institute.  We agree with Labor that it
may have been possible to develop a more refined estimate of the
overall replacement rate had we used other sources of information to
make additional assumptions about FECA beneficiaries.  We also agree
that had we developed and analyzed likely combinations of other
assumptions, we could have presented different estimates of take-home
pay replacement rates.  However, we believe that our methodology
provided a useful overall replacement rate estimate that was based on
reasonable assumptions.  Because we recognized that our result was
dependent on the different assumptions we made, we both acknowledged
this and provided a set of analyses that illustrated the sensitivity
of our result to alternative assumptions. 

Had we developed alternative estimates using additional data or
combinations of alternatives as Labor suggested, those estimates
would have been dependent on limitations inherent in these additional
sources of data and any further assumptions about the beneficiary
population.  In any event, the alternative replacement rate estimates
suggested by Labor may or may not reflect FECA beneficiaries' actual
replacement rates.  For example, regarding marital status, we assumed
that all beneficiaries who received the augmented dependent benefit
had a spouse because the automated database did not distinguish
between beneficiaries who were married or unmarried.  Although the
presence of spousal income would influence replacement rates, income
of other dependents generally would not.  Because an unknown number
of beneficiaries may not have had a spouse, but rather a dependent
such as a child or parent, we chose not to estimate the amount of
income that may be associated with an unknown number of spouses. 
Recognizing that this would tend to understate our replacement rate
calculations, we supplemented our primary analysis with examples of
how changes in assumptions on spousal income would affect our
replacement rate calculations, but we did not intend that the
examples be applied universally to all cases. 

Labor also provided several other suggestions for expanding our
analysis.  These suggestions and our detailed responses to them are
contained in appendix IV. 


---------------------------------------------------------- Letter :6.1

As agreed with your office, unless you announce the contents of this
report earlier, we plan no further distribution of this report until
10 days after its issue date.  At that time we will send copies of
this report to the Chairmen and Ranking Minority Members of the House
Committee on Education and the Workforce and its Subcommittee on
Workforce Protections; the House Committee on Government Reform and
Oversight and its Subcommittee on Government Management, Information,
and Technology; the Senate Committee on Governmental Affairs and its
Subcommittee on International Security, Proliferation and Federal
Services; other interested congressional committees and members; the
Secretaries of Labor, Transportation, and VA; the Postmaster General
of the United States; and the Directors of the Office of Management
and Budget and OPM.  Copies will be made available to others on
request. 

Major contributors to this report are listed in appendix V.  Please
contact me at (202) 512-8676 if you or your staff have any questions
concerning this report. 

Sincerely yours,

Michael Brostek
Associate Director
Federal Management and
  Workforce Issues


SCOPE AND METHODOLOGY
============================================================ Chapter I


   REPLACEMENT RATES
---------------------------------------------------------- Chapter I:1

To estimate the percentages of take-home pay replaced by FECA
benefits, we first identified, for the "chargeback year"\12 ending in
June 1997, beneficiaries on the long-term rolls who received
unreduced wage-loss compensation benefits and the dollar amounts of
their benefits.  Because collecting this information from
beneficiaries' case files maintained in OWCP's district offices would
have been time consuming and expensive, we used OWCP's automated
claims management and compensation payment systems to obtain this
information.  We did not independently verify the data obtained from
these automated systems. 

We then estimated beneficiaries' take-home pay by calculating the
current value of their pay at the time of injury and deducting
amounts for retirement benefit contributions and federal and state
income taxes.  Various workers' compensation organizations define
take-home pay as the difference between an employee's estimated gross
wages less deductions for the employee's share of mandatory
retirement contributions; federal income taxes; and, if applicable,
state income taxes.  We did not take into consideration discretionary
deductions that employees could take for items such as thrift savings
plan contributions, health and life insurance, and savings bonds
because they are not commonly taken into account in workers'
compensation take-home pay calculations.  For our calculations, we
made assumptions about beneficiaries' federal retirement plan
participation, marital status, numbers of dependents, amounts of
deductions to determine taxable income, and spouses' incomes. 
Finally, we estimated beneficiaries' replacement rates by dividing
their FECA benefits by their take-home pay. 

Of the approximately 78,000 beneficiaries who received compensation
benefits for the year ending in June 1997, 51,265 were on OWCP's
long-term rolls.  OWCP had placed most of these 51,265 beneficiaries
into one of the following three wage-earning capacity (WEC)
categories based on the extent of their disability. 

No WEC.  In general, totally disabled beneficiaries who have little
or no reemployment potential.  These beneficiaries receive unreduced
workers' compensation benefits. 

WEC undetermined.  Beneficiaries with temporary total disabilities
who also receive unreduced workers' compensation benefits.  Labor's
procedures call for it to review the status of these cases once a
year. 

WEC established.  Beneficiaries who received reduced compensation
benefits because they were partially disabled and either were working
or had the ability to work.  Compensation benefits are determined by
a formula that takes actual or potential earnings into consideration. 

