This is the accessible text file for GAO report number GAO-08-284 
entitled 'Federal Workers' Compensation Better Data and Management 
Strategies Would Strengthen Efforts to Prevent and Address Improper 
Payments' which was released on February 28, 2008. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Committees: 

United States Government Accountability Office: 

GAO: 

February 2008: 

Federal Workers' Compensation: 

Better Data and Management Strategies Would Strengthen Efforts to 
Prevent and Address Improper Payments: 

GAO-08-284: 

GAO Highlights: 

Highlights of GAO-08-284, a report to congressional committees 

Why GAO Did This Study: 

In fiscal year 2006, the Federal Employees’ Compensation Act (FECA) 
program paid over $1.8 billion in wage loss compensation to federal 
employees who were unable to work after being injured on the job. Under 
the Comptroller General’s authority to conduct evaluations on his own 
initiative, GAO examined (1) how effectively the Department of Labor’s 
(Labor) Office of Workers’ Compensation Programs (OWCP) manages the 
risk of improper FECA compensation payments; (2) what vulnerabilities 
to improper payments, if any, exist in OWCP’s procedures for making 
FECA wage loss payments; and (3) how well OWCP ensures the recovery of 
identified FECA overpayments. To address these issues, GAO reviewed 
OWCP documents, analyzed data obtained from OWCP, reviewed a random and 
projectable sample of FECA claims files, visited five OWCP district 
offices, and interviewed OWCP headquarters and district officials. 

What GAO Found: 

OWCP has not established an effective strategy for managing improper 
payments in the FECA program. The agency does not sufficiently 
emphasize preventing, detecting, and recovering improper payments. None 
of the performance goals for the program addresses improper payments. 
Further, OWCP does not collect the information it needs to accurately 
assess the FECA program’s risk of improper payments, such as 
information on their magnitude and causes. Without such data, it cannot 
focus on the most vulnerable areas. 

The FECA program is vulnerable to improper payments for several 
reasons. First, OWCP relies on unverified, self-reported information 
from claimants that is not always timely or correct. From a review of a 
sample of claims files for overpayments identified by OWCP in 2006, GAO 
found that many occurred because claimants did not inform OWCP in a 
timely manner when they returned to work. Further, because OWCP 
generally does not require claimants’ self-reported earnings to be 
verified and does not systemically match its data on FECA claimants 
with earnings data from other federal agencies, it may fail to identify 
cases of unreported earnings. An obstacle to conducting such matches, 
however, is that OWCP does not have the legal authority to access the 
database maintained by another federal agency with the most current 
earnings data. In addition, from GAO’s file reviews, GAO found that 
both overpayments and underpayments were caused by OWCP errors and that 
many overpayments occurred when OWCP’s payment-processing deadlines 
prevented payments from being quickly canceled when claimants returned 
to work or died. 

Figure: Estimated Causes of Overpayments Identified by OWCP in 2006: 

This figure is a pie chart showing estimated causes of overpayments 
identified by OWCP in 2006. 

OWCP claims examiner errors: 32%; 
Payment system limitations: 26%; 
Untimely notification: 17%; 
Miscellaneous: 17%; 
Incorrect/unverified information: 8%. 

[See PDF for image] 

Source: GAO analysis of a sample of FECA claims files. 

Note: We could not make similar estimated for OWCP's 2006 underpayments 
due to data limitations. 

[End of figure] 

What GAO Recommends: 

The Secretary of Labor should direct OWCP to, among other things, 
develop a strategy to ensure that the agency’s efforts to prevent and 
monitor improper payments are properly balanced with its other 
priorities, take steps to reduce the most common causes of improper 
payments, and focus more attention on the recovery of overpayments. In 
its comments, Labor disagreed with many of GAO’s findings and 
conclusions, but described several actions being taken by OWCP that are 
consistent with the recommendations in the report. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.GAO-08-284]. For more information, contact 
Daniel Bertoni at (202) 512-7215 or bertonid@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

OWCP Lacks an Effective Strategy for Managing the Risks of Improper 
FECA Compensation Payments: 

OWCP's Dependence on Unverified Information, Internal Errors, and 
System Limitations Leaves FECA Vulnerable to Improper Payments: 

OWCP Does Not Ensure the Recovery of FECA Overpayments: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: cope and Methodology: 

Appendix II: Comments from the Department of Labor: 

GAO's Response to Labor's Comments: 

Appendix III: GAO Contact and Acknowledgments: 

Tables: 

Table 1: National and District Office Performance Goals for the FECA 
Program: 

Table 2: Initial Balances of Compensation-Related Overpayments 
Identified in Fiscal Year 2006: 

Table 3: Distribution of the Sample of Improper Payments Identified in 
Fiscal Year 2006 by the Year in which the Payment Was Made: 

Figures: 

Figure 1: FECA Claims Process: 

Figure 2: FECA Overpayment Process: 

Figure 3: Estimated Percentage of Fiscal Year 2006 Overpayments by 
Cause: 

Figure 4: Status of Overpayments Listed on OWCP's Debt-Aging Report, 
September 2007: 

Figure 5: Estimated Recovery Status of Overpayments That OWCP Waived or 
Pursued in Fiscal Year 2006: 

Figure 6: Estimated Recovery Sources for 2006 Overpayments That Were 
Repaid or Being Collected: 

Figure 7: Initial Balances of All Wage Loss Compensation Overpayments 
Identified in Fiscal Year 2006 and in Final Determination or Terminated 
Status, as of March 13, 2007: 

Abbreviations: 

DFEC: Division of Federal Employees' Compensation: 

DOD: Department of Defense: 

FECA: Federal Employees' Compensation Act: 

GS: General Schedule: 

HHS: Department of Health and Human Services: 

IPIA: Improper Payments Information Act of 2002: 

Labor: Department of Labor: 

OIG: Office of Inspector General: 

OMB: Office of Management and Budget: 

OPM: Office of Personnel Management: 

OWCP: Office of Workers' Compensation Programs: 

SSA: Social Security Administration: 

SSN: Social Security Number: 

Treasury: Department of the Treasury: 

USPS: United States Postal Service: 

VA: Department of Veterans Affairs: 

United States Government Accountability Office: 

Washington, DC 20548: 

February 26, 2008: 

Congressional Committees: 

In fiscal year 2006, the Federal Employees' Compensation Act (FECA) 
program paid over $1.8 billion in wage loss compensation to federal 
employees who were unable to work because of injuries sustained while 
performing their federal duties. Administered by the Department of 
Labor's (Labor) Office of Workers' Compensation Programs (OWCP), FECA 
covers over 2.7 million civilian federal employees in more than 70 
different agencies, such as the U.S. Postal Service (USPS) and the 
Department of Homeland Security. OWCP bills the agencies that employ 
the injured workers for these wage loss compensation costs. The 
Improper Payments Information Act of 2002 defines an improper payment 
as any payment that should not have been made or was made in the wrong 
amount (including both overpayments and underpayments).[Footnote 1] 
According to Labor, the FECA program experienced a low rate of improper 
payments in fiscal year 2006--0.04 percent. However, recent reports 
from several federal agencies' Offices of Inspectors General have found 
weaknesses in OWCP's internal controls that suggest the actual 
percentage of improper payments may be much higher.[Footnote 2] 

We addressed the following questions under the Comptroller General's 
authority to conduct evaluations on his own initiative as part of a 
continued effort to assist Congress in assessing OWCP's strategies for 
preventing, detecting, and recovering improper payments: (1) How 
effectively does OWCP manage the risks of improper compensation 
payments? (2) What vulnerabilities to improper payments, if any, exist 
in OWCP's procedures for making wage-loss-compensation payments under 
FECA? (3) How well does OWCP ensure the recovery of identified FECA 
overpayments? 

To respond to these questions, we reviewed Labor's annual performance 
and accountability reports, the FECA procedures manual and internal 
controls, OWCP's accountability reviews, relevant agencies' Office of 
Inspector General reports, and applicable laws and regulations as well 
as interviewed officials at OWCP headquarters. We also reviewed the 
methodology used by Labor to estimate its risk of improper FECA 
payments and interviewed audit contractor staff who were involved in 
developing these estimates. In addition, we requested data from OWCP on 
the magnitude and causes of overpayments and underpayments. OWCP did 
not have data on causes or the number of improper payments that 
occurred in specific years, but it created a unique report for us that 
included all debts, including overpayments, that OWCP identified in 
fiscal year 2006. To assess the reliability of these data, we (1) 
reviewed existing documentation related to the data sources; (2) 
electronically tested the data to identify obvious problems with 
completeness or accuracy, such as missing or inconsistent data; and (3) 
interviewed knowledgeable agency officials about the data. We 
determined that the data were sufficiently reliable for the purposes of 
this report. We removed all debts that were not in final determination 
or terminated status and that were not related to wage-loss- 
compensation payments. We analyzed these data and reviewed the claims 
files for a random, projectable sample of these overpayments, as well 
as the 10 largest, to confirm whether they represented improper 
payments and to determine their causes and final outcomes. Because we 
discovered after reviewing these data that some of the debts included 
in OWCP's report either (1) were not improper payments or (2) were 
identified prior to fiscal year 2006, we excluded these debts from our 
analysis. We used the results from our file review to estimate the 
total dollar amount of improper overpayments identified by OWCP in 
2006, and the percentage that were attributable to different causes. 
Although OWCP could not identify underpayments, it identified a subset 
of payments that included underpayments. We reviewed the claims files 
for a random sample of these payments for fiscal year 2006 to identify 
underpayments, estimate their dollar value, and obtain general 
information on their causes. However, because of limitations in the 
data, we could not develop estimates of the percentage of underpayments 
attributable to different causes for all 2006 underpayments. Estimates 
based on our claims file reviews are accurate to within plus or minus 
10 percentage points at the 95 percent confidence level, unless 
otherwise noted. We also conducted site visits at 5 of OWCP's 12 
district offices: Boston, Cleveland, Dallas, San Francisco, and 
Washington, D.C. We selected these offices based on variation in office 
size, internal audit results, organizational structure, and geographic 
location. Finally, we interviewed officials responsible for managing 
the FECA program at 10 federal agencies with varying FECA caseload 
sizes. We conducted our work between September 2006 and January 2008 in 
accordance with generally accepted government auditing standards. See 
appendix I for more detailed information on our scope and methodology. 

Results in Brief: 

OWCP lacks an effective strategy for managing the risks of improper 
payments because it has not (1) emphasized preventing, detecting, and 
recovering improper payments or (2) collected the information needed to 
assess the program's risk of improper payments. None of the agency's 
performance goals for the FECA program addresses improper payments. 
Instead, they emphasize the timely processing of claims and quickly 
returning claimants to work. While these are important goals, previous 
GAO work has shown that the risk of improper payments increases when 
agencies' goals and performance measures do not strike an appropriate 
balance between service delivery and the need to ensure payment 
accuracy. Further, OWCP program staff reported that detecting and 
recovering improper payments are often lower priorities than processing 
claims quickly. In addition, OWCP lacks useful information on the 
magnitude of improper payments or their causes, making it difficult to 
identify vulnerabilities that lead to payment errors or determine their 
impact on program operations. While Labor estimated that the FECA 
program made $703,000 in improper payments in fiscal year 2006 by 
reviewing a sample of all payments made during the year, this estimate 
provides OWCP with limited information to use in identifying and 
managing the FECA program's risk of improper payments. For example, 
this estimate does not capture all types of improper payments and it 
does not include the improper payments that OWCP identified during the 
year, which we estimated to be $13.3 million for 2006--$7.1 million in 
overpayments and $6.2 million in underpayments. Without comprehensive 
information on risks, OWCP may not be taking all of the precautions 
necessary to focus on its most vulnerable areas. 

The FECA program is also vulnerable to improper payments because OWCP 
relies on self-reported eligibility information from claimants without 
verifying it, makes internal payment errors, cannot stop certain 
payments that OWCP knows to be in error, and receives inaccurate wage 
and benefits information from claimants' employing agencies. OWCP 
relies on claimants to inform it when they return to work, but our 
review of a sample of overpayments identified by OWCP in fiscal year 
2006 found that an estimated 11 percent of overpayments occurred 
because claimants did not notify OWCP of their return to work in a 
timely manner. Further, because OWCP does not generally require claims 
examiners to verify claimants' self-reported earnings statements and 
does not conduct systematic data matches with the Social Security 
Administration's (SSA) wage records, it may fail to identify cases of 
unreported outside earnings: a recent report by SSA's Office of 
Inspector General found that nearly 7 percent of claimants OWCP found 
to have no wage-earning capacity in 2004 actually had earnings that 
were reported to SSA. In addition, we estimated that 6 percent of the 
overpayments occurred when OWCP was not notified immediately after 
claimants died. Despite the fact that OWCP conducts monthly data 
matches with SSA's death records, we found instances in which the 
agency continued to send FECA payments to a claimant or a claimant's 
survivor for more than a year after the individual died. We also found 
that both overpayments and underpayments were caused by OWCP errors, 
such as when claims examiners made calculation errors, did not take 
timely action to stop payments after being notified that a claimant had 
returned to work or died, and or incorrectly reduced FECA payments. In 
one instance, OWCP paid a claimant for nearly 13 years after he 
returned to work, despite numerous notifications of the error by the 
claimant. In addition, about 26 percent of overpayments OWCP identified 
in 2006 occurred because limitations in its payment systems prevented 
it from quickly canceling payments when eligibility changes occurred, 
such as when claimants returned to work. In other instances, both 
underpayments and overpayments occurred when claimants' employing 
agencies provided inaccurate wage and benefits data to OWCP. 

OWCP does not sufficiently ensure the recovery of FECA overpayments-- 
specifically, it does not always process overpayments in a timely 
manner and misses opportunities for recovering them. Our analyses of 
OWCP data confirmed that overpayments are not always processed within 
OWCP's required 60-day time frame. For example, almost half of the 
identified overpayments listed in OWCP's September 2007 debt report 
were over 6 months old, but OWCP had not yet notified the claimants of 
the overpayments. Claims examiners in several district offices we 
visited told us that they sometimes delayed processing and recovering 
identified overpayments to focus on other tasks, such as paying initial 
claims quickly. OWCP cannot recover overpayments until the required 
overpayment notices have been issued and past GAO work suggests that 
overpayments are less likely to be repaid if they are not confirmed and 
processed promptly. While OWCP does not track the recovery status of 
overpayments, based on our review of 2006 overpayments that OWCP waived 
or pursued, we estimated that about 71 percent were repaid or were in 
the process of being collected. However, although many of them were 
collected, claims examiners missed opportunities to recover other 
overpayments. In one case, for example, a claims examiner made a 
$29,000 payment to a claimant while a $10,000 overpayment that had been 
discovered 12 months earlier was still pending. Claims examiners can 
miss such opportunities when their focus on paying initial claims is 
not adequately balanced with an emphasis on recovering overpayments, 
when they are not completely familiar with OWCP's recovery processes, 
or when OWCP's data system does not alert them to recover overpayments 
from claimants' subsequent FECA payments. 

We are making several recommendations to the Secretary of Labor to help 
OWCP strengthen its efforts to prevent and address improper FECA 
payments. Specifically, we recommend that the Secretary of Labor direct 
OWCP to (1) develop a management strategy to ensure that preventing and 
monitoring improper payments is properly balanced with the need to 
quickly process and pay claims; (2) take specific steps to reduce the 
most common causes of improper payments; (3) develop a legislative 
proposal to obtain the legal authority to enter into a data-matching 
agreement with the Department of Health and Human Services in order to 
identify individuals who are receiving FECA payments and have earnings 
reported in the National Directory of New Hires, and (4) focus more 
attention on the recovery of FECA overpayments. In its comments, Labor 
disagreed with many of GAO's findings and conclusions; however, the 
agency described several actions being taken by OWCP that are 
consistent with the recommendations in the report. 

Background: 

The FECA program provides wage loss compensation and payments for 
medical treatment to federal employees who are injured in the 
performance of their federal duties.[Footnote 3] During fiscal year 
2006, OWCP made over $1.8 billion in wage-loss-compensation payments to 
injured federal employees ("claimants") and processed approximately 
20,000 new wage loss claims. At the end of fiscal year 2006, over 
55,000 claimants were receiving regular monthly wage-loss-compensation 
payments from OWCP. 

Federal agencies use their own annual appropriations to reimburse Labor 
for wage-loss-compensation payments made to their employees each year, 
while most of the program's administrative costs are covered by direct 
appropriations from the Congress. For fiscal year 2008, Labor requested 
that the Congress provide $93.4 million in administrative funding for 
the FECA program and sought an additional $52.3 million from certain 
federal agencies for administrative purposes.[Footnote 4] In total, 
this funding would provide 895 full-time equivalent positions for the 
FECA program. 