We obtained information on 30,057 beneficiaries on the long-term
rolls who either did not have a WEC or had an undetermined WEC and
whose last two benefit payment checks for the 1997 chargeback year
were for the same amount.  We selected cases in which the last two
checks were the same to eliminate cases in which beneficiaries
received either lump-sum payments or payments for only a portion of
the 4-week period normally covered by a payment. 

For our analyses, we excluded FECA beneficiaries who (1) were
expected to receive benefits for relatively short periods before
returning to work; (2) received schedule awards;\13 (3) had
established WECs; (4) lived overseas; or (5) received FECA benefits
that were less than the minimum authorized under FECA because they
were part-time or nonfederal employees (e.g., Civil Air Patrol). 

In many cases, the calculation of take-home pay replacement rates
required the computation of individual states' income taxes.  To
limit the number of states for which we needed to make these
calculations, we limited our replacement rate analyses to about 75
percent of the 30,057 beneficiaries selected.  We chose states where
the largest number of these beneficiaries resided as of June 1997,
until we had selected enough states--19--to include about 75 percent
of beneficiaries.  We then developed replacement rate information for
23,257 beneficiaries (77 percent), who resided in 19 states, 4 of
which did not have an income tax.\14

Beneficiaries' actual pay at the time of injury could not be
efficiently determined because this information is only available
from beneficiaries' case files located in OWCP's district offices. 
We therefore made several calculations to estimate the current value
of beneficiaries' take-home pay.  First, we recomputed beneficiaries'
workers' compensation benefits to reflect benefits received at the
time of injury by reducing their June 1997 benefits by the amount of
periodic cost-of-living allowances they received.  Second, based on
this recomputation of benefits received at the time of injury and
whether they had at least one dependent as of June 1997, we
calculated employees' pay before their injury based on either the
66-2/3 or 75 percent benefit level.  Third, by increasing employees'
pay at the time of injury by average federal pay comparability
increases\15 authorized since then, we calculated the current value
of beneficiaries' pay\16 at the time of injury.  Fourth, from this
amount, we made deductions for retirement benefit contributions;
federal income taxes; and, where applicable, state income taxes in
computing the current value of beneficiaries' take-home pay.  Lastly,
we compared current FECA benefits received to these take-home pay
amounts to determine take-home pay replacement rates. 

Postal Service, blue-collar, and certain other federal employees are
in pay plans that differ from the general schedule plan that covers
most federal civilian workers.  In computing the current value of
workers' pay before the injury, we used OPM information on average
federal pay comparability increases applicable to the general
schedule because OWCP's automated databases did not contain
sufficient information to identify either the occupations of over
two-thirds of the beneficiaries we analyzed or the pay plans of
beneficiaries. 

From OWCP and OPM information, we determined FECA benefit levels, the
presence or absence of dependents, whether the beneficiary resided in
a state with an income tax, and beneficiaries' estimated pay at the
time of injury.  However, to develop our estimates of take-home pay
replacement rates, we also needed to make assumptions regarding
beneficiaries' retirement and Medicare contributions, numbers of
exemptions, amounts of itemized deductions (if taken), and spousal
income.  Changing the assumptions would change the estimated ratio of
FECA benefits to take-home pay.  The assumptions we made in
calculating take-home pay replacement rates for our principal
analyses follow.  While information was not readily available to
support different assumptions, we used different assumptions about
numbers of exemptions, income tax deductions, and spousal income to
illustrate how they could influence take-home pay replacement rates. 

Retirement and Medicare contributions.  We assumed that all
beneficiaries participated in CSRS and Medicare and that total
deductions for these programs were 8.45 percent.  Our profile
information indicated that a high percentage of FECA beneficiaries on
the long-term rolls were over 55 years old or were injured many years
ago.  Because the Federal Employees Retirement System (FERS) was not
established until 1986, we assumed that most beneficiaries would have
been CSRS participants. 

Under both CSRS and FERS, deductions for retirement and Medicare
benefits totaled 8.45 percent.  Under CSRS, deductions in 1997 were 7
percent for retirement benefits and 1.45 percent for Medicare
benefits.  Under FERS, deductions were 6.2 percent for Social
Security retirement benefits, 0.8 percent for a FERS annuity, and
1.45 percent for Medicare benefits.  However, under FERS, the 6.2
percent contribution for Social Security retirement benefits applied
to only the first $65,400 of pay in 1997.  Thus, take-home pay under
our assumptions would be understated for the relatively small number
of FECA beneficiaries whose 1997 pay was over $65,400 and who were in
FERS.  In these cases, replacement rates would be lower. 

Exemptions (dependents).  In those cases in which FECA beneficiaries
received augmented FECA benefits of 8-1/3 percent, the database did
not indicate the exact number of dependents because the benefit is
the same whether the beneficiary had one or more dependents.  We
assumed that such beneficiaries had only one dependent and that the
dependent was a spouse.  We made this assumption based on the average
age of the beneficiaries analyzed and to simplify our tax and
take-home pay calculations.  In cases where there is more than one
dependent, and therefore more exemptions for tax purposes, take-home
pay would be higher and the replacement rate would be lower. 
Appendix II, table II.6 shows the effects of different exemption
assumptions on take-home pay replacement rates.  Because few
beneficiaries were injured and added to the long-term rolls after
they were 65 years old, we did not consider whether additional
exemptions for age or blindness may have applied in computing
take-home pay. 

Itemized deductions.  In our computations of federal and state income
taxes, we used federal and state standard deduction amounts for both
single and married beneficiaries except in cases where state taxes
could have exceeded federal standard deduction amounts.  In these
cases, our computations of federal income taxes used itemized
deductions based on state income tax amounts rather than standard
deduction amounts.  To support our use of the standard deduction for
computing income taxes, we used 1995 Internal Revenue Service (IRS)
information on tax filers who itemized deductions to show that lower
income tax filers generally did not itemize deductions.  Appendix II,
table II.4 shows how the use of different itemized deduction amounts
in computing federal income taxes would reduce take-home pay
replacement rates.  Appendix II, figure II.5 shows the range of
take-home pay replacement rates by amount of pay for single and
married beneficiaries claiming different deduction amounts. 

Spousal income.  We assumed that if a FECA beneficiary received the
dependent benefit, the dependent was a spouse.  For our principal
analyses, we assumed the spouse had no income.  If spouses did have
income, the beneficiaries' effective take-home pay replacement rates
would have been higher.  Examples of the effect of spousal income on
take-home pay replacement rates are shown in appendix II, figure
II.4. 

In estimating federal income taxes for our principal analyses, we
generally computed taxable income by deducting amounts for federal
standard deductions (i.e., $4,150 for a single individual or $6,900
for a couple filing a joint return) and exemptions (i.e., $2,650 for
each exemption) from beneficiaries' gross pay adjusted to 1997 levels
and applied 1997 federal income tax rates.  Because over 25 percent
of the FECA beneficiaries analyzed were single and because the
average age of all beneficiaries analyzed was 61, we did not consider
the effects of earned income tax credits\17 in computing federal
income taxes.  If we had considered this credit for eligible FECA
beneficiaries, effective take-home pay replacement rates would have
been lower. 

In computing take-home pay for FECA beneficiaries who resided in
states with an income tax, we took into account amounts the states
allowed for standard deductions, spousal exemptions, and, where
appropriate, other deductions or tax credits that were based on gross
income in computing state income taxes.  We obtained information on
1997 state income tax rates, exemptions, and standard deductions from
the Research Institute of America's All States Tax Handbook and
individual state's income tax forms and instructions.  The residents
of some states could be subject to county or city income taxes. 
However, we did not attempt to identify and take these types of taxes
into consideration in computing FECA beneficiaries' take-home pay
because it would have been time consuming and expensive to do so.  If
applicable, deductions for these taxes from pay would serve to
increase take-home pay replacement rates. 

Our comparison of FECA benefits with the current value of take-home
pay did not take into consideration beneficiaries' projected salary
growth that might have resulted from merit pay increases or
promotions had they not been injured.  Assumptions about
beneficiaries' potential promotions would have been very speculative. 
Also, other studies we reviewed in developing our replacement rate
methodology did not consider future promotion potential to be a
factor in calculating replacement rates. 


--------------------
\12 A chargeback year is the FECA billing year for which accumulated
benefit outlays are billed to employing agencies for whom injured
employees once worked. 

\13 Schedule awards are benefits for the permanent loss of, or loss
of use of, certain parts or functions of the body.  Benefits are
calculated on the basis of schedules in the law that specify the
number of weeks employees are to receive benefits. 

\14 We categorized one state as a nonincome tax state because,
although it had an income tax, it did not tax income from salaries
and wages. 

\15 We used average pay comparability increases because in some years
pay rate increases were not the same for all employees.  For example,
beginning in 1994 federal workers received different locality-based
comparability adjustments based on disparities between federal and
nonfederal salaries where they worked. 

\16 About 4 percent of the employees were injured before 1966, the
first year in which cost-of-living increases were paid under FECA. 
Because this percentage was relatively small, we did not adjust for
pay increases received before 1966.  If we had considered pre-1966
pay increases, take-home pay replacement rates would have been lower. 

\17 In 1997, tax filers with earned income under $25,760, if there
were one qualifying child, or $29,290, if there were more than one
qualifying child, may have been eligible for an earned income tax
credit that would have either reduced their taxes or enabled them to
receive a cash payment from the federal government.  The amount of
the credit would have depended on the filer's income and could have
amounted to as much as $2,210, if there were one qualifying child, or
$3,656, if there were more than one qualifying child. 