USPS pays more in FECA compensation than any other federal agency. In 
2004, USPS paid approximately $852 million in wage loss compensation. 
During this same period, the Departments of Navy and Army paid the 
second and third highest amounts in FECA wage loss compensation to 
injured civilian employees of their agencies, approximately $245 and 
$177 million dollars, respectively. 

Claims Management: 

Claims examiners at OWCP's 12 district offices determine applicants' 
eligibility for FECA benefits and process claims for wage loss 
payments. FECA divides work-related injuries into two categories: 
"traumatic injuries" and "occupational illnesses or diseases." 
Traumatic injuries are wounds or other conditions that occur within a 
single day or work shift, such as when an employee slips at work and 
sprains his ankle. An occupational illness or disease is a physical 
condition produced by the work environment over a period longer than 
one workday or shift, such as carpal tunnel syndrome. In this report, 
we use the term "injuries" to refer to both workers who have sustained 
traumatic injuries and workers who have experienced an occupational 
illness or disease. 

FECA regulations specify complex criteria for computing compensation 
payments. Using information provided by the employing agency and the 
claimant on a claims form, OWCP calculates compensation based on a 
number of factors, including the claimant's rate of pay, deductions for 
health and life insurance benefits, the claimant's marital status, and 
whether or not the claimant has dependents. In addition, claimants 
cannot receive FECA benefits at the same time they receive certain 
other federal disability or retirement benefits. For example, claimants 
cannot receive both FECA wage-loss-compensation payments and disability 
payments from the Department of Veterans Affairs (VA) for the same 
injury. Further, claimants cannot receive federal retirement benefits 
paid through the Office of Personnel Management (OPM) concurrently with 
FECA benefits and must elect to receive one or the other.[Footnote 5] 
However, a claimant can receive both FECA and SSA retirement benefits, 
although the claimant's FECA wage-loss-compensation payments should be 
reduced by the amount of SSA retirement benefits attributable to 
federal service.[Footnote 6] Similarly, a claimant can receive both 
FECA and SSA disability benefits, although SSA is required to reduce 
the level of disability benefits it pays by the amount of FECA wage 
loss compensation received by the claimant. Figure 1 details the 
process for filing and calculating claims for wage loss compensation 
under the FECA program. 

Figure 1: FECA Claims Process: 

This figure is a flowchart of the FECA claims process.

[See PDF for image] 

Source: GAO analysis; Art Explosion. 

[End of figure] 

Based on various eligibility factors, the amount of wage loss 
compensation OWCP pays claimants varies widely. For example, a claimant 
who earned $2,500 per month ($30,000 a year), paid $67 a month for 
health insurance benefits, and had no dependents would receive 
approximately $1,600 a month in FECA wage loss compensation. A married 
claimant who earned $8,500 per month ($102,000 a year) and paid $150 a 
month for health insurance benefits would receive approximately $6,050 
a month in FECA compensation.[Footnote 7] 

Compensation payments are issued on a monthly or weekly basis. 
Claimants who are expected to experience wage loss for longer than 3 
months receive automatic monthly payments as long as their eligibility 
for wage loss compensation continues. Alternatively, claimants who are 
expected to recover more quickly and return to work within 3 months are 
required to file new claims forms each payment cycle in order to prove 
that they were off work and receive manually generated 
payments.[Footnote 8] OWCP provides wage loss compensation until 
claimants can return to work in either their original positions or 
other suitable positions that meet medical work restrictions. If 
claimants return to work but do not receive wages equal to that of 
their prior positions--such as claimants who return to work part-time-
-FECA benefits cover the difference between their current and previous 
salaries.[Footnote 9] 

FECA regulations require claims examiners to verify annually that 
claimants who are receiving automatic monthly compensation payments 
remain eligible for compensation. This verification process relies 
almost entirely on information provided by claimants on a form that 
OWCP mails them each year. Claims examiners are responsible for 
following up and taking necessary action to ensure that the forms are 
completed and returned, and can suspend compensation payments if a 
claimant fails to submit the form within the specified time period. 
Once returned, claims examiners review the forms for indications that a 
claimant's eligibility has changed and adjust the compensation payments 
accordingly. For example, if a married claimant indicates that he 
divorced his wife and does not have any other dependents, the claims 
examiner should reduce his wage-loss-compensation payment from three- 
quarters of his salary to two-thirds of his salary. 

Improper Payments: 

The Improper Payment Information Act enacted in 2002 requires the heads 
of federal agencies to annually review all programs and activities they 
administer, identify those that may be susceptible to significant 
improper payments, and estimate and report the annual amount of 
improper payments in those programs and activities. The Office of 
Management and Budget (OMB) defines significant improper payments as 
payments in any program that exceed both 2.5 percent of total payments 
and $10 million annually. In addition, OMB has previously identified 
other programs, including the FECA program,[Footnote 10] as being at a 
high risk of improper payments because its total payments exceed $2 
billion annually.[Footnote 11] Because of this high risk designation, 
Labor must annually estimate the improper payment rate for the FECA 
program and report this rate in its Performance and Accountability 
Report, as well as identify the causes of improper payments and report 
the corrective actions it plans to take to address them. 

Overpayment Recovery: 

Once an overpayment has been identified by OWCP, the actions taken to 
recover it depend on the amount of the overpayment. Because of the 
administrative costs associated with recovering overpayments, OWCP 
allows claims examiners to waive overpayments less than $200 without 
taking any action to recover them. For overpayments $200 or greater, 
claims examiners must send a notice of the overpayment (called a 
"preliminary notice") to the claimant within 30 days. The claimant then 
has 30 days to respond to the preliminary notice and contest the 
overpayment. If a claimant does not repay the overpayment or appeal the 
overpayment decision within 30 days of the preliminary notice, claims 
examiners must issue an additional notice (called the "final notice") 
that provides a recovery strategy to the claimant, such as a suggested 
repayment schedule. As shown in figure 2, the recovery options 
available to claims examiners depend on whether the claimant continues 
to receive FECA or other federal compensation payments, the amount of 
the overpayment, and whether the claimant was found to be at fault in 
the creation of the overpayment. If the debt is delinquent for 180 days 
and the claimant has not responded to two additional letters from OWCP 
demanding repayment, the claims examiner is required to refer the debt 
to the Department of the Treasury (Treasury) for recovery. 

Figure 2: FECA Overpayment Process: 

This figure is a flowchart of FECA overpayment process. 

[See PDF for image] 

Source: GAO analysis. 

[End of figure] 

OWCP Lacks an Effective Strategy for Managing the Risks of Improper 
FECA Compensation Payments: 

OWCP has not established an effective strategy for managing improper 
FECA payments. The agency does not sufficiently emphasize preventing, 
detecting, and recovering improper payments. Program staff reported 
that they focused on the aspects of the claims process that are 
regularly tracked and measured by managers and said preventing, 
detecting, and recovering improper payments are often lower priorities. 
In addition, OWCP lacks the data needed to accurately assess the 
program's risk of improper payments. The agency does not collect data 
on the magnitude or causes of improper payments, making it difficult to 
identify vulnerabilities that lead to payment errors, implement 
procedures to prevent them, or evaluate their effectiveness. 

OWCP Does Not Emphasize Preventing, Detecting, or Recovering Improper 
Payments: 

For the past 5 years, none of the national goals established for the 
FECA program have addressed improper payments but have focused 
primarily on improving service delivery. Two of the five national goals 
set time frames for returning claimants to work; two focus on 
minimizing medical and compensation costs, and the fifth goal addresses 
improving customer service by quickly responding to claimant and agency 
inquiries. Past GAO work has shown that emphasizing the prevention, 
detection, and recovery of improper payments at the managerial level by 
establishing goals for reducing improper payments is a key aspect of an 
effective strategy for managing improper payments.[Footnote 12] 

Similarly to the national goals, OWCP's performance goals for its 
district offices emphasize timely case management, but do not focus on 
payment accuracy. None of the 21 performance goals for district offices 
contained in OWCP's fiscal year 2007 operational plan address 
preventing or detecting improper payments, although one focuses on 
recovering delinquent overpayments. As table 1 illustrates, the 
majority of OWCP's performance goals for district offices either 
support the national goals for the program or focus on quickly 
adjudicating claims and processing payments. While one goal requires 
district offices to quickly process wage-loss-compensation payments, 
OWCP has not established a corresponding goal to ensure that these 
payments are accurate. Although quickly processing claims and payments 
are important goals, previous GAO work has shown that the risk of 
improper payments increases when these goals are not balanced with an 
emphasis on payment accuracy.[Footnote 13] In commenting on this 
report, Labor reported that it has included a new measure on the 
timeliness of processing overpayments in its 2008 operational plan for 
the FECA program. 

Table 1: National and District Office Performance Goals for the FECA 
Program: 

Strategic goal: National goals for the FECA program for fiscal years 
2003-2008 (targets listed are for fiscal year 2007); 
Reduce the consequences of work-related injuries: Return claimants to 
work quickly: 1. For USPS claimants, achieve a lost-production-days 
rate of 129.8 days; 2. For claimants from all other agencies, achieve a 
lost-production-days rate of 49 days; 
Reduce the consequences of work-related injuries: Minimize compensation 
and medical costs; 3. Produce $8 million in savings by returning 
claimants receiving long- term wage loss compensation to work; 4. Keep 
the inflationary trend in FECA medical costs below the nationwide 
trend; 
Reduce the consequences of work-related injuries: Improve customer 
service: 5. Achieve targets for 4 of 5 communications performance 
areas, such as reducing average call response times. 

Reduce the consequences of work-related injuries: Reduce the 
consequences of work-related injuries: Reduce the consequences of work-
related injuries: 

Strategic goal: District office workload and performance goals; (21 
goals established for fiscal year 2007); Reduce the consequences of 
work-related injuries: 3 goals for quickly returning claimants to work: 
1. For claimants who have received wage loss compensation for 12 
months, return them to work within 156 days, on average; 2. Of the 
claimants assigned nurses to aid in their recovery, return 6,440 to 
work; 3. For claimants unable to return to the jobs they held when 
injured, return 550 to work by training them for new positions
Reduce the consequences of work-related injuries: 1 goal for 
compensation cost savings: 1. Achieve resolutions (such as returning 
claimants to work) in 2,342 claims for long-term wage loss 
compensation; 
Reduce the consequences of work-related injuries: 5 goals for providing 
quality and timely customer service: 
1. Authorize 95% of medical referrals within 3 days; 
2. Review 95% of incoming mail within 3 days; 
3. Respond to 90% of priority written inquiries within 14 days; 
4. Respond to 90% of general written inquiries within 30 days; 
5. Improve communications efforts, such as: 
--Keep callers waiting for 3 min; 
--Respond to 65% of telephone calls the same day; 
Reduce the consequences of work-related injuries: 11 goals for 
accepting claims and processing benefits quickly, including: 
--Process 90% of claims for traumatic injuries within 45 days; 
--Process 85% of claims for basic occupational illnesses within 90 
days; 
--Process 85% of wage- loss claims for payments within 14 days; 
Reduce the consequences of work-related injuries: 1 goal focused on 
collecting delinquent overpayments: 
1. Resolve or refer to Treasury for collection 95% of debts that are 
180 days delinquent.[A]. 

Source: GAO analysis of Labor data. 

Note: We reviewed the following documents: Labor's fiscal year 2007 
performance and accountability report, Labor's fiscal year 2008 
performance budget, and OWCP's 2007 operational plan. 

[A] In 2007, OWCP resumed measurement of and created a separate report 
to identify and track delinquent overpayments that have been referred 
to Treasury. However, data were not yet available to monitor the 
success of these efforts at the time of our review. 

[End of table] 

Several district office staff we interviewed reported that they focused 
most of their attention on aspects of the claims process that are 
regularly tracked and measured by OWCP, whereas tasks such as 
identifying and processing improper payments are given lower priority. 
In a recent rep6ort by Labor's Inspector General, officials in one 
district office stated that claims examiners' primary focus was to 
process claims and that time constraints prevented examiners from 
focusing on efforts to prevent and detect improper payments, such as 
confirming claimant eligibility information on a regular 
basis.[Footnote 14] 

While OWCP officials monitor payment accuracy as part of their biennial 
reviews of the district offices, the results are not incorporated into 
Labor's annual assessment of the FECA program's risk of improper 
payments, nor are they considered by outside auditors when they 
evaluate the program's internal controls.[Footnote 15] Each OWCP 
district office undergoes a review every other year, during which OWCP 
officials review a sample of claims files to evaluate a comprehensive 
range of district office operations--such as the appropriateness of 
decisions to accept or deny claims and the accuracy of payments-- 
against program-wide performance standards. For example, a sample of 
initial payments is reviewed to ensure that they are accurate, based on 
the appropriate pay rate from the employing agency, and are properly 
certified by a second claims examiner. 

However, permissible error rates are high--the permissible error rate 
for payment accuracy is 20 percent--and district offices with error 
rates that exceed the acceptable rate are not always required to 
implement corrective actions to address the identified 
deficiencies.[Footnote 16] Five of the six district offices OWCP 
reviewed in fiscal year 2006 failed to meet the standard because over 
20 percent of their payments were either inaccurate or lacked 
documentation to support the amount paid; one office had an error rate 
of 43 percent.[Footnote 17] However, only three of the offices were 
required to take corrective action to ensure that their future payment 
calculations are accurate. OWCP officials told us that offices with 
error rates slightly below the standard may not be required to develop 
corrective action plans if they have already taken steps to address the 
particular issue. They noted, however, that a corrective action plan is 
required whenever an office misses the performance standard by a large 
margin. Further, in commenting on this report, Labor also noted that 
findings from several review items may be combined into a single 
corrective action plan; as a result, formal remedies for payment 
accuracy may be in place even though they were not specifically cited 
in the reports reviewed by GAO. 

OWCP Lacks the Information Needed to Accurately Assess the FECA 
Program's Risk of Improper Payments: 

OWCP lacks the information that it needs to accurately assess the FECA 
program's risk of improper payments. Previous GAO work has shown that 
agencies must know the magnitude of improper payments and the causes of 
these errors in order to assess program risks and take actions to 
address them.[Footnote 18] A risk assessment entails a comprehensive 
review and analysis of program operations to determine where 
vulnerabilities exist and what those vulnerabilities are and to measure 
their potential or actual impact on program operations. Information on 
the magnitude and causes of improper payments form the foundation upon 
which management can determine the nature and type of corrections 
needed and give management baseline information for measuring progress 
in reducing improper payments. Without this information, a program is 
vulnerable to improper payments because managers may not be taking all 
of the precautions necessary to ensure that payments are accurate. 

First, OWCP does not collect or use available data to determine the 
magnitude of improper payments or their causes, making it difficult to 
identify vulnerabilities that lead to payment errors and determine 
their impact on program operations. For example, although the reports 
OWCP uses to manage the program list the amount of debts--including 
potential overpayments--for collection purposes, OWCP does not analyze 
the data to identify the magnitude of improper overpayments that occur 
during a particular year or whether they are increasing or decreasing 
over time. In addition, OWCP does not collect data on the magnitude of 
underpayments identified each year because of limitations in its data 
systems. Further, although agency officials cited several potential 
causes for improper payments, the agency does not collect aggregate 
information on how frequently these errors occur. As a result, the only 
way to determine why an overpayment or underpayment occurred is to 
review the information in each claims file. While OWCP began using a 
new data system in 2005 that has increased the amount and quality of 
data available on improper payments, program officials told us that 
they have no plans to collect data on the causes of improper payments. 
Without accurate data on improper payment risks, OWCP cannot target its 
resources towards preventing or reducing the errors that are the most 
prevalent or costly, nor can it monitor the effectiveness of its 
efforts to prevent such errors. In commenting on this report, Labor 
stated that it is (1) developing codes to track the reasons for 
overpayments and (2) considering a method to collect information on the 
reasons for underpayments. 