   CAREER PATTERNS
---------------------------------------------------------- Chapter I:2

To obtain information on the career patterns of workers in selected
occupations that were the same as the occupations of FECA
beneficiaries, we first used occupational code data from OWCP's
automated systems to identify the most frequently coded occupations
of FECA beneficiaries.  Usable information on beneficiaries'
occupations was available for only 9,900 of the 30,057 workers we
analyzed.  According to an OWCP analyst, Labor has required agencies
to furnish occupational code information for injured workers since
October 1986.  However, many of the cases that we analyzed were
established before then. 

For the 9,900 FECA beneficiaries for which occupational code
information existed, over 550 different occupations were represented. 
As agreed with your office, we developed career pattern information
for workers in the following occupations--letter carrier, postal
distribution, nurses, and air traffic controllers.  We selected these
occupations because they were either the occupations (1) that were
coded the most frequently or (2) for which many beneficiaries were
likely to be employed by the same agency.  We interviewed officials
from the Postal Service, FAA, and VA who were familiar with the
career patterns of employees in these occupations.  We supplemented
and compared this information with readily available personnel data
on active employees obtained from either these agencies or OPM.  In
addition, for workers in other frequently cited occupations, we
compared aggregate age and pay information from OPM's Central
Personnel Data File\18 with FECA information on beneficiaries with
the same occupations.  Due to time constraints, it was beyond the
scope of our review to analyze the many factors that could be
involved in determining the extent to which beneficiaries' career
progression was affected by their injuries. 


--------------------
\18 The Central Personnel Data File is an automated information
system containing individual records for most federal civilian
employees.  Information on an employee's date of birth, occupation,
and basic pay is included in this system. 


   CHARACTERISTICS OF
   BENEFICIARIES
---------------------------------------------------------- Chapter I:3

To determine beneficiaries' FECA benefit amounts, current ages, ages
when injured, and other characteristics, we relied on data from
OWCP's automated claims management and compensation payment system. 
We developed information on beneficiaries' characteristics for 30,057
beneficiaries--nearly 23,250 beneficiaries for whom we developed
replacement rate information and approximately 6,800 of the remaining
11,460 beneficiaries on the long-term rolls who were receiving FECA
wage-loss compensation benefits of either 66-2/3 or 75 percent of
gross pay. 

We did not verify the information on beneficiaries' characteristics
obtained from OWCP's automated systems.  According to OWCP officials,
they generally believed the information from these systems to be
highly reliable when used in the aggregate.  For purposes of our
analyses, we used the date of injury for computing FECA benefits and
pay at the time of injury.  An OWCP analyst told us that information
on effective dates of some beneficiaries' pay rates may not always be
available or accurate because beneficiaries may have (1) been on and
off the rolls over a period of years or (2) suffered from
occupational diseases rather than traumatic injuries. 

Our work was done in Washington, D.C., between October 1997 and July
1998 in accordance with generally accepted government auditing
standards.  We requested comments on a draft of this report from the
Secretary of Labor.  Labor's comments are summarized at the end of
the letter and are presented in full in appendix IV. 


ADDITIONAL INFORMATION ON FACTORS
AND ASSUMPTIONS AFFECTING
REPLACEMENT RATES
========================================================== Appendix II

The following sections discuss in more detail the factors and
assumptions that influenced the estimated replacement rates presented
in the letter. 


   DATES OF INJURY INFLUENCED
   REPLACEMENT RATES
-------------------------------------------------------- Appendix II:1

More recently injured beneficiaries generally had lower replacement
rates, on average, than those who were injured many years ago.  Over
the years, FECA benefits were increased by cost-of-living allowances
that exceeded general schedule pay comparability increases that
beneficiaries would have received had they not been injured.  Table
II.1 shows average replacement rates based on year of injury and the
number of beneficiaries injured during each period. 



                               Table II.1
                
                   Replacement Rates Based on Date of
                                 Injury

                                         Number of             Average
Dates of injury                      beneficiaries   replacement rates
------------------------------  ------------------  ------------------
After 1989                                   4,408                90.2
1985 to 1989                                 5,674                90.3
1980 to 1984                                 3,669                90.2
1975 to 1979                                 4,204                98.6
1970 to 1974                                 3,170               108.5
Before 1970                                  2,132               105.8
----------------------------------------------------------------------
Source:  GAO analyses of OWCP data. 

Since 1966, FECA cost-of-living and general schedule pay
comparability increases have generally differed in amounts and
effective dates.  Amounts of cost-of-living or pay comparability
increases to which beneficiaries would have been entitled depended on
their dates of injury.  Figure II.1 compares FECA cost-of-living
increases with pay comparability increases for each year from 1966 to
1997. 

   Figure II.1:  Comparison of
   Cost-of-Living and Average Pay
   Comparability Percentage
   Increases (1966 to 1997)

   (See figure in printed
   edition.)

Source:  GAO analysis of OWCP and OPM information. 

Table II.2 shows the cumulative amount of cost-of-living and pay
comparability increases that beneficiaries injured before selected
dates would have received through June 1997. 