Further, while Labor is required to annually estimate the magnitude of 
improper payments in the FECA program, its estimate provides OWCP with 
limited information to identify and address program vulnerabilities. 
Labor estimated that the FECA program made $703,000 in improper 
payments in fiscal year 2006 based on a review of claims files for a 
sample of all payments made during the year and determined that the 
program had a low risk of improper payments because this estimate did 
not exceed $10 million and 2.5 percent of program payments--the 
threshold that federal guidance defines as high risk.[Footnote 19] 
While Labor followed the required guidance in developing this estimate, 
it may be understated because it does not include certain types of 
improper payments that recent audits suggest may be fairly prevalent in 
the program. For example, Labor's estimate would fail to catch improper 
payments that occurred because the information in the claims files that 
it reviewed--such as information on a claimant's work status--was 
inaccurate. If a claimant returned to work but failed to notify OWCP, 
this information would not be reflected in the claims file. A 2007 SSA 
Inspector General audit of the FECA program found that claimants failed 
to report approximately $12.6 million in wages to OWCP in 2004, 
suggesting that they had returned to work and were no longer eligible 
for wage loss compensation--but had not notified OWCP.[Footnote 20] 

Labor is not required to include the amount of improper payments that 
OWCP identified during the fiscal year in its estimate and OWCP does 
not track these data.[Footnote 21] However, these data provide 
comprehensive information on the FECA program's risk of improper 
payments that is useful in managing the program. From our analysis of 
potential overpayments and underpayments, we estimated that OWCP 
identified $13.3 million in improper payments in fiscal year 2006 ($7.1 
million in overpayments and $6.2 million in underpayments), some of 
which occurred in prior fiscal years.[Footnote 22] From our review of 
the claims files, we found that many of the improper payments spanned 
multiple fiscal years and some were not identified until several years 
after they occurred. Without comprehensive information on improper 
payments identified each year or trends in these payments over time, it 
is difficult for OWCP to identify vulnerabilities in the program that 
can lead to improper payments. 

OWCP's Dependence on Unverified Information, Internal Errors, and 
System Limitations Leaves FECA Vulnerable to Improper Payments: 

The FECA program is vulnerable to improper payments because OWCP relies 
on unverified self-reported eligibility information from claimants, 
makes internal payment errors, cannot stop certain payments, and 
receives inaccurate wage and benefits information from claimants' 
employing agencies. To identify the causes of improper payments and to 
estimate the total dollar values of improper overpayments and 
underpayments identified by OWCP in fiscal year 2006, we selected a 
sample of overpayments and another sample of potential underpayments 
and reviewed the selected claims files. Based on our overpayment file 
review, we estimated the causes of all overpayments identified by OWCP 
in 2006.[Footnote 23] We also collected information on the causes of 
underpayments, but were unable to estimate their prevalence because of 
limitations in OWCP's data.[Footnote 24] We found that overpayments 
commonly occurred when claimants failed to notify OWCP in a timely 
manner when they returned to work, or when family members did not 
quickly notify OWCP that a claimant had died. In other instances, 
overpayments occurred because claimants did not report earnings to 
OWCP. Claims examiners created improper overpayments and underpayments 
when they made payment calculation errors. They also overpaid claimants 
when they did not promptly stop payments after being notified that a 
claimant had returned to work or died. In addition, overpayments 
occurred because OWCP's administrative payment processing deadlines 
prevented claims examiners from quickly canceling some payments after 
being notified of changes in claimants' eligibility status. Finally, 
both overpayments and underpayments occurred when claimants' employing 
agencies provided inaccurate wage and benefits data to OWCP. Figure 3 
shows the causes we identified from our review of the claims files for 
overpayments. 

Figure 3: Estimated Percentage of Fiscal Year 2006 Overpayments by 
Cause: 

This pie chart is a pie chart showing estimated percentage of fiscal 
year 2006 overpayments by cause. 

[See PDF for image] 

Source: GAO analysis of a sample off FECA claims files. 

[End of figure] 

OWCP Relies on Claimants to Provide Key Eligibility Information: 

The FECA program is vulnerable to improper payments because it relies 
on claimants to report key eligibility information, such as when they 
return to work at their agencies or earn wages from other employment, 
and does not verify that the data are timely or accurate. 

Late or No Notice When Claimants Return to Work: 

Some overpayments occur because OWCP relies on claimants--rather than 
their employing agencies--to inform it when they return to work, and 
claimants do not always do so in a timely manner. From our review of 
claims files from a sample of the overpayments identified by OWCP in 
2006, we estimated that about 11 percent of all of OWCP's 2006 
overpayments occurred because claimants did not immediately notify OWCP 
when they returned to work. In some of these instances, claimants did 
not notify OWCP that they had returned to work at all--instead, their 
employing agencies notified OWCP. Until 1999, OWCP required employing 
agencies to submit a notification form when claimants returned to 
work.[Footnote 25] However, OWCP discontinued use of this form, and 
agencies are no longer required to notify OWCP when a claimant returns 
to work. Officials from one employing agency told us that they would 
like OWCP to reinstate use of the notification form in order to better 
ensure that wage-loss-compensation payments are terminated when 
claimants return to work.[Footnote 26] In commenting on this report, 
Labor noted that it is working to allow agencies to use its online 
system to notify OWCP electronically when a claimant returns to work. 

Late or No Notice When Claimants or Their Survivors Die: 

OWCP is also not always notified in a timely manner of the death of 
claimants or their survivors who were receiving survivor benefits. From 
our review of the claims files for OWCP's 2006 overpayments, we 
estimated that about 6 percent of all overpayments occurred when a 
claimant or survivor died but OWCP was not quickly notified. OWCP 
relies on claimants' survivors to inform the agency when claimants die. 
Among the overpayments we reviewed, survivors usually notified OWCP 
within a few months of a claimant's death. However, in a few 
situations, the claims examiners were never informed of the death but 
became aware of it through other means, such as when the annual forms 
they sent to the claimant were returned to OWCP as undeliverable or 
after an investigation by the claimant's employing agency. In one claim 
we reviewed, OWCP paid a claimant for more than a year after he died, 
until a U.S. Postal Service investigation uncovered that the claimant 
was dead, and the claimant's cousin had fraudulently accessed his bank 
account and withdrawn the funds. In addition, OWCP is not always 
notified in a timely manner of the death of survivors--such as a spouse 
or eligible dependent of a deceased claimant--who received survivor 
benefits.[Footnote 27] One of the 10 largest overpayments identified in 
2006, which totaled over $130,000, occurred when the widow of a FECA 
claimant died and OWCP was not notified. OWCP continued to 
automatically deposit her FECA survivor benefit payments to her bank 
account every month for more than 2 ˝ years after her death. 

A recent report by SSA's Inspector General found that nearly $2 million 
in wage-loss-compensation payments were made in 2004 to claimants who 
died in 2003 or earlier.[Footnote 28] OWCP headquarters officials told 
us the agency conducts monthly data matches for all FECA claimants with 
SSA's death records to prevent long-term overpayments to claimants who 
died.[Footnote 29] However, they acknowledged that there had been a 
recent 8-month lapse in these monthly data matches because OWCP's 
contract with SSA for data matching services had temporarily expired. 
In addition, OWCP officials told us that, because they do not collect 
the social security numbers (SSNs) of claimants' spouses or other 
eligible dependents who collect survivor benefits, they cannot conduct 
matches of their records against SSA's death records for these 
individuals. Because OWCP does not conduct death matches for claimants' 
spouses or other dependents, it cannot use this information to identify 
overpayments that occur when it is not notified that (1) a deceased 
claimant's spouse or other dependent who was receiving survivor 
benefits died or (2) when a claimant's spouse or dependents died-- 
making the claimant ineligible to receive a higher wage-loss- 
compensation payment based on having a spouse or other eligible 
dependents. 

Unverified Self-Reported Data on Earnings and Other Federal Benefits: 

OWCP also relies on claimants to report whether they earn wages, which 
may affect their eligibility for wage loss compensation, but claims 
examiners generally do not verify this information. Unlike other 
federal agencies such as SSA and the Department of Veterans Affairs, 
OWCP does not conduct a systemic data match of its records against 
SSA's wage records to identify unreported earnings. Instead, OWCP 
conducts these matches on an ad-hoc basis for individual claimants if a 
claims examiner suspects that a claimant has unreported 
earnings.[Footnote 30] Among claims in our sample of 212 overpayments, 
claims examiners only verified about 22 percent of claimants' annual 
earnings statements by comparing them to SSA's data between 2002 and 
2007. Four of the seven cases of unreported earnings included in our 
review were not uncovered by OWCP, but by fraud investigations 
undertaken by the claimants' employing agencies. Further, a recent 
report by SSA's Inspector General found that, in 2004, about 7 percent 
of the approximately 1,800 claimants that OWCP determined to be unable 
to work at all actually had earnings that were reported to 
SSA.[Footnote 31] 

Beyond the limitations associated with OWCP's ad-hoc verification of 
individual claimants' earnings, the effectiveness of OWCP's 
verification process is undermined by the fact that the data are not 
current. OWCP officials told us that SSA's earnings data are about 2 
years old. More current earnings data are available from another 
federal database, the National Directory of New Hires, a database 
maintained by the Department of Health and Human Services (HHS) to 
assist states in locating parents and enforcing child support orders. 
The database includes quarterly wage data for up to eight quarters, 
which can be compiled into annual data for matching purposes. OWCP 
could use these data to conduct systematic data matches with all of its 
claimants to identify those with unreported earnings. Before 
implementing such a match, OWCP and HHS would have to ensure that 
claimants' privacy and personal information were protected. At present, 
OWCP does not have legislative authority to access the database. In the 
last several years, the Congress has authorized some expanded use of 
this database, allowing other benefit programs to obtain the data. For 
example, it has allowed SSA to use the database to establish 
individuals' eligibility for Supplemental Security Income and the 
Department of Education to use it to collect student loan repayments. 
OWCP officials told us they have sought access to the National 
Directory of New Hires, but did not provide us with any formal 
legislative proposals requesting such authority. 

Some overpayments also occur because OWCP does not regularly verify 
whether claimants are receiving SSA retirement benefits. For FECA 
claimants in the current federal retirement system (those hired after 
1983) who are also collecting SSA retirement benefits, OWCP is required 
to reduce their FECA payments by the amount of their SSA payments 
attributable to their federal service.[Footnote 32] However, during our 
interviews, some claims examiners reported that identifying these 
claimants is difficult. The Department of Defense, which has undertaken 
an initiative to ensure that FECA payments made to its former employees 
are correctly reduced by the amount of their SSA retirement benefits, 
has helped OWCP institute 230 of these reductions. However, in 2006, 
fewer than 30 FECA claimants from all other employing agencies had 
their payments reduced because of SSA retirement benefits. This small 
number suggests that OWCP has not undertaken a serious effort to 
identify and reduce the wage-loss-compensation payments of claimants 
receiving SSA retirement payments as required. Because the number of 
FECA claimants covered by the current federal retirement system will 
substantially increase as time goes on, it is likely that the risk of 
these improper payments will also substantially increase. 

Errors by OWCP Claims Examiners Caused Overpayments and Underpayments: 

OWCP does not sufficiently ensure that its claims examiners correctly 
calculate payment amounts to claimants; promptly stop payments after 
they have been notified that a claimant has returned to work or died; 
make accurate decisions to deny, reduce, or terminate claimants' wage 
loss payments; or ensure that claimants file their annual eligibility 
and earnings statements. 

Inaccurate Payment Calculations by Claims Examiners: 

While OWCP requires an experienced claims examiner to certify the 
accuracy of the first payment made on each claim before it is issued, 
many errors are still undetected. According to a 2006 internal audit by 
OWCP, about 14 percent of the initial payments sampled had calculation 
errors, and an additional 14 percent could not be verified because the 
claims file contained insufficient information to support the 
calculations made by the claims examiner. From our review of the claims 
files for OWCP's 2006 overpayments, we estimated that about 15 percent 
of all overpayments occurred because claims examiners made payment 
calculation errors; we also found that some underpayments were caused 
by incorrect payment calculations. 

Calculation errors we found in our reviews of overpayments and 
underpayments included situations in which claims examiners incorrectly 
withheld health or life insurance premiums, paid claimants for the 
wrong amount of hours of lost wages, or paid claimants at the wrong 
rate. For example, several overpayments we reviewed occurred when 
claims examiners paid claimants twice for the same period. In addition, 
both overpayments and underpayments resulted when claims examiners 
withheld incorrect insurance premium amounts. For example, one 
underpaid claimant spent 4 years trying to get OWCP to stop incorrectly 
deducting premiums for family health insurance from the claimant's wage-
loss-compensation payments. The claims examiner then reimbursed this 
claimant for the error twice--effectively overpaying the claim by more 
than $14,000. 

Some of these payment errors reflect the complicated nature of 
accurately determining the amount of compensation to which a claimant 
is entitled. For example, several overpayments and underpayments we 
reviewed occurred because claims examiners incorrectly determined the 
pay rate of a claimant who had returned to work and was later re- 
injured. To correctly calculate payments in such cases, a claims 
examiner must determine whether the claimant was working in a different 
position when he or she was re-injured, compare this amount to what the 
claimant would have earned in his or her original position, and use the 
higher of the two values. Such errors underscore the need for adequate 
training for claims examiners in calculating wage-loss-compensation 
payments: 

Untimely Termination of Payments to Claimants Who Return to Work or 
Die: 

Claims examiners do not always promptly stop payments when they are 
notified that claimants have returned to work or died. From our review 
of OWCP's 2006 overpayments, we estimated that about 17 percent 
occurred when claimants returned to work or died but claims examiners 
did not stop payments quickly after they were notified of these events. 
We estimated that it took claims examiners an average of more than 5 
weeks to stop these payments.[Footnote 33] In some cases we reviewed, 
claims examiners had to be notified several times before finally 
halting payments. For example, a $106,000 overpayment resulted when a 
claimant continued to be paid wage loss compensation for nearly 13 
years after returning to work, despite numerous notifications. In 
another case, it took OWCP more than 2 ˝ years after a claimant's death 
to cancel wage-loss-compensation payments, even though the claims file 
contained more than 20 letters from OWCP to the claimant that had been 
returned as undeliverable. 

Incorrect Decisions to Deny, Reduce, or Terminate Wage Loss Payments: 

From our review of the claims files for potential underpayments, we 
estimated that OWCP identified about $6.2 million in underpayments in 
fiscal year 2006. Many large underpayments occurred because claims 
examiners either inappropriately determined that compensation should be 
denied, reduced, or terminated, or did not follow proper procedures 
when decreasing benefits.[Footnote 34] In one case we reviewed, for 
example, OWCP was required to pay one claimant over $29,000 in back 
compensation because it did not provide a notice explaining the 
claimant's due process rights when it proposed to terminate benefits. 
In another case, OWCP underpaid a claimant by over $83,000 because a 
claims examiner inappropriately terminated the claimant's benefits for 
refusing a job offer from an employing agency. After the claimant 
appealed the termination, it was determined that the claimant was 
justified in refusing the job because it did not meet the physical 
restrictions required by the injury. 

Failure to Ensure That Claimants Submit Annual Eligibility Forms or 
Follow Up on the Information on the Forms: 

Finally, OWCP does not consistently ensure that claimants return their 
annual eligibility forms or adjust benefits when the information 
reported by claimants indicates a change in their eligibility. 
According to OWCP's internal audits, 14 percent of the claims files the 
agency sampled in fiscal year 2006 were either missing annual 
eligibility forms entirely or contained forms with incomplete 
information on which claims examiners failed to follow up. In addition, 
4 percent of the sampled claims files contained information from 
claimants indicating that they may have been collecting dual benefits, 
had outside earnings, or had changes in their dependent status that 
affected their payments, but the claims examiners did not follow up on 
this information. In one claims file we reviewed, a claimant reported 
the death of a spouse on two separate annual eligibility forms, but the 
claims examiner never reduced the wage loss payments. 

Some Overpayments Occur because of Limitations in OWCP's Payment 
System: 

Based on our review of the claims files, we estimated that about 26 
percent of all of 2006 overpayments occurred because OWCP had already 
processed payments or mailed checks before claims examiners were 
notified of events affecting claimants' eligibility for wage loss 
payments, such as a claimant's return to work or death. Because of 
payment system limitations, claims examiners cannot cancel or make 
changes to automated monthly payments for a 10-day period prior to the 
end of each pay period. Therefore, if a claims examiner is informed 
during this period that a claimant has returned to work or died, the 
examiner cannot prevent an overpayment from being issued. Many OWCP 
claims examiners we interviewed cited the payment processing deadline 
as a frequent cause of overpayments in their caseloads. These 
overpayments are for relatively short periods and tend to be smaller 
than other overpayments; according to our estimates, these overpayments 
averaged about $900 each in 2006.[Footnote 35] Although they tended to 
be small, they occurred frequently and took time for claims examiners 
to process. 

Inaccurate Data from Employing Agencies also Lead to Improper Payments: 

Some improper payments occur because claimants' employing agencies 
provide inaccurate or incomplete wage and benefits data to OWCP. While 
OWCP relies on these agencies to report claimants' wage and benefits 
data on initial claims forms so that claims examiners can calculate 
wage-loss-compensation payments, it does not require them to provide 
evidence of their accuracy, such as by submitting copies of claimants' 
pay stubs. In fact, claims examiners in each of the five district 
offices we visited reported that the information provided by employing 
agencies on claims forms was frequently incomplete or incorrect. They 
also said that obtaining corrected information from employing agencies 
could be difficult. In our reviews of OWCP's 2006 improper payments, we 
found that both overpayments and underpayments were caused by 
inaccurate pay rate information provided by claimants' employing 
agencies. Underpayments also occurred because employing agencies failed 
to provide complete data on claimants' pay rates. For instance, several 
underpayments we reviewed occurred because employing agencies failed to 
indicate that claimants were entitled to extra pay for working at night 
or on Sundays. 

While OWCP depends on employing agencies to help identify payment 
errors, it does not provide them with sufficient tools to easily do so. 
Some employing agency officials told us that they are less able to 
monitor the accuracy of payments because they do not have ready access 
to detailed compensation data needed to verify OWCP's payment 
calculations.[Footnote 36] Employing agencies can track the total 
amount being paid to a claimant by OWCP in an online system maintained 
by OWCP; however, they are unable to view details used to calculate 
payments, such as base and premium pay rates, the amount withheld for 
health and life insurance premiums, and whether claimants have 
dependents. If the employing agencies want to audit their employees' 
claims and identify potential improper payments, they have to send 
representatives to OWCP's district offices to review the claims files. 
One of the ten largest overpayments identified in fiscal year 2006, 
totaling nearly $127,000, was uncovered when an employing agency sent a 
representative to OWCP to review claims files. The agency had 
previously informed OWCP that it was paying the claimant based on an 
incorrect pay rate. However, it was only during its claims file review 
nearly 4 years later that the agency discovered that the correct pay 
rate was still not being used. Recently, OWCP officials told us that 
OWCP was revising its online system to include more detailed payment 
information, which should be available by the spring of 2008. 

OWCP Does Not Ensure the Recovery of FECA Overpayments: 

OWCP does not sufficiently ensure the timely recovery of FECA 
overpayments and misses some opportunities for recovering them. 
Further, OWCP cannot identify how many overpayments are waived each 
year or what percentage of overpaid dollars are repaid. As a result, it 
cannot assess the success of its recovery efforts. 

OWCP Does Not Ensure Timely Processing of Overpayments: 

OWCP does not ensure that its claims examiners process overpayments of 
wage loss compensation in a timely manner, which delays their recovery. 
Before seeking recovery of an overpayment, OWCP requires a claims 
examiner to (1) issue a preliminary notice within 30 days of 
identifying the overpayment to explain the circumstances of the 
overpayment and give the claimant an opportunity provide additional 
information or contest the decision and (2) issue a final notice within 
30 days of the preliminary notice. 

Timely processing of overpayments is critical to recovering the amounts 
owed. First, OWCP cannot attempt to recover overpayments until both the 
preliminary and final notices have been sent to the claimants. Second, 
prior GAO work has shown that successful recovery of overpayments is 
directly related to the time it takes to confirm and process the 
overpayment.[Footnote 37] Specifically, the longer it takes to process 
an overpayment, the less likely it will be that a claimant will still 
be receiving FECA benefits from which the overpayment can be recouped. 
Finally, an overpayment that OWCP is unable to recover cannot be 
transferred to Treasury for additional recovery efforts until a final 
overpayment notice has been issued to the claimant. Treasury has 
recovery tools not available to OWCP, such as deducting overpayments 
from a claimant's federal tax refund or other federal payments. When 
OWCP does not issue a final overpayment notice promptly, it results in 
delays in transferring debts to Treasury and in applying these 
additional tools. 

Available data indicate that OWCP does not ensure that overpayments are 
confirmed and processed within its required time frames. As shown in 
figure 4, OWCP's September 2007 debt-aging report indicates many delays 
in processing overpayments. Almost half of the identified overpayments 
were more than 6 months old, and OWCP had not yet issued preliminary 
overpayment notices to the claimants.[Footnote 38] Similar delays were 
evident when OWCP assessed overpayment processing during its internal 
reviews of the district offices. For the six district offices reviewed 
in 2006, the district office had not issued a preliminary notice to 
claimants for over one-third of the overpayments OWCP reviewed. 
Further, based on our review of the 2006 overpayments, we estimated 
that, on average, OWCP issued the final overpayment notice to claimants 
64 days after the preliminary notice, with a range of 26 days to 470 
days.[Footnote 39] In several district offices we visited, claims 
examiners or district office officials told us that they sometimes 
delayed issuing required overpayment notices to claimants because their 
first priority is to pay claims. A claims examiner in one district 
office, for example, told us about a backlog of overpayment cases for 
which preliminary or final overpayment notices had not been sent to the 
claimant. 

Figure 4: Status of Overpayments Listed on OWCP's Debt-Aging Report, 
September 2007: 

This figure is a combination bar chart showing status of overpayments 
listed on OWCP's debt-aging report, September 2007. The X axis 
represents GAO analysis of OWCP data, and the Y axis represents the 
percentage of overpayments in each status. 

Pending overpayment: Less than 30 days: 27; 
Pending overpayment: Over 180 days: 45. 

Preliminary notice issued: Less than 30 days: 23; 
Preliminary notice issued: Over 180 days: 36. 

Final notice issued: Less than 30 days: 58; 
Final notice issued: Over 180 days: 16. 

[See PDF for image] 

Source: GAO analysis of OWCP data. 

Note: If a claimant appeals OWCP's preliminary overpayment decision, 
the overpayment remains in preliminary status until an appeal decision 
is issued. Some of the preliminary overpayments that were more than 6 
months old could have been waiting for an appeal decision. 

[End of figure] 

We also found that some OWCP district offices experienced more delays 
in processing overpayments than others. OWCP's fiscal year 2006 
overpayments, for example, showed that most district offices had issued 
preliminary notices by March 2007 for at least 90 percent of pending 
overpayments. However, four offices had issued preliminary overpayment 
notices for a lower percentage of overpayments, with one office issuing 
preliminary notices for less than half of its pending overpayments. 

While some cases we reviewed involved complicated issues and may have 
required extra time for OWCP to confirm the existence and amount of the 
overpayment before sending a preliminary overpayment notice to the 
claimants, other claims files we reviewed had no readily apparent 
reasons for the delays. In one case, for example, a postal worker 
aggravated a pre-existing knee injury as a result of prolonged bending 
and walking while delivering mail. The claimant subsequently returned 
to work and was overpaid almost $700. However, the claims examiner did 
not send out the preliminary overpayment notice for nearly a year after 
being notified that the claimant had returned to work. 

OWCP officials acknowledged that implementation of OWCP's new data 
system had resulted in some unreliable data on the debt-aging report 
and disrupted the program's ability to track overpayments for the past 
few years. They told us that they are taking steps to improve the 
reliability of the debt-aging report and to ensure that overpayments 
are processed within OWCP's required time frames. For example, in 2007 
OWCP directed its district offices to review the debt-aging report in 
order to identify inaccurate data and outstanding overpayments that 
required action. OWCP officials told us that the reliability of the 
data has improved since this review, and the number of potential 
overpayments listed on the report has been significantly reduced. They 
confirmed that the data we analyzed from the September 2007 debt-aging 
report should be reasonably reliable. 

About 70 Percent of Overpayments Are Repaid, but OWCP Overlooks 
Opportunities to Recover Overpayments: 

While OWCP does not track the recovery status of overpayments, our 
review of claims files for 2006 overpayments indicated that an 
estimated 71 percent were repaid or being collected at the time of our 
review.[Footnote 40] An additional 19 percent of the overpayments were 
waived by OWCP, in most cases because the overpayment was less than 
$700 and the cost of trying to recover the overpayment was expected to 
be greater than the amount actually recovered. See figure 5 for the 
recovery outcomes for our claims file sample. OWCP officials told us 
that they planned to collect more information on the outcome of OWCP's 
recovery efforts--such as whether overpayments are waived or repaid--by 
the end of fiscal year 2008. 

Figure 5: Estimated Recovery Status of Overpayments That OWCP Waived or 
Pursued in Fiscal Year 2006: 

This figure is a pie chart showing estimated recovery status of 
overpayments that OWCP waived or pursued in fiscal year 2006. 

Repaid: 61%; 
Waived: 19%; 
Being collected: 10%; 
Delinquent: 6%; 
Unknown: 3%; 
Referred to Treasury: 2%. 

[See PDF for image] 

Source: GAO analysis of a sample of FECA claims files. 

[End of figure] 

Note: Percentages do not add to 100 due to rounding. 

We estimated that about half of the overpayments that were repaid or 
being collected were repaid directly by the claimant, while 21 percent 
were withheld from the claimant's FECA payments (see fig. 6). 

Figure 6: Estimated Recovery Sources for 2006 Overpayments That Were 
Repaid or Being Collected: 

This figure is a pie chart showing estimated recover sources for 2006 
overpayments that were repaid or being collected. 

Voluntary repayment: 58%; 
Withholding from FECA payments: 22%; 
Recoupment from deceased claimant's bank or estate: 17%; 
Court mandated restitution: 3%; 
Refund from OPM retirement benefits: 1%. 

[See PDF for image] 

Note: Percentages do not add to 100 due to rounding. 

[End of figure] 

Based on our review of the 2006 overpayments, we estimated that OWCP 
waived about $118,000 (about 2 percent) of all 2006 overpaid dollars at 
the time of our review. Given the potential costs of recovering an 
overpayment, OWCP claims examiners may immediately waive any FECA 
overpayments under $200 without notifying the claimant that an 
overpayment has occurred. Claims examiners can also waive overpayments 
under $700 after sending the claimant a preliminary overpayment notice 
if the additional recovery costs are expected to exceed the amount to 
be recovered. From our review of the 2006 overpayments, we estimated 
that 77 percent of the overpayments $200 or under were waived, compared 
with 38 percent of the overpayments between $200 and $700, and 1 
percent of the overpayments that were $700 or greater.[Footnote 41] 

Other benefit programs GAO has reviewed have established lower minimum 
thresholds below which they will not seek recovery; in other words, 
they seek to recover more overpayments than OWCP does. For example, 
SSA's Supplemental Security Income program, which provides cash 
assistance to certain categories of people who have limited income and 
resources, waives overpayments of $500 or less if the recipient was not 
at fault in the creation of the overpayment and cannot afford to repay 
it. An SSA program official told us that, because most of its 
overpayments are automatically generated, the program attempts to 
recover most overpayments over $1.[Footnote 42] In addition, according 
to the Director of the Veterans Affairs' (VA) Debt Management Center, 
the VA's pension and disability compensation programs have totally 
automated overpayment -processing systems, making it cost effective to 
attempt recovery on small overpayments. The VA attempts to recover any 
overpayment over $5. Because OWCP's data system does not automatically 
generate overpayment notices, these waiver thresholds may not be cost- 
effective for FECA at this time. However, OWCP implemented a new data 
system in 2005 and is continuing to add new capabilities related to 
tracking improper payments. As its overpayment processing capabilities 
improve, OWCP may be able to reduce its waiver thresholds and seek to 
recover more overpayments. 

We found two cases during our review of the 2006 overpayments in which 
the claims examiners waived the overpayments even when the claimant 
indicated a willingness to pay back the amount. In one case, a claimant 
had incurred a $430 overpayment and was sent a preliminary overpayment 
notice. The claimant sent a letter to OWCP indicating that he wanted to 
repay the overpayment in four installments. Instead, however, the 
claims examiner waived the overpayment. Similarly, a claimant called 
OWCP to establish a payment plan for her $575 overpayment, but the 
claims examiner told her that the overpayment would be waived because 
it was less than $700. For waivers of overpayments under $700, OWCP 
procedures state that the claims examiner should consider such factors 
as whether the claimant can be located, the likelihood of recovery, and 
the potential costs of pursuing the case. 

We also found instances when claims examiners missed opportunities to 
deduct overpayments from other FECA payments, including wage loss 
compensation and other types of payments from OWCP. In one case, for 
example, a claims examiner made a $29,000 payment to a claimant with an 
outstanding potential overpayment of almost $10,000. While the 
claimant's FECA payments were stopped, the claims examiner did not 
issue a preliminary overpayment notice to the claimant for over a year. 
The claims examiner issued the final overpayment notice 2 months later, 
in the same month that the $29,000 payment was made. At the time of our 
review, the $9,940 debt had not been repaid and was considered 
delinquent. In another case, the claimant incurred a $660 overpayment 
because health and life insurance premiums had not been deducted from 
his FECA payments. The overpayment was waived in July 2006 despite the 
fact that the claimant continued to receive manually generated FECA 
wage-loss-compensation payments until September 2006 and began 
receiving automatic monthly payments in October 2006. These missed 
opportunities may have occurred, in part, because OWCP's data system 
does not easily identify when a claimant with an overpayment is also 
receiving FECA payments. For example, one district office official 
pointed out that when a claims examiner processes a new payment, the 
system does not notify the claims examiner if the claimant has any 
outstanding overpayments. Further, the debt-aging report does not 
include a claimant's payment status, which could alert claims examiners 
when a claimant with an outstanding overpayment begins receiving FECA 
payments. In commenting on this report, Labor noted that it is planning 
to create an automated prompt that will alert claims examiners 
preparing a FECA payment that a claimant has an existing overpayment. 

OWCP's relatively infrequent use of wage garnishment may also represent 
a missed opportunity to recover overpayments, especially given the 
large number of overpayments that occur when claimants return to work 
at a federal agency. OWCP procedures identify wage garnishment as a 
recovery option but do not include any detailed instructions on how to 
implement it. Instead, claims examiners are encouraged to transfer 
overpayments for which garnishment is an option to Treasury and have 
that agency garnish the claimant's wages. Most claims examiners we 
interviewed did not mention wage garnishment as a recovery option. 
Further, none of the overpayments in the claims files we reviewed were 
recovered through wage garnishment. Many overpayments created when 
claimants return to work are small and may not be worth the effort of 
arranging for wage garnishment with the employing agency. However, some 
overpayments are large and could be recovered in this manner. OWCP 
headquarters officials told us that wage garnishment is a viable 
recovery option, but noted that it may be difficult for claims 
examiners to identify the correct employing agency official who can 
arrange for the claimant's wages to be garnished. 

Claims examiners may miss these recovery opportunities for several 
reasons. As noted earlier, they may be more focused on paying initial 
wage loss claims than recovering overpayments. In addition, some claims 
examiners we interviewed were not completely familiar with FECA's 
overpayment recovery process, usually because they did not process 
overpayments very often. For example, some claims examiners were unsure 
about which overpayments were eligible to be waived; however, they said 
they reviewed the FECA procedures manual when they had to process an 
overpayment. While some claims examiners reported that they 
aggressively pursued the recovery of overpayments from available 
recovery sources, others said they were more limited in what they could 
do. For example, one claims examiner told us that he could not recover 
overpayments from a claimant's wage-loss-compensation payments, even 
though the FECA procedures manual recommends that overpayments over 
$200 be finalized and recovered from ongoing wage-loss-compensation 
payments as quickly as possible. In one district office, two claims 
examiners also told us that they were not allowed to use wage 
garnishment to recover overpayments, although OWCP headquarters 
officials told us that claims examiners can garnish wages to collect 
overpayments. 

Several claims examiners told us that it would be more effective to 
assign responsibility to one person for recovering overpayments, rather 
than expecting claims examiners to focus on recoveries in addition to 
all their other responsibilities. Such an approach could reduce 
instances when a claims examiner overlooks a potential recovery source 
or unnecessarily waives an overpayment. One district office we visited 
had recently implemented this approach, giving responsibility to a 
staff person in the fiscal office to recover all overpayments once the 
final overpayment notice was sent to the claimant. In dividing the 
overpayment process, the district office manager said he hoped that 
claims examiners would focus on processing overpayments more quickly, 
while the fiscal staff member would specialize in the recovery process. 
In another district office, a fiscal staff member provided assistance 
to claims examiners by pointing out potential recovery sources for 
overpayments, such as OPM retirement benefits. Fiscal staff in the 
other three district offices we visited were not involved in the 
recovery of overpayments. 

Conclusions: 

OWCP appropriately places a high priority on making timely wage-loss- 
compensation payments to injured federal workers who might otherwise 
face financial hardship while, at the same time, helping them return to 
work when their injuries have been resolved. However, without a 
counterbalancing emphasis on preventing and identifying improper 
payments as well as recovering overpayments, claimants may not receive 
all of the compensation to which they are entitled and program dollars 
may be spent on ineligible claimants, threatening the overall integrity 
of the program. To effectively address program vulnerabilities, OWCP 
must first identify those vulnerabilities. The current process for 
assessing FECA's risk of improper payments does not provide sufficient 
information on the extent of improper payments. Further, because OWCP 
lacks data on the causes of improper payments, it cannot focus its 
efforts on preventing them before they occur or addressing those areas 
that are at the highest risk for errors. It is unlikely that the 
vulnerabilities we have identified will be addressed without a change 
in OWCP's current management strategies with respect to improper 
payments. 