                               Table II.2
                
                Cumulative Amounts of Cost-of-Living and
                Pay Comparability Increases for Selected
                 Beneficiaries\a From Date of Injury to
                               June 1997

                                  Increases between January 1970 and
                                              June 1997
                                --------------------------------------
                                 Sum of FECA cost-  Sum of average pay
                                         of-living       comparability
Beneficiaries injured before             increases    increases if not
January                                  (percent)   injured (percent)
------------------------------  ------------------  ------------------
1970                                         143.9               118.7
1975                                         110.0                91.8
1980                                          71.1                62.1
1985                                          40.4                40.2
1990                                          27.3                27.6
1995                                           8.5                 8.0
----------------------------------------------------------------------
\a Cost-of-living increases are provided to injured employees who
stopped work on account of an injury more than 1 year prior to the
effective date of the increase. 

Source:  GAO analyses of OWCP and OPM data. 


   HIGHER PAY AND PROGRESSIVE TAX
   RATES INCREASED REPLACEMENT
   RATES
-------------------------------------------------------- Appendix II:2

Replacement rates vary for workers receiving different amounts of
pay.  Because federal and many state tax rates are progressive,
higher pay levels generally mean higher taxes.  Higher tax rates
reduce take-home pay, thereby increasing replacement rates. 
Conversely, in states with no state income taxes, replacement rates
for beneficiaries with the same income would be lower than they would
be in states with an income tax. 

In 1997, for single individuals, federal income tax rates were 15
percent on taxable income up to $24,650, 28 percent on taxable income
up to $59,750, and 31 percent on taxable income up to $124,650.  For
married individuals filing jointly, federal income tax rates were 15
percent of taxable income up to $41,200, and 28 percent of taxable
income up to $99,600.\19

Table II.3 shows that average replacement rates generally increased
as beneficiaries' pay increased.  Higher pay would generally be
subject to higher income tax rates, which cause an increase in
replacement rates. 



                               Table II.3
                
                Average Take-Home Pay Replacement Rates
                for Beneficiaries at Various Pay Levels
                  Before Their Injury as of June 1997

                                           Number of           Average
Current value of pay                   beneficiaries  replacement rate
----------------------------------  ----------------  ----------------
Under $20,000                                  1,407              90.7
20,000 to 29,999                               7,204              94.9
30,000 to 39,999                               9,293              95.1
40,000 to 49,999                               3,019              95.8
50,000 to 59,999                               1,074             100.1
60,000 to 69,999                                 654             105.7
70,000 to 79,999                                 349             107.4
80,000 and over                                  257             106.5
----------------------------------------------------------------------
Source:  GAO analysis of OWCP data. 

In addition to changes in take-home pay replacement rates related to
progressive federal income tax rates, many FECA beneficiaries lived
in states that also taxed income.  Beneficiaries living in states
with income taxes would have less take-home pay and thus higher
replacement rates.  Of the 23,257 beneficiaries for whom we developed
replacement rate information, about 17,200 lived in 15 states with a
state income tax.  Of these 15 states, 3 had flat tax rates ranging
from 2.8 to 5.95 percent of income, and 12 had progressive tax rates
ranging from 0.5 to 9.3 percent of income.  In addition, one state
had an income tax but did not tax earnings from salaries or wages. 
In computing state income taxes, we considered exemption and standard
deduction amounts allowed by the states in making our calculations. 

Our estimate of the average take-home pay replacement rate for all
beneficiaries for whom we developed information was about 95 percent;
for beneficiaries in states without an income tax, about 94 percent;
and for beneficiaries in states with an income tax about, 96 percent. 
Figure II.2 shows how different state income tax rates would
influence replacement rates for beneficiaries earning various amounts
of pay. 

   Figure II.2:  Influence of
   State Income Tax Rates on
   Replacement Rates

   (See figure in printed
   edition.)

Note:  For these scenarios, in which no specific states' income tax
laws apply, we computed state income taxes based on federal taxable
income amounts.  Using these amounts has the effect of slightly
understating amounts of state taxes paid because federal standard
deduction and exemption amounts were generally higher than amounts
allowed by the states. 

Source:  GAO computations. 

Some beneficiaries lived in areas, such as counties and cities, that
also taxed income.  For our analysis, however, we did not identify or
consider amounts of income taxes paid to local jurisdictions.  Had we
included these taxes, they would have further reduced beneficiaries'
take-home pay and increased replacement rates. 


--------------------
\19 While higher federal income tax rates exist for single and
married taxpayers, these rates would generally not apply to federal
workers unless they had income from sources other than their salaries
or their spouses had taxable income. 


   AUGMENTED BENEFITS INCREASED
   REPLACEMENT RATES
-------------------------------------------------------- Appendix II:3

Under FECA, injured workers with one or more dependents receive
workers' compensation benefits based on 75 percent of their pay
before they were injured compared with benefits of 66-2/3 percent of
pay for beneficiaries without dependents.  Without the additional
dependent's allowance, married injured workers whose spouses did not
work would have lower take-home pay replacement rates than those who
were single because their standard deduction and exemption amounts
would be higher than single beneficiaries and thus their taxes would
be lower.  Figure II.3 shows replacement rates for (1) single
beneficiaries and beneficiaries with dependents based on their
respective benefit levels and various pay amounts received and (2)
beneficiaries with dependents if additional FECA benefits of 8-1/3
percent were not provided. 