We recognize that determining eligibility for FECA wage-loss benefits 
and calculating wage-loss payments is complicated and depends on 
employing agencies to provide accurate information to OWCP, which 
leaves the process at risk for errors. However, OWCP's own errors and 
reliance on self-reported data with respect to when claimants return to 
work and their earnings also contributes to the risk of improper 
payments. Taking more proactive steps to reduce the number of improper 
payments would allow OWCP to provide better customer service to 
claimants because such action would decrease the considerable time it 
takes to document, process, and recover overpayments. While OWCP can do 
much on its own to reduce the risk of improper payments, its ability to 
detect improper payments related to unreported earnings is hindered by 
the cumbersome process it must use to verify claimants' earnings 
through SSA and the outdated data produced by this process. While OWCP 
does not have the legislative authority to access the more timely and 
accurate earnings data available in the National Directory of New 
Hires, Congress has recently expanded access to these data for other 
programs to use in verifying individuals' eligibility for federal 
benefits. 

Similarly, OWCP can do more to focus on the recovery of overpayments. 
With little data on the recovery status of overpayments, OWCP cannot 
monitor the effectiveness of its recovery efforts or use this 
information to improve these efforts. While our review indicates that 
OWCP is recovering many overpayments, it also suggests that recoveries 
could be increased if OWCP acts on the missed recovery opportunities we 
identified. 

Recommendations for Executive Action: 

We recommend that the Secretary of Labor direct OWCP to develop a 
management strategy to ensure that the program's emphasis on quickly 
processing and paying FECA claims is balanced with the need for payment 
accuracy. Specifically, the agency should take the following two 
actions: 

* revise its program performance measures to ensure increased emphasis 
on payment accuracy, adequate internal controls, and overpayment 
recoveries and: 

* collect more detailed information on improper payments, such as the 
causes of overpayments and underpayments, and use these data to better 
identify improper payment risks and to address areas of high risk. 

We also recommend that the Secretary direct OWCP to take steps to 
reduce common causes of improper payments, such as: 

* requiring agencies to report to OWCP when a FECA claimant returns to 
work and provide incentives for agencies to notify OWCP quickly; 

* ensuring that its data match with SSA's death records is conducted 
regularly and consistently and that it includes individuals who are 
receiving survivor death benefits; 

* taking steps to ensure that wage-loss-compensation payments for 
claimants covered by the current federal retirement system are 
appropriately reduced by the amount of their SSA benefits that are 
attributable to their federal service; 

* considering ways to reduce the time it takes to process automated 
monthly payments; 

* determining what additional training claims examiners may need to 
improve payment accuracy; and: 

* exploring options for improving information sharing between OWCP and 
employing agencies so that OWCP can make accurate payments and agencies 
can help identify payment errors. 

To allow OWCP to more effectively verify the earnings information 
reported by FECA recipients and identify instances in which a claimant 
receiving wage loss compensation has unreported earnings, we recommend 
that the Secretary of Labor direct OWCP to develop a legislative 
proposal seeking legal authority to enter into a data-matching 
agreement with the Department of Health and Human Services to identify 
FECA claimants who have earnings reported in the National Directory of 
New Hires. Any such data-matching agreement would need to include 
appropriate safeguards for protecting claimants' privacy and personal 
information. 

We further recommend that the Secretary direct OWCP to take steps to 
focus attention on the recovery of FECA overpayments, such as: 

* collecting more detailed information on how overpayments are resolved 
in order to monitor the effectiveness of OWCP's recovery efforts; 

* holding staff accountable to ensure that overpayments are processed 
in a timely manner; 

* considering reducing the dollar threshold for waiving overpayments as 
OWCP's overpayment processing data system develops additional 
capabilities; 

* determining whether having fiscal staff dedicated to recovering 
overpayments would increase their recovery; and: 

* developing system modifications that would automatically identify 
claimants who have outstanding overpayments in order to ensure that 
debts are repaid from future benefit payments. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to Labor for review and comment. The 
agency provided comments, which are reproduced in appendix II. Labor 
expressed some concerns about our findings and conclusions but did not 
specifically comment on our recommendations. However, Labor also 
indicated that it is taking a number of steps that are consistent with 
our recommendations. In its comments, Labor stated that FECA 
overpayments and underpayments should not be considered improper 
because Labor adjusts them once additional information becomes 
available. We used the definition of improper payments from the 
Improper Payments Information Act: "any payment that should not have 
been made or that was made in an incorrect amount." In addition, Labor 
stated that our analysis of improper payments identified in 2006 does 
not accurately reflect its performance in terms of managing improper 
payments because it includes payments that occurred in previous years. 
We determined, however, that using currently available data on improper 
payments that occurred in 2006 would not have provided an accurate 
assessment of the risk of the FECA program to improper payments because 
OWCP does not identify some improper payments until a year or more 
after they occur. As a result, we based our analysis on improper 
payments that were identified in 2006 without respect to when they 
occurred and describe in the report how our measure of improper 
payments differs from Labor's. This analysis allowed us to demonstrate 
the type of information that OWCP could collect to identify and address 
the largest risks that lead to improper payments. Labor also stated 
that system improvements and staff training have improved its data and 
processing of overpayments. We acknowledge in our report that Labor's 
data on improper payments have recently improved. We believe that the 
improvements in Labor's performance in addressing improper payments 
have resulted from its efforts to focus more attention on these 
payments since our review began. This supports our assertion that the 
risks of improper payments are reduced when agencies strike an 
appropriate balance between service delivery and payment accuracy. We 
continue to believe that our findings, conclusions, and recommendations 
are sound. Labor's comments and our responses are reproduced in their 
entirety in appendix II. In addition, we incorporated clarifications in 
the report as appropriate. 

We are sending copies of this report to the Secretary of Labor, 
relevant congressional committees, and other interested parties. We 
will make copies available to others upon request. In addition, the 
report will be available at no charge on GAO's Web site at [hyperlink, 
http://www.gao.gov]. 

If you or your staff have any questions or wish to discuss this report 
further, please contact me at (202) 512-7215 or at bertonid@gao.gov. 
Contact points for our Offices of Congressional Relations and Public 
Affairs may be found on the last page of this report. The major 
contributors are listed in appendix III. 

Signed by: 

Daniel Bertoni: 

Director, Education, Workforce, and Income Security Issues: 

List of Congressional Committees: 

The Honorable Edward M. Kennedy: 
Chairman: 
Committee on Health, Education, Labor, and Pensions: 
United States Senate: 

The Honorable Patty Murray: 
Chairman: 
Subcommittee on Employment and Workplace Safety: 
Committee on Health, Education, Labor, and Pensions: 
United States Senate: 

The Honorable Joseph I. Lieberman: 
Chairman: 
Committee on Homeland Security and Governmental Affairs: 
United States Senate: 

The Honorable Daniel K. Akaka: 
Chairman: 
Subcommittee on Oversight of Government Management, the Federal 
Workforce, and the District of Columbia: 
Committee on Homeland Security and Governmental Affairs: 
United States Senate: 

The Honorable Joe Wilson: 
Ranking Member: 
Subcommittee on Workforce Protections Committee on Education and Labor: 
House of Representatives: 

[End of section] 

Appendix I: Scope and Methodology: 

To address our objectives, we requested data from OWCP on the volume, 
causes, and recovery outcomes of improper payments it identified. 
However, OWCP was only able to provide us with limited information on 
overpayments, and could not provide us with any data on underpayments 
because it could not separate them from other payment types. As a 
result, we decided to review a sample of claims files containing 
identified overpayments to learn more about their characteristics and 
reviewed a sample of other claims files to look for underpayments and 
estimate the total dollar value of underpayments OWCP identified in 
fiscal year 2006. We used the data gathered through these file reviews 
to estimate the magnitude of OWCP's improper payments. We also reviewed 
results from OWCP's internal audits, called accountability reviews, to 
learn more about risks OWCP identified. Additional information on how 
we conducted our claims file reviews and other analyses are discussed 
below. We conducted our review between September 2006 and January 2008 
in accordance with generally accepted government auditing standards. 

Review of Overpayments Identified by OWCP in Fiscal Year 2006: 

OWCP collects limited information on overpayments, but was able to 
create a report for us listing all FECA debts identified in fiscal year 
2006. This data extract was run on March 13, 2007, and included basic 
data on all 2006 FECA debts as of that date, including type, processing 
status, and initial and current balances. This data source contained 
records of both wage loss compensation overpayments and other types of 
debts, such as medical provider debts and debts associated with legal 
settlements. To assess the reliability of these data, we (1) reviewed 
existing documentation related to the data sources, (2) electronically 
tested the data to identify obvious problems with completeness or 
accuracy, and (3) interviewed knowledgeable agency officials about the 
data. We determined that the data extract was sufficiently reliable for 
our purposes. Before selecting our sample, we removed all debt records 
that did not appear to be related to wage-loss-compensation payments. 
In addition, we removed all debts that had been voided. As shown in 
table 2, the large majority of the remaining debt balance was due to 
wage loss compensation overpayments, with smaller amounts related to 
benefits fraud, employing agency overpayments, and miscellaneous 
situations. Their initial balances totaled just over $14.7 million. 

Table 2: Initial Balances of Compensation-Related Overpayments 
Identified in Fiscal Year 2006: 

Debt type: Wage loss compensation; 
Pending: $2,067,950.82; 
Preliminary: $2,244,362.57; 
Final: $3,615,526.16; 
Terminated: $4,522,698.35; 
Suspended: $4,781.53; 
Total: $12,455,319.43. 

Debt type: Benefits fraud; 
Pending: $201,010.26; 
Preliminary: $78,212.21; 
Final: $484,054.69; 
Terminated: $220,625.29; 
Suspended: $0.00; 
Total: $983,902.45. 

Debt type: Employing agency; 
Pending: $0.00; 
Preliminary: $0.00; 
Final: $0.00; 
Terminated: $10,614.75; 
Suspended: $0.00; 
Total: $10,614.75. 

Debt type: Miscellaneous; 
Pending: $202,561.11; 
Preliminary: $12,717.80; 
Final: $348,374.76; 
Terminated: $699,456.67; 
Suspended: $0.00; 
Total: $1,263,110.34. 

Debt type: Total; 
Pending: $2,471,522.19; 
Preliminary: $2,335,292.58; 
Final: $4,447,955.61; 
Terminated: $5,453,395.06; 
Suspended: $4,781.53; 
Total: $14,712,946.97. 

Source: GAO analysis of OWCP data. 

[End of table] 

Before selecting our sample, we removed pending, preliminary 
determination, and suspended overpayments from the universe in order to 
avoid analyzing characteristics of overpayments that might later be 
voided or overturned.[Footnote 43] All remaining overpayments were 
either in final determination or terminated status. Final determination 
overpayments had neither been waived nor overturned on appeal and were 
being collected as of the date of the data run. Once final 
determination overpayments are collected in full, they become 
terminated. Overpayments may also be terminated if they are waived or 
referred to Treasury because of non-repayment. OWCP groups all of these 
overpayments together because it does not track how overpayments are 
resolved. 

The initial balances of the nearly 2,800 remaining final and terminated 
overpayments totaled about $9.9 million. As seen in figure 7, 
terminated overpayments made up the bulk of the universe; they 
represented 82 percent of all records. Final determination overpayments 
accounted for 18 percent of the universe and had a higher median dollar 
amount--about $2,600 as compared to just over $600 for terminated 
overpayments. 

Figure 7: Initial Balances of All Wage Loss Compensation Overpayments 
Identified in Fiscal Year 2006 and in Final Determination or Terminated 
Status, as of March 13, 2007: 

This figure is a combination bar chart showing initial balance of all 
wage loss compensation overpayments identified in fiscal year 2006 and 
in final determination or terminated status, as of March 13, 2007. The 
X axis represents the overpayment amounts, and the Y axis represents 
the number of claims. 

Under $200: Final determination: 1; 
Under $200: Terminated: 565. 

$200-under $700: Final determination: 58; 
$200-under $700: Terminated: 651. 

$700-under $5000: Final determination: 282; 
$700-under $5000: Terminated: 873. 

$5000-under $10,000: Final determination: 63; 
$5000-under $10,000: Terminated: 102. 

$10,000 and over: Final determination: 91; 
$10,000 and over: Terminated: 100. 

[See PDF for image] 

Source: GAO analysis of OWCP data. 

[End of figure] 

We selected a random sample of 331 overpayment records, stratified by 
initial dollar value, from the universe of almost 2,800 final and 
terminated overpayments. Some claims had multiple overpayment records, 
but OWCP officials informed us that some claims examiners may have 
incorrectly created separate overpayment records each time a claimant 
submitted a repayment on a preexisting debt. As a result, we sampled 
cases with multiple overpayments in separate strata and reviewed all 
overpayments listed under each claim to better estimate the prevalence 
of incorrectly recorded overpayments. 

Once we selected our sample, we requested the electronic file for each 
claim that had a sampled overpayment. We reviewed these files to 
collect information including what caused overpayments, how long they 
lasted, how long it took claims examiners to process them, whether they 
were appealed or waived, whether they were repaid, and whether claims 
files contained annual eligibility updates. A second person verified 
the information collected. We only included overpayments that we 
considered to be improper in our analysis. Debts that were not due to 
improper payments, such as debts for copying expenses or overpayments 
that occurred when claimants retroactively elected to join other 
benefits programs, were excluded from our analysis. We also excluded 
cases that should not have been included in the universe of 2006 debts. 
(Some debt records were actually repayments on debts identified in 
previous fiscal years.) 

Using only the 212 overpayments records that remained after these 
exclusions, we analyzed data on overpayment characteristics and made 
projectable estimates to the universe of final determination and 
terminated fiscal year 2006 overpayments. Unless otherwise noted, 
estimates are accurate to within plus or minus 10 percentage points, at 
the 95 percent confidence level. For instance, our estimate of the 
total dollar value of improper payments has a much wider confidence 
interval. 

While OWCP's debt database was sufficiently reliable for our purposes, 
there are some limitations to our analysis. First, because we excluded 
overpayments that were in pending, preliminary, and suspended statues 
from our analysis, our results are only projectable to the universe of 
overpayments identified in fiscal year 2006 and placed in final 
determination or terminated status by March 13, 2007. As a result, our 
estimate of the dollar value of improper overpayments identified by 
OWCP is low. Additionally, our estimates may capture characteristics 
that are not representative of all overpayments. Since the overpayments 
we reviewed had already been finalized or terminated, they could have 
been processed faster by claims examiners, have fewer appeals by 
claimants, or exhibit quicker repayment than the pending, preliminary, 
and suspended overpayments we excluded from the universe. Further, 
since we were only able to analyze recovery outcomes of overpayments 
that had been terminated as of March of 2007, the proportions of all 
2006 overpayments that were repaid, waived, or referred to Treasury for 
recoupment may differ from our estimates. 

Review of Potential Underpayments Identified by OWCP in Fiscal Year 
2006: 

OWCP does not track underpayments and was therefore unable to provide 
us with data on the volume, dollar value, or causes of underpayments it 
identified in fiscal year 2006. However, it was able to provide us with 
an extract of all payments made through its payment override system, 
called the "direct payment" system, which includes some underpayments. 
The direct payment system allows claims examiners to issue payments in 
special circumstances, such as when a claimant has already been paid 
for a particular period, but is owed additional compensation. 
Underpayment reimbursements are issued via direct payment if a claimant 
was previously compensated for a particular period and was underpaid. 
If a claimant was not originally compensated for a period, but should 
have been, the underpayment reimbursement would then be issued through 
OWCP's standard payment process--not the direct payment system. Other 
instances in which payments must be issued through the override system 
include when payments cover periods extending far into the future, when 
payments are very large, and when payments are issued to nonclaimants, 
such as survivors of deceased federal workers. 

We used the data set provided to create an estimate of the total dollar 
value of underpayments OWCP identified in fiscal year 2006. We limited 
the universe to the type of payments most likely to contain 
underpayments. First, we excluded payments to individuals other than 
claimants because these payments must be made via direct payment. Then 
we excluded payments coded as schedule awards because lump sum schedule 
awards, whose payment periods extend into the future, must be issued as 
direct payments. Finally, we excluded payments that had been canceled. 
All remaining payments were coded as wage-loss compensation payments to 
claimants. 

We selected a random sample of the remaining records that was 
stratified by dollar value. We selected the 90 claims with the highest 
total dollar value of direct payments, and 110 additional claims for 
review. These 200 claims had a total of 344 direct payment records 
because some claims had multiple direct payments. We then reviewed each 
payment in these claims files and collected information on whether the 
direct payments listed were issued as reimbursement for improper 
underpayments or whether they were issued for other reasons. We also 
collected information on what caused each underpayment, who was 
responsible for it, and who identified it. When OWCP made a direct 
payment because a decision to deny or terminate payment was overturned 
on appeal, we did not consider it an underpayment if the decision was 
overturned based on new information that was not available when the 
original decision was made. We did consider it an underpayment when the 
denial or termination was overturned because of an error in judgment or 
because a claims examiner did not follow appropriate FECA procedures. A 
second person verified the information collected, as well as the 
judgments made by the original reviewer. From our review, we determined 
that 172 of the 344 payment records represented improper underpayments. 

The key limitation of this data set was the fact that an unknown number 
of underpayments were not included in the universe. However, because 
sampling the direct payment universe was the only available way to 
identify underpayments, we determined that it was sufficiently reliable 
for our purposes. Since the universe does not include all underpayment 
reimbursements, our estimate of the total dollar value of underpayments 
identified in fiscal year 2006 is understated. Additionally, because we 
designed our sampling methodology primarily for the purposes of 
developing an estimate of the total dollar value of underpayments, we 
were unable to estimate the prevalence of causes of underpayments with 
sufficient precision. 

Analysis of Labor's Improper Payment Risk Estimate: 

We relied on the results from our claims file reviews of potential 
overpayments and underpayments to develop an estimate of improper 
payments identified by OWCP in 2006. As noted above, OWCP's data 
systems do not collect information on improper underpayments. OWCP did 
provide data on overpayments identified in 2006, including $9.9 million 
for overpayments that were finalized or terminated by March 2007. In 
reviewing a sample of these claims files, we found that not all 
overpayments included on the list represented improper payments. 
Consequently, we developed our own estimate of improper overpayments 
based on the results from our file review. We estimated that OWCP 
identified $13.3 million in improper payments in fiscal year 2006 ($7.1 
million in improper overpayments and $6.2 million in improper 
underpayments). The 95 percent confidence interval for this estimate 
ranged from approximately $10.4 million to $16.2 million. This is an 
estimate of the magnitude of improper payments that OWCP identified in 
fiscal year 2006, not the magnitude of improper payments that OWCP made 
in fiscal year 2006. Our estimate of $13.3 million includes some 
payments made to claimants in previous fiscal years (such as fiscal 
year 2004 or 2005) that OWCP discovered were improper in 2006. Table 3 
provides information on the proportion of the sample of improper 
payments we reviewed that were made in fiscal year 2006 and the 
proportion made in prior fiscal years. 

Table 3: Distribution of the Sample of Improper Payments Identified in 
Fiscal Year 2006 by the Year in which the Payment Was Made: 

Total improper payments; 
Sample of overpayments: Number (% total): 212; 
Sample of overpayments: Dollars (% total): $1,469,000; 
Sample of underpayments: Number (% total): 172; 
Sample of underpayments: Dollars (% total): $3,483,000. 

Some or all of the improper payment was in fiscal year 2006; 
Sample of overpayments: Number (% total): 162 (76.4%); 
Sample of overpayments: Dollars (% total): $576,000 (39.2%); 
Sample of underpayments: Number (% total): 104 (60.5%); 
Sample of underpayments: Dollars (% total): $1,817,000 (52.2%). 

Entire improper payment occurred prior to fiscal year 2006; 
Sample of overpayments: Number (% total): 50 (23.6%); 
Sample of overpayments: Dollars (% total): $892,000 (60.8%); 
Sample of underpayments: Number (% total): 68 (39.5%); 
Sample of underpayments: Dollars (% total): $1,666,000 (47.8%). 

Source: GAO analysis of Labor data. 

[End of table] 

As previously discussed, this estimate likely understates the amount of 
overpayments identified during the fiscal year for two reasons. First, 
our estimate only includes overpayments that were finalized by March 
2007, when OWCP provided us with its 2006 debt universe. An additional 
$4.7 million in debts that had been declared by OWCP, but not yet 
finalized, were not included in our universe. Additionally, our 
estimate of improper underpayments is likely to be understated because 
not all underpayments are issued as direct payments. 

Analysis of OWCP's Accountability Reviews: 

We used OWCP's biennial internal program audits, called accountability 
reviews, to provide additional evidence about OWCP vulnerability to 
improper payments. These reviews are conducted on a district office 
basis, with samples drawn from each office's universe of claims files. 
To aggregate these results across all six district offices evaluated in 
fiscal year 2006, we weighted the errors that were identified in each 
office to account for differences in the size of universes across 
district offices. Specifically, we used the following formula to weight 
the results in each district office: (number of errors in the district 
office/sample size for district office) multiplied by (universe of 
items reviewed for district office /universe of items reviewed across 
all district offices). 

[End of section] 

Appendix II Comments from the Department of Labor: 

Note: GAO comments supplementing those in the report text appear at the 
end of this appendix. 

U.S. Department of Labor: 
Assistant Secretary for Employment Standards: 
Washington, D.C. 20210: 

February 8, 2008: 

Mr. Daniel Bertoni: 
Director, Education, Workforce, and Income Security Issues: 
United States Government Accountability Office: 
Washington, D. C. 20548: 

Dear Mr. Bertoni: 

Thank you for the opportunity to comment on the GAO draft report 
entitled Federal Workers' Compensation: Better Data and Management 
Strategies Would Strengthen Efforts to Prevent and Address Improper 
Payments (GAO-08-284). The report resulted from a study in which you 
reviewed the Office of Workers' Compensation Programs (OWCP) Division 
of Federal Employees' Compensation (DFEC) benefit payments for 
accuracy. Although we were pleased that the volume of improper payments 
under the Federal Employees' Compensation Act (FECA) estimated by GAO 
was very low (less than 0.75% of the more than $1.8 billion in wage 
loss compensation payments made during the year studied), even though 
erroneous payments made in other years were included in your estimate, 
we do have some concerns with the findings and conclusions. I would 
like to provide a few general comments on the overall report, followed 
by our specific responses to the study's findings: 

* It is the mission of the FECA program to promptly pay wage loss and 
medical benefits to injured Federal employees in an effort to minimize 
hardship. Accurate payments based on verified, complete, and accurate 
data are essential. However, there are instances where payment may 
result in over or underpayments, but these payments should not be 
considered improper, since adjustments are made once additional 
information becomes available. 

(See response 1.): 

* Unlike in prior GAO reviews,[Footnote 44] we were not provided case 
names and numbers. This greatly limited our ability to comment or 
verify the accuracy of the GAO's findings as they related to particular 
cases or circumstances, including a determination as to whether a 
payment was in fact improper. Our request for case numbers associated 
with specific errors was declined. Additionally, without the case names 
and numbers, we are unable to identify localized training needs based 
on specific problems that may have been attributable to a single 
office.[Footnote 45] 

(See response 2.): 

* As stated in the report, the GAO included erroneous payments 
identified in the year studied, but actually made in prior years to 
determine its estimate of the volume of improper payments under the 
FECA during the year studied. We believe this approach is not an 
accurate measure of the program's improper payment performance. 

(See response 3.): 

Notwithstanding the low volume of errors identified in the report and 
our inability to examine the underlying data, the Department of Labor 
(DOL) will follow up appropriately on the recommendations to better 
analyze and track the causes of improper payments under FECA and 
thereby further reduce their incidence. 

Our comments on specific findings of the study follow the order 
presented in the report. 

GAO Item: 0WICP Lacks an Effective Strategy for Managing the Risks of 
Improper FECA Compensation Payments 

"OWCP Does Not Emphasize Preventing, Detecting, or Recovering Improper 
Payments" 

The report is correct in its assertion that the FY 2007 program plan's 
twenty-one (21) performance goals established for the district offices 
"emphasize timely case management, but do not focus on payment 
accuracy." While these performance goals specifically address 
timeliness measures, OWCP's staff training, payment procedures (which 
require payment review and certification), and subsequent internal 
accountability review processes provide controls to ensure the 
prevention, timely detection, and recovery of improper payments. 
Additionally, the DOL's A-123 Reviews and the DOL Office of the 
Inspector General's annual audits further address qualitative issues, 
including a test of our internal control processes. 

(See response 4.): 

The DFEC Operational Plan for 2008 (presented in draft in September 
2007) did include a new measure of the timely processing of identified 
overpayments (both pending and preliminary), in keeping with the other 
timeliness measures found in the plan (which addresses a concern raised 
later in the report). 

(See response 5.): 

The report indicates that DOL does not use the results of the DFEC 
accountability review findings in its assessment of the program's risk 
of improper payments. We do not concur with this finding in that the 
independent auditors for both the DOL OIG SAS-70 audit and the DOL's A-
123 Review incorporate our accountability reviews in their assessment 
of the agency's internal controls. The effectiveness of the 
accountability review process is assessed during the course of the 
audits. Results of our accountability reviews are referenced in the 
audit reports and serve as a basis for findings and recommendations of 
the audit entity. For FY 2007, FECA received an unqualified audit 
opinion. 

(See response 6.): 

The GAO states that the accountability review's permissible error rate 
for initial payments is high (20%). However, the accountability review 
definition of errors is not comparable to the GAO's, thus the GAO's 
interpretation of the accountability review error rate is inaccurate. 
In the accountability review, an error is not only an incorrect 
payment, but also a procedural error. Thus, a single payment may have 
multiple errors attributed to it in the accountability review process, 
even if the payment was correct. Procedural factors may include 
incomplete documentation or certification of an otherwise proper 
payment. Therefore, error rates in the accountability reviews do not 
only reflect errors for actual payments. 

(See response 7.): 

The report stated that the program had failed to require some district 
offices to form corrective action plans when the accountability review 
revealed an initial payment rate below the acceptable standard. There 
are two explanations for this. First, it should be noted that there is 
always a meeting immediately after every accountability review between 
the National and district office upper management. This meeting is 
conducted to analyze the findings and to determine what corrective 
action plans are needed. In these meetings, 'management reviews other 
extraneous factors that may have seriously impacted the error rate - 
factors which may obviate the need for a corrective action on a failed 
item. For instance, district office management may have recognized a 
problem in the course of the review year and already implemented 
corrective measures. Secondly, findings from several items may be 
combined into a single corrective action plan, instead of creating an 
individual plan for each one. So, formal remedies may be in place, 
although not directly noted in the Initial Payments section of the 
final accountability review report. We will undertake more 
comprehensive documentation of corrective actions and the basis for 
establishing corrective action plans. 

(See response 8.): 

"OWCP Lacks the Information Needed to Accurately Assess the FECA 
Program's Risk of Improper Payments" 

We believe that the detailed internal accountability review process 
provides a good basis to assess the types and frequency of payment 
errors. In addition the statistical sampling performed for the annual 
improper payments reporting requirement also provides data as to 
payment errors. To further improve this process, we are developing ways 
to collect information in the Integrated Federal Employees' 
Compensation System (iFECS) for the analysis of potential erroneous 
payments. In the realm of overpayments, we are developing reason codes 
in the system so that we can track the various types of overpayments 
that occur, and the reasons for them. As for underpayments, one of the 
DFEC district offices is already conducting an audit of potential 
erroneous payments. The program plans to evaluate this audit and 
consider implementing it on a national level. Analysis of both 
underpayments and overpayments will be used to identify training needs 
in an effort to improve performance. With the availability of this 
data, we will consider establishing performance goals in this area. 

(See response 9): 

SSA Audit Report: 

The report cited a 2007 audit report from the Social Security 
Administration (SSA) Inspector General that found that "claimants 
failed to report approximately $12.6 million in wages to OWCP in 2004." 
What is not indicated in the GAO report is that in making this finding, 
the SSA/OIG did not review case files to determine if the wages in 
question did in fact create improper payments by the DFEC. SSA/OIG 
simply ran the DFEC payment system's case status codes against their 
database to identify apparent matches. However, their methodology did 
not take into account legitimate earnings or coding errors, which could 
only be determined by case review. For this reason, the SSA/OIG 
indicated that the $12.6 million was a universe of "potential" 
overpayments, and not an assertion of actual improper payments. The 
SSA/OIG audit was performed at the request of another agency and DFEC 
was not given the opportunity to review the data in detail or provide 
comment. 

(See response 10): 

Improper Payments Identified in 2006: 

The GAO has estimated a total of $13.3 million in improper payments 
identified in 2006, spanning multiple fiscal years. This figure was 
derived from a review of known underpayments and overpayments and not 
from random sampling of all payments. The report speculates that this 
figure is probably understated, since there were frequent delays in 
creating overpayment records in the system. We do not believe this 
statement is reflective of the current status of DFEC overpayments. The 
overpayments reviewed by the GAO spanned FY 2004, FY 2005, and FY 2006. 
During that period DFEC made extensive system changes, including the 
inception of the new iFECS in FY 2005. As a result of the normal 
problems associated with a system start-up, there were difficulties in 
identifying and tracking overpayments, resulting in longer than usual 
delays in creating overpayment records. However, most of these problems 
were corrected by the end of FY 2006, which has allowed the DFEC to 
shorten this timeframe. In fact, since iFECS now automatically 
identifies and tracks many types of overpayments (in particular, 
overpayments attributable to prior returns to work), this enables OWCP 
to act more promptly than before the new system was implemented. 

(See response 11): 

GAO Item: 0WCP 's Dependence on Unverified Information, Internal 
Errors, and System Limitations Leaves FECA Vulnerable to Improper 
Payments. 

"OWCP Relies on Claimants to Provide Key Eligibility Information" 
Claimant Notification of Return to Work: 

The report indicates that the DFEC relies on claimants themselves 
reporting their return to work, allowing for improper payments to be 
created. GAO is correct in that DFEC does take action when claimants 
report their return to work. However, DFEC obtains return to work 
information from injury compensation coordinators in employing agencies 
as well as nurses and vocational rehabilitation staff. Injury 
compensation specialists at the employing agencies report claimant 
returns to work directly to the DFEC (as noted in the report on page 
21). In addition, return to work status is monitored and reported by 
OWCP's field nurses and vocational rehabilitation specialists whose 
principal role is to assist with return to work. In the late 1990's 
OWCP began a two-pronged nurse intervention program that contracts with 
nurses to contact injured workers to facilitate medical care and assist 
with return to work issues. For claimants in vocational rehabilitation, 
rehabilitation professionals will have firsthand knowledge of a 
claimant's return to work. Both nurses and vocational rehabilitation 
contractors provide notification when an employee has returned to work 
or scheduled a return to work. 

(See response 12.): 

Further, the DFEC is working on creation of an electronic version of 
the discontinued Form CA-3. The program envisions an interactive 
component in the Agency Query System (AQS) where the employer can 
quickly report the return to work without requiring a paper form and 
mailing. The previously-used manual form was completed by employing 
agencies to formally notify DFEC when a claimant returned to work, but 
was discontinued because it was frequently not completed, or submitted 
too late to be of use. Until the implementation of the electronic 
process, the employing agencies will continue to rely on their injury 
compensation coordinators to report returns to work directly to the 
DFEC District Office. 

(See response 13.): 

Claimant or Beneficiary Deaths: 

As the report notes on page 22, there was a lapse in the SSA Death 
Match agreement that allows the DFEC to utilize SSA data to identify 
claimant deaths. That problem has since been rectified and matches were 
later conducted for some of the six month lapsed period, allowing the 
program to track claimant deaths in a timelier manner. The report also 
cites a finding in the same SSA/OIG study mentioned earlier that nearly 
$2 million in wage loss benefits were paid to deceased claimants in FY 
2004. Again, the SSA/OIG report acknowledges that the study was limited 
to a computer match which neither verified the beneficiaries' deaths 
nor included reviews of the individual cases to verify that 
overpayments actually occurred. We believe that the match included 
cases where we terminated compensation payments but erroneously failed 
to change our case status. As a result, we believe that the reported 
amount is overstated. As for tracking the death of survivor 
beneficiaries to avoid improper payments, the program has previously 
pledged to begin to collect these beneficiaries' Social Security 
Numbers and to input these into our system to enable crossmatch with 
the SSA death listings for these individuals. OWCP has been authorized 
by Congress (Public Law 103-333, 108 Stat. 2539, September 30, 1994) to 
require persons who file notices of injury and or claims for 
compensation under the FECA and its extensions to disclose SSNs. 
Consequently, applicable regulations concerning the filing of a notice 
of injury and claim for compensation, including 20 C.F.R. § 10.100 
(traumatic injury), 10.101 (occupational disease), and 10.105 (death), 
now expressly require the reporting of the injured worker's SSN. For 
death claims, the survivor must also disclose the SSNs of all survivors 
for whom benefits are claimed. 