   Figure II.3:  Replacement Rates
   for Beneficiaries Under
   Different Dependent and Benefit
   Rate Assumptions

   (See figure in printed
   edition.)

Source:  GAO computations. 


   DIFFERENT INCOME TAX
   ASSUMPTIONS CHANGE REPLACEMENT
   RATES
-------------------------------------------------------- Appendix II:4

In addition to the above discussed factors that affected replacement
rates, the rates we computed would have been different if we had used
different assumptions in calculating federal and state income taxes
for each beneficiary.  For example, replacement rates would have
increased if we had assumed that a beneficiary's spouse had taxable
income.  Replacement rates would have decreased if we had assumed
that the number of exemptions or amounts of itemized deductions
claimed for income tax purposes were greater than the amounts we used
in our calculations. 

Each of these factors and the extent of their effects are discussed
in more detail in the following subsections. 


      SPOUSES' INCOME COULD
      INCREASE REPLACEMENT RATES
------------------------------------------------------ Appendix II:4.1

The presence of a spouse with income could raise the value of
nontaxable workers' compensation benefits because the couple's
combined taxable income had there not been an injury might be subject
to a higher tax rate.  Higher tax rates equate to higher wage
replacement rates.  Pay earned by married workers when they returned
to work after they had been injured would not be accompanied by
additional exemptions or, in most cases, deductions for the couple. 
However, additional taxable wages based on both incomes could be
subject to the same or higher tax rates than the last dollars earned
by the injured worker's spouse.  Compared with single-income couples,
replacement rates for two-income couples are typically higher at both
lower and higher incomes, according to a Workers Compensation
Research Institute study. 

Figure II.4 shows for beneficiaries receiving different amounts of
benefits that the more taxable income a beneficiary's spouse had, the
higher the replacement rate. 

   Figure II.4:  Replacement Rates
   for Beneficiaries With Spouses
   with Different Amounts of
   Taxable Income

   (See figure in printed
   edition.)

Source:  GAO computations. 


      MORE TAX DEDUCTIONS REDUCE
      REPLACEMENT RATES
------------------------------------------------------ Appendix II:4.2

Standard deductions for 1997 federal income tax purposes for single
and joint return filers were $4,150 and $6,900, respectively.  Using
these deduction amounts and our other assumptions, percentages of pay
at the time of injury adjusted to 1997 pay levels replaced by FECA
benefits for single beneficiaries and beneficiaries with dependents
were 92 and 97 percent, respectively.  If itemized deductions were
two or three times higher than the standard deduction amounts we
used, replacement rates would decrease by amounts ranging from about
2 to 7 percent depending on beneficiaries' pay.  Table II.4 shows
examples of changes in replacement rates for single and married
beneficiaries with pay of $30,000 or $60,000 if their itemized
deductions were double or triple the 1997 standard deduction amounts. 



                               Table II.4
                
                  Replacement Rates Based on Different
                           Deduction Amounts

                                 Single   Married    Single   Married
                                --------  --------  --------  --------
                                 Example   Example   Example   Example
                                       1         2         3         4
------------------------------  --------  --------  --------  --------
Income                           $30,000   $30,000   $60,000   $60,000
FECA benefits                     20,000    22,500    40,000    45,000
CSRS contributions                 2,535     2,535     5,070     5,070
Exemption(s)                       2,650     5,300     2,650     5,300
Deductions
 Standard                          4,150     6,900     4,150     6,900
 Double                            8,300    13,800     8,300    13,800
 Triple                           12,450    20,700    12,450    20,700
Federal taxes
 Standard                          3,480     2,670    11,692     8,028
 Double                            2,858     1,635    10,530     6,135
 Triple                            2,235       600     9,368     5,100
Take-home pay
 Standard                         23,985    24,795    43,239    46,902
 Double                           24,608    25,830    44,401    48,795
 Triple                           25,230    26,865    45,563    49,830
FECA as a percentage of take-
 home pay
 Standard                           83.4      90.7      92.5      95.9
 Double                             81.3      87.1      90.1      92.2
 Triple                             79.3      83.8      87.8      90.3
----------------------------------------------------------------------
Source:  GAO computations. 

As shown, replacement rates were highest for married beneficiaries at
the higher income level who claimed standard deductions and lowest
for single beneficiaries at the lower income level whose itemized
deductions were three times the standard deduction amount.  Figure
II.5 shows these differences across different income levels.  Other
single or married beneficiaries whose itemized deductions were two or
three times the standard deduction amounts would have replacement
rates that would fall between these rates. 

   Figure II.5:  Deduction Amounts
   Influence Take-Home Pay
   Replacement Rates

   (See figure in printed
   edition.)

Source:  GAO computations. 

The number of FECA beneficiaries who would itemize their deductions
versus those who would use the standard deduction is unknown. 
According to IRS data on 1995 income tax filers with adjusted gross
incomes between $10,000 and $99,999, of the 50.9 million single
taxpayers, 42.7 million (84 percent) did not itemize deductions.  Of
the 49.0 million taxpayers filing jointly, 25.5 million (52 percent)
did not itemize deductions.  IRS information shows that as income
increases, the percentage of taxpayers itemizing deductions
increases.  While average amounts of itemized deductions increased
with income, these increases were relatively small.  Table II.5
shows, for various income groups, the percentage of returns claiming
itemized deductions and the average amounts of deductions claimed. 
In 1995, standard deduction amounts for single and married
beneficiaries were $3,900 and $6,550, respectively. 



                                    Table II.5
                     
                     Percentages of Returns Claiming Itemized
                         Deductions and Average Amount of
                        Deductions Claimed for Single and
                           Married Income Groups (1995)

                        Single                             Married
          ----------------------------------  ----------------------------------
             Percentage of                       Percentage of
Adjusted      returns with    Average amount      returns with    Average amount
gross             itemized       of itemized          itemized       of itemized
income\a        deductions        deductions        deductions        deductions
--------  ----------------  ----------------  ----------------  ----------------
$10,000                8.6            $9,650              11.4           $11,435
 to
 $19,999
20,000                18.1             8,332              20.0            11,741
 to
 29,999
30,000                35.6             9,282              30.7            11,536
 to
 39,999
40,000                50.7            10,468              47.2            11,954
 to
 49,999
50,000                72.1            12,491              69.1            13,247
 to
 74,999
75,000                83.8            17,970              86.8            16,918
 to
 99,999
--------------------------------------------------------------------------------
\a Data provided for filers with incomes over $10,000 or under
$100,000 because the pay for most FECA beneficiaries would be within
this range. 

Source:  GAO analysis of 1995 IRS data. 

In general, if FECA beneficiaries were similar to all individuals
filing income tax returns in 1995, FECA beneficiaries with more pay
at the time of injury would be more likely to claim itemized
deductions in excess of standard deduction amounts than would those
with lower pay.  In such cases, our replacement rates would be
overstated, particularly for beneficiaries at higher income levels. 
Likewise, if beneficiaries' deductions were equal to or greater than
their income (thereby owing no tax), the replacement rate for single
and married beneficiaries would be about 73 and 82 percent,
respectively, because the relationship between take-home pay (gross
pay less retirement and Medicaid contributions) and FECA benefits
would always be the same. 


      MORE EXEMPTIONS/
      DEPENDENTS REDUCE
      REPLACEMENT RATES
------------------------------------------------------ Appendix II:4.3

FECA beneficiaries are entitled to augmented benefits if they have
one or more dependents.  However, information on the specific number
of dependents claimed by each beneficiary is not available from FECA
automated data.  Beneficiaries receiving the dependent benefit
allowance who were eligible to claim more than the two we assumed in
our income tax calculations would have lower take-home pay
replacement rates than those shown in our analyses. 

In 1997, taxpayers were allowed to reduce their taxable income by
$2,650 for each exemption claimed on their tax return.  Table II.6
shows examples of how increases in the number of exemptions would
decrease replacement rates. 



                               Table II.6
                
                 Number of Exemptions Influences Take-
                       Home Pay Replacement Rates

                        $30,00  $30,00  $30,00  $60,00  $60,00  $60,00
Income                       0       0       0       0       0       0
----------------------  ------  ------  ------  ------  ------  ------
Number of exemptions         2       3       4       2       3       4
Exemption amount        $5,300  $7,950  $10,60  $5,300  $7,950  $10,60
                                             0                       0
Take-home pay           24,795  25,192  25,590  46,902  47,644  48,386
Take-home pay             90.7    89.3    87.9    95.9    94.5    93.0
 replacement rate
----------------------------------------------------------------------
Source:  GAO. 

In general, each additional exemption decreased the replacement rate
by about 1.5 percent. 


      OTHER FACTORS INFLUENCE
      REPLACEMENT RATES
------------------------------------------------------ Appendix II:4.4

In addition to the above factors, workers' actual take-home pay could
be affected by other deductions that we did not consider in our
calculations of FECA take-home pay replacement rates because
employees have a choice of whether to have their take-home pay
reduced by their share of the cost of fringe and other benefits to
which they may be entitled. 

Examples of the deductions not included in our calculations of
take-home pay were employees' thrift savings plan contributions,
allotments for U.S.  savings bonds, and deductions for health, life,
or disability insurance.  Typically, these deductions are
discretionary.  In the case of health and life insurance, injured
workers are eligible to participate in these federal programs and
could have payments for these types of insurance withheld from their
workers' compensation benefits. 