(See response 14.): 

(See response 15.): 

Failure to Send Wage Data Requests on Every Claimant: 

In addition, the DFEC has also made recent efforts, not noted in the 
report, to improve the number of FERS offset deductions being made for 
Social Security benefits attributable to Federal service. A new process 
is being implemented to speed up the response time from SSA on FERS 
data requests made by the district offices. In addition, a project is 
now underway with the United States Postal Service to review their 
claims that may be eligible for FERS reductions. 

(See response 16.): 

"Errors by OWCP Claims Examiners Caused Overpayments and Underpayments" 

CE Errors/ Incorrectly Denied Payments: 

It is difficult to assess the magnitude of this problem, since, 
contrary to past practice, the DFEC has not been given the opportunity 
to view cases in which the GAO is citing errors. Without knowing for 
certain if these cases are in fact errors, or what type of error 
patterns may exist, a specific response is impossible. This is 
especially true in the analysis of "incorrectly denied payments." 
Reviewing a denied claim is a very technical undertaking, and one that 
the DFEC only gives to its more experienced and expert employees. In 
addition, assessment of such cases is subject to an interpretation of 
the current evidence, which may not be reflective of the circumstances 
of a case at the time the initial decision was done. 

(See response 17.): 

Nevertheless, the program is intent on improving claims examiner 
performance, and is creating new training modules to that end. One of 
the planned modules will deal specifically with establishing pay rates 
and initial payments, with the goal of improved accuracy. The modules 
will cover many aspects of the claims process, and will be available on-
line for either initial claims training or for updating the training of 
existing examiners. 

Lack of Eligibility Forms: 

It should be noted that the GAO review of this subject was based on the 
2006 internal accountability reviews conducted by the DFEC. The reviews 
focus on the work in a given district office over the prior twelve 
months, meaning that they would include cases from early 2005 through 
mid 2006. In a system update on March 5, 2006, the DFEC implemented 
substantial improvements to its Periodic Entitlement Review (PER) 
system. This is the application that is used to track claimant 
entitlement and ensure that the annual eligibility forms have been 
issued and tracked. As a result, the 2007 Financial Audit, conducted by 
Labor/OIG, showed considerable improvement in this area. We are 
confident that continued use of this tracking system will be an 
effective tool to reduce this type of improper payment. 

(See response 18.): 

"Some Overpayments Occur Because of Limitations in OWCP's Payment 
System" 

The payment process schedule that the DFEC has established with the 
U.S. Treasury was created to allow the Treasury to issue DFEC payments 
in a timely manner. To avoid claimant hardship, the Treasury schedules 
its check releases so that they arrive within a couple of days of the 
date of issuance. To accomplish this, the Treasury established the 
timeline that is currently followed. Recent discussions with Treasury 
on this matter indicate that we could only move back the schedule a 
single day if we still wish to release the claimant's payments in a 
timely manner equivalent to present service. However, if we did move 
the schedule back a day (allowing claims examiners one more day to 
react to a return to work) there would be a danger that a transmission 
or processing problem with the payment schedules (without a back up 
day) would create severe delays in payment issuance for all of our 
claimants. In the interest of having a fail-safe day, the program 
intends to leave the current schedule in place. 

(See response 19.): 

"Inaccurate Data From Employing Agencies Also Leads to Improper 
Payments" 

The OWCP has already discussed with the GAO its plans to provide 
employing agencies with more detailed compensation calculation 
information that would allow them to identify errors resulting from 
incorrect pay rate information. Specifically, the DFEC has agreed to 
accelerate a planned enhancement to the iFECS/Agency Query System, 
which will allow employing agency staff to access all the details of a 
given payment online. This should allow the agencies the ability to 
monitor payments and discover any flaws in the data they have 
submitted. It should be noted that agencies have always been provided 
the pay rate information on form CA 1049, but will now have the ability 
to cross check this against actual pay history in real time.

(See response 20.): 

GAO Item: OW/CP Does Not Ensure the Recovery of FECA Overpayments "OWCP 
Does Not Ensure the Timely Processing of Overpayments" 

We believe that this finding unfairly characterizes the program's 
current debt performance. As the concept of an automatic identification 
of a potential overpayment ("pending" debts) was newly introduced to 
the program with the inception of the iFECS system in 2005, claims 
personnel were unfamiliar with it and failed to make proper use of this 
aspect of the system. Though the report notes that during the period of 
review nearly half of the program's pending debts were over 180 days 
old (page 29), that figure is currently down to 26% and steadily 
improving. Although we note again that we are unable to comment on the 
particular cases reviewed by the GAO because the GAO declined to 
identify the specific cases they reviewed, we would note that many of 
the aged pending debts that we reviewed proved to be duplicates upon 
review. Because of the system changes, there were longer than usual 
delays in debt processing for 2005 and part of 2006. However, training 
with the district offices and continued improvements in the iFECS 
system have greatly improved the program-wide performance. 

(See response 21): 

"About 70 Percent of Overpayments Are Repaid, but OWCP Overlooks 
Opportunities to Recover Overpayments" 

Waiver: 

Despite the fact the SSA and the Department of Veterans Affairs (VA) 
use a much lower threshold for writing off small overpayments, lowering 
the current amount may not be feasible for DFEC. The waiver limits are 
set in an effort to weigh the costs associated with pursuing the 
overpayment against the potential benefits received. As the FECA 
statute requires a finding of whether the individual is with or without 
fault and thus FECA regulations require that a fault finding be made at 
the preliminary decision stage, we cannot automate the process. As a 
result, our overpayment process requires more up-front analysis by a 
claims person than may be needed by the other agencies cited. The 
current dollar amounts for waiver are set because of this need for 
early human intervention. 

(See response 22): 

Missed Opportunities on PR Payments: 

The program already has scheduled an enhancement to the iFECS 
compensation system that will prompt the user when a payment is being 
made on a claim that has an existing overpayment. There will also be a 
prompt if the claimant being paid has an overpayment in a separate 
case. The effect of these prompts will be to warn the claims staff to 
consider collection of outstanding overpayments, and should greatly 
reduce any missed opportunities for these types of collections. 

(See response 23): 

Wage Garnishment: 

The DFEC Procedure Manual outlines the various types of collection 
strategies that should be pursued when trying to collect on an improper 
payment. Specifically, PM §6-100-3(f) states that although wage 
garnishment is an available collection opportunity, it is preferable 
for the debt to be referred to the U.S. Treasury for servicing. Since 
the Treasury conducts both wage garnishment and the offset of any 
outgoing federal payments, they are better equipped to make these 
collections on behalf of the DFEC. 

(See response 24): 

OWCP is committed to continual improvement in payment accuracy and 
overpayment collection, while recognizing our priority of prompt 
payment of wage loss compensation to avoid hardship. We believe we have 
effective systems in place and enhancements being developed to ensure 
an even higher degree of payment accuracy. We will give your 
recommendations full consideration and we will continue to work to 
improve our automated systems and our employees' performance. 

Again, we appreciate the opportunity to review and comment on the 
report. 

Sincerely, 

Signed by: 

Victoria A. Lipnic

GAO's Response to Labor's Comments: 

The following are GAO's responses to the Department of Labor's comments 
on our draft report as outlined in Labor's February 8, 2008, letter. 

1. Labor commented that FECA's mission is to promptly pay wage loss and 
medical benefits to injured federal employees to minimize hardship, 
noting that accurate payments are essential. Labor also noted that, 
while payments can result in overpayments or underpayments, they should 
not be considered improper because adjustments are made once additional 
information becomes available. As stated our report, we used the 
definition of improper payment in the Improper Payment Information Act: 
any payment that should not have been made or was made in the wrong 
amount. We continue to believe that Labor needs to ensure that it 
strikes the appropriate balance between processing claims quickly and 
preventing improper payments. 

2. Labor asserted that its ability to comment on or verify the accuracy 
of GAO's findings with respect to specific cases was limited because we 
did not provide the case names and numbers associated with specific 
categories of improper payments. Because it is GAO's policy not to 
release our work papers before a review is completed, we did not 
provide the requested information to Labor. The previous review cited 
in Labor's comments involved a different set of circumstances in which 
GAO identified errors during its review of case files and worked with 
OWCP to confirm the errors. GAO also did not provide its work papers to 
Labor for that study. To verify the findings of our current report, 
OWCP could have reviewed the case files it provided to us for our 
review. 

3. Labor stated that we reviewed all improper payments identified in 
2006 without regard to when they actually occurred, describing our 
methodology as an inaccurate measure of the program's performance with 
regard to improper payments. In conducting our analysis, we determined 
that reviewing available data on improper payments that occurred in 
2006 would not provide an accurate assessment of the risk of the 
program to improper payments because OWCP does not identify some 
improper payments for a year or more after they occur. As a result, we 
focused on improper payments that were identified in 2006 without 
respect to when they occurred. We described the specific methodology we 
used in the report and detailed how our measurement of improper 
payments differs from Labor's. Further, while Labor's random sample of 
payments is consistent with OMB's guidance for developing an improper 
payment rate, it provides little information on the magnitude and 
causes of improper payments. Our review of improper payments identified 
by Labor in a particular year provides much more qualitative 
information on the magnitude and causes of improper payments that the 
agency can use to reduce the incidence of improper payments. We 
continue to believe that Labor should use its existing data and collect 
additional data on the causes of improper payments to use in 
identifying and addressing the most common causes of FECA improper 
payments. 

4. While OWCP has payment procedures, provides training to staff, and 
conducts biennial reviews to determine how well district offices follow 
established procedures, the lack of performance measures for payment 
accuracy leaves the FECA program at risk of improper payments. As we 
noted in the report, previous GAO work has shown that the risk of 
improper payments increases when payment timeliness goals are not 
balanced with an emphasis on payment accuracy. Further, some OWCP staff 
told us that they focused more on claims processing tasks that are 
tracked by OWCP, while placing a lower priority on processing improper 
payments. We continue to believe that the agency should establish 
performance goals related to payment accuracy. 

5. Labor did not provide us with its fiscal year 2008 Division of 
Federal Employees' Compensation (DFEC) Operational Plan during our 
review. In response to this comment, we asked for a copy of the plan 
and Labor provided it. We added the information about the new 
performance measure to the report. 

6. Labor's Office of Inspector General's (OIG) SAS-70 audit for 2006 
assessed the FECA program's internal controls. As part of this audit, 
the OIG inspected OWCP's accountability reviews with respect to payment 
accuracy and determined that the reviews were properly conducted. 
However, the OIG's SAS-70 audit and Labor's A-123 review did not 
incorporate the results of the accountability reviews with respect to 
payment accuracy into their findings. 

7. We added language to the report to recognize that OWCP's 
accountability reviews include procedural errors that may not result in 
an improper payment. We separately reviewed payment and procedural 
errors and found that 14 percent of the payments sampled in OWCP's 2006 
accountability reviews had calculation errors rather than procedural 
errors. Further, if district offices are not following the agency's 
payment procedures in over 20 percent of its cases, the program is 
still at risk of improper payments, even if some procedural errors do 
not create an improper payment. 

8. During our review, OWCP officials did not tell us that findings from 
several review items could be combined into a single corrective action 
plan. We added this information to our report. 

9. Labor's statistical sampling is intended to develop an improper 
payment rate and provides little useful information about the program's 
risks of improper payments. While the results from its accountability 
reviews could potentially provide data on program risks, OWCP officials 
gave us no indication that that they using the results for this 
purpose. While OWCP officials did not share with us during our review 
their plans to develop reason codes for improper payments, we support 
the steps they are proposing to better track improper payments, steps 
that are consistent with our recommendations. We also encourage OWCP to 
develop performance goals for reducing improper payments based on these 
improved data. 

10. While the SSA OIG did not review individual cases to verify that an 
improper payment occurred, it focused on a specific population of FECA 
payment recipients who should not be earning any wages--individuals who 
were totally disabled for the entire year. Conducting automated data 
matches is an important internal control for means-tested benefit 
programs and a critical first step in the process of identifying 
potential improper payments that should be further investigated. 
Without such data matches, Labor cannot begin to identify overpayments 
made when claimants do not report earnings. In addition, if the 
information that the SSA OIG reviewed was inaccurate because OWCP 
improperly coded some of its claims, OWCP should take steps to ensure 
that all claims are properly coded. 

11. In our report, we note that OWCP's new data system provides better 
data for tracking the status of overpayments and that OWCP has taken 
steps to address data reliability issues that arose during the initial 
implementation of the new system. We agree that the new system should 
more quickly identify some overpayments, such as those that occur when 
claimants return to work. However, from our reviews of the claims 
files, we found other types of overpayments that can take a long time 
to identify--such as errors in the wage rates provided by the 
claimants' agencies to OWCP--that will not be identified by the new 
system. Further, OWCP data and interviews with some claims examiners 
indicated that significant delays occurred in processing overpayments 
after they were identified. As noted in our report, managing 
overpayments has not been a priority for OWCP in the past. However, we 
believe that improvements in OWCP's performance are the result of its 
recent increased emphasis on improper payments. 

12. We clarified in the report that nurses and vocational 
rehabilitation staff can also notify OWCP that a FECA claimant returned 
to work. However, because nurses or vocational rehabilitation staff are 
not assigned to every claim, we believe the employing agency is in the 
best position to notify OWCP when a claimant returns to work, and OWCP 
should require agencies to do so in a timely manner. 

13. Although, during our review, OWCP officials did not tell us about 
their plan to be notified electronically when claimants return to work, 
we support OWCP's efforts to develop systems to facilitate timely 
notification when claimants return to work. We recognize, however, that 
agencies may not always provide this information promptly to OWCP. As a 
result, we recommended that OWCP provide incentives for agencies to 
notify it quickly. For example, OWCP tracks how quickly agencies submit 
required claims forms and could similarly track how quickly agencies 
notify it when claimants return to work. 

14. While we recognize that the SSA OIG did not review individual cases 
to verify that an improper payment occurred, some of these cases may 
represent actual overpayments. However, without conducting regular data 
matches with SSA's death records for all individuals receiving FECA 
payments, OWCP cannot easily identify such overpayments. As with 
matches of earnings data, conducting automated data matches with death 
records is an important internal control and a critical first step in 
the process of identifying potential improper payments that should be 
further investigated. In addition, if the information that the SSA OIG 
reviewed was inaccurate because OWCP improperly coded some of its 
claims, OWCP should take steps to ensure that all claims are properly 
coded. 

15. Because OWCP officials told us during our review that the FECA 
program does not collect the SSNs of individuals receiving survivor 
death benefits, we contacted an OWCP official to discuss this comment. 
He clarified that the form survivors use to apply for death benefits 
does not request their SSNs. As a result, OWCP cannot currently include 
survivors in its data match with SSA's death records. However, the 
official also stated that, when the death benefit application form 
expires in 2010, OWCP plans to propose that the form be revised to 
include the survivor's SSN, a proposal subject to OMB approval. We 
continue to believe that OWCP should take all needed steps to include 
survivors who claim death benefits in its monthly data matches with 
SSA's death records. 

16. OWCP officials told us the agency is taking steps to improve the 
number of FERS offset deductions being made for Social Security 
benefits but the specific steps being taken were not completely clear. 
We requested a copy of a letter that the officials told us they had 
prepared related to this initiative, but OWCP did not provide it to us, 
so we could not confirm the steps being taken. OWCP officials also 
mentioned that the agency was taking steps to obtain FERS information 
more quickly from SSA. While this is a useful step, our concern is that 
some claims examiners may not be aware of this issue or how to 
implement a FERS offset deduction. 

17. We agree that making determinations about claims files requires 
some judgment and interpretation. To ensure that appropriate and 
consistent interpretations were used, GAO took the following steps in 
reviewing case files. First, we established case file review protocols 
with specific criteria about how to treat different types of 
circumstances. For example, in reviewing underpayments created when a 
denied claim was overturned on appeal, the protocol instructed 
reviewers not to consider it an underpayment if the decision was 
overturned based on new information that was not available when the 
original decision was made. We pretested the protocol and revised it in 
response to the pretest results. Our protocol also called for a second 
GAO analyst to independently verify the case file responses recorded by 
the initial analyst. If the two analysts did not agree, they discussed 
the case until they came to consensus on the appropriate response. In 
addition, GAO supervisors and technical staff reviewed the results of 
the analysts' work. Finally, we support OWCP's plans to develop a 
training program to improve payment accuracy, plans that are consistent 
with our recommendations. 