COMPARISON OF FECA AND OPM
INFORMATION FOR INDIVIDUALS IN THE
SAME OCCUPATIONS
========================================================= Appendix III

Occupati                                             OPM information (September
  ons           FECA information (June 1997)                   1997)
--------  ----------------------------------------  ----------------------------
                                         Estimated
                                     current value
                    Average age at  of pay at time             Average   Average
            Number  time of injury       of injury    Number       age       pay
--------  --------  --------------  --------------  --------  --------  --------
Pipefitt       274            43.8         $39,251     3,935      47.1   $38,569
 er
Laborer        234            43.5          26,555     4,380      39.7    22,494
Secretar       210            45.3          26,741    63,453      44.7    28,301
 y
Material       164            43.6          29,058    15,347      46.5    30,208
 handler
Clerk/         161            46.4          25,303    60,100      43.3    26,621
 assista
 nt
Electric       156            46.2          39,275     2,048      46.6    42,643
 ian
Motor          144            47.3          31,266     6,943      48.5    30,983
 vehicle
 operato
 r
Carpente       140            44.7          34,196     2,721      48.6    35,637
 r
Miscella       128            40.0          28,648       175      43.0    37,081
 neous
 occupat
 ions
Custodia       125            47.0          24,780    11,306      47.7    22,998
 l
 worker
Maintena       125            45.7          34,538    11,271      47.5    35,291
 nce
 mechani
 c
--------------------------------------------------------------------------------
Source:  GAO analysis of FECA and OPM data. 




(See figure in printed edition.)Appendix IV
COMMENTS FROM THE DEPARTMENT OF
LABOR
========================================================= Appendix III



(See figure in printed edition.)


The following are comments on the Department of Labor's letter dated
July 20, 1998. 

GAO COMMENTS

1.  Labor said we made an underlying but unstated assumption that the
state where beneficiaries currently resided was the state where they
lived when injured.  Labor added that about 26 percent of the
beneficiaries did not live in the state in which they were injured,
but the report did not state how many individuals or what income
groups moved from states with an income tax to states without an
income tax. 

Regarding Labor's comment that we made an unstated assumption about
beneficiaries residing in states where they lived when they were
injured, we did not need to make such an assumption.  Although our
profile information showed that about 26 percent of the beneficiaries
currently lived in states that were different from the ones in which
they were injured, any differences between beneficiaries' states of
residence at the time of injury and their current residences were not
relevant to our computation of beneficiaries' current take-home pay
replacement rates. 

2.  Labor suggested that our estimates of take-home pay replaced by
FECA benefits be further qualified by adding language stating that we
assumed beneficiaries had not received any promotions from the time
of injury through the present.  Labor said that it was almost certain
that some percentage of injured workers would have received
promotions, thus lowering the replacement rate. 

Labor is in effect saying that for at least some workers the
take-home pay replacement rates we developed were overstated because
our estimated replacement rates were based on pay at the time of
injury adjusted to 1997 pay levels and did not take into
consideration the possibility that some workers, had they not been
injured, would have received promotions.  Higher pay rates reflecting
assumed promotions, if compared to compensation benefits based on pay
at the time of injury, would result in lower replacement rates. 

While the subject of forgone promotions may be relevant to assessing
the effects of work-related injuries on individuals' careers, neither
we nor the workers' compensation studies we reviewed in developing
our replacement rate methodology considered future promotion
potential to be a factor in calculating replacement rates.  In
addition, although some employees may have been promoted had they not
been injured, an assumption by us on which employees would have
received one or more promotions would be very speculative. 
Therefore, we did not consider it necessary to further qualify our
estimates of take-home pay replaced by FECA benefits.  We have
revised our scope and methodology to note the reasons why we did not
make an assumption regarding forgone promotions and merit pay
increases. 

3.  In commenting on table II.4, Labor said that our estimated
replacement rate of 95.9 percent for a married beneficiary who was
paid $60,000 and who took the standard deduction would not be
reflective of the norm for that group because higher income
individuals tend to itemize deductions.  Likewise, Labor noted that
our replacement rate of 79.3 percent based on a single person who was
paid $20,000 and whose itemized deductions were three times the
standard deduction amount would not be reflective of the norm for
that group because most single people with pay of $20,000 would not
be itemizing deductions.  We did not intend the information in table
II.4 to be reflective of norms for those groups of individuals. 
Rather, we provided these hypothetical examples to show the
sensitivity of our replacement rate analyses to different assumptions
about individual beneficiaries' standard or itemized deductions. 


MAJOR CONTRIBUTORS TO THIS REPORT
=========================================================== Appendix V

GENERAL GOVERNMENT DIVISION

Larry H.  Endy, Assistant Director
Edward R.  Tasca, Evaluator-in-Charge
Gregory H.  Wilmoth, Supervisory Social Science Analyst
George H.  Quinn, Jr., Computer Specialist

In addition to those named above, the following individuals from the
General Government Division made important contributions to this
report:  Wayne Barrett, Senior Evaluator; Cathy Hurley, Senior
Computer Specialist; Kim Wheeler, Graphics; and Ernestine Burt, Issue
Area Assistant. 


*** End of document. ***