18. We based our findings on the most recent data and reports available 
at the time of our review. We believe that Labor's efforts to ensure 
that claimants submit annual eligibility forms will help decrease the 
risk of improper payments. 

19. Given the large number of overpayments created by the current 
payment schedule established by OWCP, we continue to believe that OWCP 
should look for opportunities to reduce the time needed to process its 
automated monthly FECA payments. 

20. In our report, we describe OWCP's plans to provide employing 
agencies with more detailed FECA payment information. We support OWCP's 
efforts in this area, which are consistent with our recommendations. 

21. We used data from OWCP's September 2007 debt aging report, which 
was the most recent data we could incorporate into our report and that 
OWCP officials told us was reasonably reliable. If more recent data 
indicate that OWCP's performance in processing overpayments is 
improving, we commend OWCP for its efforts. As stated previously, we 
believe that OWCP's performance may be improving because of its 
increased emphasis on managing improper payments, and we encourage the 
agency to continue these efforts. 

22. We recognize that agencies use different computer systems and that 
the FECA program has different requirements from the programs 
administered by SSA and VA. However, as OWCP continues to improve its 
overpayment processing capabilities, we believe that it should review 
the thresholds used by these and other agencies to identify potential 
opportunities to reduce its waiver thresholds. 

23. OWCP officials did not inform us of these planned enhancements to 
their data system during our review. We added this information to our 
report. We support the steps OWCP is proposing to improve its debt 
collection activities, steps that are consistent with our 
recommendations. 

24. In our report, we describe OWCP's policy and procedures with 
respect to wage garnishment. We agree that Treasury may be better 
equipped to garnish the wages of FECA claimants with overpayments, 
especially if claims examiners process claims quickly and refer them to 
Treasury as soon as possible. However, if OWCP does not process and 
finalize overpayments quickly, opportunities to garnish wages may be 
lost since a claimant may no longer be working at a federal agency by 
the time the overpayment is referred to Treasury. We encourage OWCP to 
continue its efforts to process overpayments more promptly and refer 
uncollected overpayments to Treasury as quickly as possible. 

[End of section] 

Appendix III: GAO Contact and Acknowledgments: 

GAO Contact: 

Daniel Bertoni, Director, (202) 512-7215, bertonid@gao.gov: 

Staff Acknowledgments: 

Revae Moran, Assistant Director, and Michelle St. Pierre, Analyst-in- 
Charge, managed this assignment, and Jordan H. Holt and M. Ellen Phelps 
Ranen made key contributions throughout the assignment. In addition, 
Carla Craddock and Keith Howey assisted with the audit work; Carla 
Lewis provided advice on financial audit requirements and improper 
payment issues; Daniel Schwimer provided legal assistance; Charles 
Willson provided writing assistance; and Laurie Hamilton, Jean McSween, 
Karen O'Conor, and Beverly Ross assisted with the methodology and 
statistical analyses. 

Footnotes: 

[1] Pub. L. No. 107-300, (2002). 

[2] For example, see (1)Department of Labor, Office of Inspector 
General, Mechanisms Used to Identify Changes in Eligibility Are 
Inadequate at the FECA District Office in Jacksonville, Florida, 04-07- 
004-04-431 (Washington, D.C.: Sept. 28, 2007) and (2) Social Security 
Administration, Office of the Inspector General, Federal Employees 
Compensation Act: A Nationwide Review of Federal Employees Who Received 
Compensation for Lost Wages for Periods When "Earned Wages" Were 
Reported on the Social Security Administration's Master Earnings File, 
A-15-06-16037 (Baltimore, Md.: May 18, 2007). 

[3] 5 USC §8101, et seq. 

[4] Several mixed-ownership government corporations, such as USPS and 
the Tennessee Valley Authority, are required to provide additional 
funds to cover their "fair share" of the costs for administering the 
program for their employees. See 5 USC §8147(c). 

[5] FECA claimants are not required to retire at a certain age and can 
continue to receive FECA wage-loss-compensation payments for as long as 
they remain eligible. 

[6] For claimants covered under the Civil Service Retirement System, 
none of their SSA retirement benefits are based on federal service. 
Therefore, their FECA benefits are not reduced by the amount of SSA 
retirement benefits that they receive. However, for FECA claimants 
covered under the Federal Employees Retirement System, a portion of 
their SSA retirement benefits is based on their federal service. The 
Civil Service Retirement System was replaced by the Federal Employees 
Retirement System, and almost all federal employees hired after 1983 
are covered by the Federal Employees Retirement System. 

[7] Minimum and maximum monthly payments are set by law (5 USC 
§8112(a)). The minimum monthly FECA payment is 75 percent of the basic 
monthly pay for a GS-2, step 1 employee ($18,700 per year as of Jan. 1, 
2007). The maximum monthly FECA compensation payment cannot exceed 75 
percent of the basic monthly pay for a GS-15, step 10 employee 
($121,000 per year as of Jan. 1, 2007). 

[8] OWCP refers to claimants who receive automatic monthly payments as 
being on the "periodic roll" and those who receive manually generated 
payments as being on the "daily roll." 

[9] In general, OWCP continues to pay claimants the difference between 
their current salary and the salary they were earning at the time of 
their injury for as long as this difference exists and their medical 
work restrictions remain the same. OWCP would not continue to pay this 
difference for claimants who quit their job without good cause (for 
example, if they quit because they did not like the work hours). 

[10] Office of Management and Budget, OMB Circular No. A-11 (2002), 
Appendix C, "Section 57-5." However, OMB Circular No. A-123 provides 
that agencies with this high-risk designation are not permanently 
subject to the improper payment reporting requirements if the program 
has documented a minimum of 2 consecutive years of improper payments 
that are less than $10 million annually. 

[11] In fiscal year 2006, the FECA program made over $1.8 billion in 
wage-loss-compensation payments and paid $668 million for claimant 
medical and rehabilitation services, for a total of approximately $2.5 
billion in benefits. 

[12] GAO, Executive Guide, Strategies to Manage Improper Payments: 
Learning from Public and Private Sector Organizations, GAO-02-69G 
(Washington, D.C.: October 2001). 

[13] GAO, Executive Guide, Strategies to Manage Improper Payments: 
Learning from Public and Private Sector Organizations, GAO-02-69G 
(Washington, D.C.: October 2001). 

[14] Department of Labor, Office of Inspector General, Mechanisms Used 
to Identify Changes in Eligibility Are Inadequate at the FECA District 
Office in Jacksonville, Florida, 04-07-004-04-431 (Washington, D.C.: 
Sept. 28, 2007). 

[15] Labor's assessment of the FECA program's risk of improper payments 
is based on a statistical estimate of the amount of improper payments 
made by the program during the fiscal year. When evaluating the 
program's internal controls, outside auditors verify that OWCP 
conducted accountability reviews but do not factor the results of the 
reviews into their evaluation. 

[16] In reviewing initial payments, a sampled payment that was not 
properly certified would be considered an error, even if the actual 
payment was correct. 

[17] For three of these district offices, virtually all of the errors 
were payment calculations. For one office, one-third of the errors were 
payments calculations, while most of the remaining errors involved 
discrepancies between the pay rate provided by the employing agency and 
the pay rate used for the FECA calculations. For the last office, 
almost all the errors were due to the fact the case file lacked 
sufficient information to determine if the payment was calculated 
correctly. 

[18] GAO, Executive Guide, Strategies to Manage Improper Payments: 
Learning from Public and Private Sector Organizations, GAO-02-69G 
(Washington, D.C.: October 2001). 

[19] The Improper Payment Information Act requires federal agencies to 
annually estimate the amount of improper payments in programs that 
agencies have identified as susceptible to improper payments. OMB 
guidance requires agencies to obtain this estimate using statistically 
valid techniques. Labor based its estimate on a review of a sample of 
102 wage-loss-compensation payments made that year in which it found 1 
overpayment that totaled $228. From this review, Labor projected that 
the FECA program made $703,000 in improper payments in fiscal year 
2006. This estimate is intended to be a gross total of overpayments and 
underpayments in the program. In fiscal year 2007, Labor estimated that 
the FECA program made $2.6 million in improper payments. 

[20] Social Security Administration, Office of the Inspector General, 
Federal Employees Compensation Act: A Nationwide Review of Federal 
Employees Who Received Compensation for Lost Wages for Periods When 
"Earned Wages" Were Reported on the Social Security Administration's 
Master Earnings File, A-15-06-16037 (Baltimore, Md.: May 18, 2007). 

[21] To ensure consistency in the error rates reported by the agencies, 
OMB requires agencies to provide an improper payment rate based on a 
statistical sample of payments projected to the universe of payments 
made that year. OMB does not require agencies to include the amount of 
improper payments identified each year in their estimates. Following 
federal guidance, Labor's 2006 estimate is of the dollar value of 
improper payments made during the fiscal year and does not include any 
payments from previous fiscal years (such as 2004 or 2005) that OWCP 
identified as improper in fiscal year 2006. 

[22] In reviewing claims files from OWCP's list of the $9.9 million in 
overpayments identified and finalized by OWCP in 2006, we found that it 
included overpayments that we did not consider to be improper, as well 
as overpayments from time periods outside our purview. Therefore, in 
order to estimate the magnitude of improper overpayments identified by 
OWCP in 2006, we reviewed the claims files for a sample of potential 
overpayments that it identified that year. Similarly, because OWCP's 
data systems do not have the ability to distinguish underpayments from 
other payments made, we reviewed the claims files for a sample of 
potential underpayments in order to estimate the magnitude of improper 
underpayments identified by the agency that year. Based on the results 
of both of these reviews, we estimated that OWCP identified 
approximately $13.3 million in improper payments in 2006. The 95 
percent confidence interval for this estimate ranged from $10.4 million 
to $16.2 million. Whereas Labor estimated the magnitude of improper 
payments that OWCP made in fiscal year 2006, we estimated the magnitude 
of improper payments that OWCP identified during the fiscal year. 

[23] OWCP provided a list of all debts identified in the FECA program 
in fiscal year 2006, which included potential overpayments. From the 
information provided, we excluded debts that (1) were not in final 
determination or terminated status or (2) were unrelated to wage-loss- 
compensation payments. From this revised data, we selected a random 
sample of wage loss compensation overpayments and reviewed the claims 
file for each debt to identify the cause of the overpayment and its 
recovery status. After reviewing the claims files, we excluded 
overpayments that (1) were not improper payments or (2) were repayments 
of debts identified prior to our fiscal year 2006 time frame. After 
completing our analysis, we estimated the total dollar amount of 
improper overpayments identified by OWCP in 2006, and the percentage of 
improper overpayments that were attributable to different causes. See 
appendix 1 for additional information on our methodology. 

[24] Because OWCP was unable to identify its universe of underpayments, 
we sampled from a universe of different types of payments, including 
reimbursements when claimants were underpaid. Given this limitation in 
the universe, we could not estimate both the magnitude of underpayments 
and the frequency of different causes using a reasonable sample size. 
We designed a sampling methodology to allow us to develop a total 
dollar value estimate of underpayments and were therefore unable to 
estimate the prevalence of causes of underpayments with sufficient 
precision. See appendix I for additional information on our 
methodology. 

[25] Although the OWCP headquarters officials we interviewed were 
unsure when this form was discontinued, a DOD official told us OWCP 
stopped using the form in 1999. 

[26] OWCP also assigns nurses and vocational rehabilitation staff to 
work with certain claimants and these staff can also notify OWCP when a 
claimant returns to work. 

[27] If a claimant dies as a result of a work-related injury or 
illness, certain survivors, such as the claimant's spouse, children, 
and dependent parents, are eligible for FECA compensation (known as 
death benefits). The law specifies the percentage of a claimant's 
monthly salary that is due to each eligible survivor, but monthly 
payments for all survivors cannot exceed 75 percent of the claimant's 
monthly salary. 

[28] Social Security Administration, Office of Inspector General, 
Federal Employees' Compensation Act: A Nationwide Review of Federal 
Employees Who Received Compensation for Lost Wages for Periods When 
"Earned Wages" Were Reported on the Social Security Administration's 
Master Earnings File, A-15-06-16037 (Baltimore, Md.: May 18, 2007). 
Because the SSA Inspector General did not review the specific cases to 
confirm that an improper payment had been made, it described these 
cases as potential overpayments. 

[29] In fiscal year, 2007, this data match identified 847 FECA 
claimants who were listed in SSA's death records. 

[30] Without a systemic data match, OWCP must obtain a release from the 
claimant and send this release to SSA along with a request for earnings 
statements for the period covered in the release--a cumbersome and time 
intensive process. 

[31] Social Security Administration, Office of Inspector General, 
Federal Employees' Compensation Act: A Nationwide Review of Federal 
Employees Who Received Compensation for Lost Wages for Periods When 
"Earned Wages" Were Reported on the Social Security Administration's 
Master Earnings File, A-15-06-16037 (Baltimore, Md.: May 18, 2007). 
Because the SSA Inspector General did not review the specific cases to 
confirm that an improper payment had been made, it described these 
cases as potential overpayments. 

[32] As of fiscal year 2004, 70 percent of civilian federal employees 
were enrolled in the Federal Employees Retirement System, which covers 
employees hired since 1984. Thirty percent were enrolled in the Civil 
Service Retirement System, which covers employees hired prior to 1984. 

[33] The 95 percent confidence interval for this estimate ranged from 
26 to 55 days. 

[34] Because of the manner in which payments are recorded, we did not 
capture some underpayments that occurred when OWCP inappropriately 
denied or terminated wage loss payments. Our estimate only pertains to 
the underpayments we were able to identify. See appendix I for more 
information on the methodology for reviewing underpayments. 

[35] The 95 percent confidence interval for this estimate ranged from 
$660 to $1,162. 

[36] Employing agencies receive copies of letters sent by OWCP to 
claimants after their first automatic monthly payment has been issued. 
These letters list the amount of insurance premiums deducted but do not 
include specific information about how claimants' pay rates are 
calculated. Further, after this initial letter, agencies do not receive 
copies of subsequent benefits statements sent by OWCP to claimants. 

[37] GAO, Welfare Benefits: Potential to Recover Hundreds of Millions 
More in Overpayments, GAO/HEHS-95-111 (Washington, D.C; June 20, 1995) 
and GAO, Benefit Overpayments: Recoveries Could Be Increased in the 
Food Stamp and AFDC Programs, GAO/RCED-86-17 (Washington, D.C.: Mar. 
14, 1986). 

[38] When a potential overpayment is first identified, OWCP categorizes 
it as a pending overpayment. Once its existence and amount has been 
confirmed, a preliminary notice is sent to the claimant and the 
overpayment is then categorized as a preliminary overpayment. OWCP 
officials told us that most pending overpayments are confirmed as 
actual overpayments. 

[39] This is based on our review of 27 overpayments in which (1) OWCP 
issued both a preliminary and final overpayment notice to the claimant, 
and (2) the claimant did not request an appeal in response to the 
preliminary overpayment notice. The confidence interval for our 
estimate of the mean days between the preliminary and final notice is 
44 days to 85 days. 

[40] The claims files we reviewed were only those that OWCP had waived 
or finalized, i.e., issued both the preliminary and final notices of 
overpayment to claimants. We do not know the recovery status of 
overpayments that OWCP had not finalized at the time of our review. 
These overpayments may involve more complicated issues and therefore 
may have different recovery outcomes. 

[41] The confidence interval for the $200 and under category is 66 
percent to 87 percent, while the confidence interval for overpayments 
between $200 and $700 is 27 percent to 49 percent. 

[42] For those few cases in which an overpayment notice is manually 
generated, the Supplemental Security Income program attempts to collect 
any overpayment over $10. 

[43] We excluded preliminary overpayments because they could have 
outstanding appeals and might be overturned. We excluded pending 
overpayments because of concern that some should be voided. (Pending 
overpayments may be automatically generated by OWCP's computer system. 
OWCP officials informed us that some claims examiners did not know how 
to process these overpayments, and may have created and processed 
duplicate records.) Suspended overpayments were excluded because they 
are associated with pending legal action. 

[44] See GAO-03-158R Postal Service Workers' Compensation Claims, 
Letter of Shelby Hallmark, pp 25-26. [hyperlink, 
http://www.gao.gov/new.items/dO3158r.pdf.] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability.  

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "Subscribe to Updates."  

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to:  

U.S. Government Accountability Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548:  

To order by Phone: 
Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061:  

To Report Fraud, Waste, and Abuse in Federal Programs:  

Contact:  

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470:  

Congressional Relations:  

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548:  

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: