This is the accessible text file for GAO report number GAO-08-112 
entitled 'Veterans' Affairs: Improved Management Would Enhance VA's 
Pension Program' which was released on February 14, 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 Requesters: 

United States Government Accountability Office: 

GAO: 

February 2008: 

Veterans' Benefits: 

Improved Management Would Enhance VA's Pension Program: 

Veterans' Benefits: 

GAO-08-112: 

GAO Highlights: 

Highlights of GAO-08-112, a report to congressional requesters 

Why GAO Did This Study: 

In 2006, the Department of Veterans Affairs (VA) paid about $3.5 
billion in means-tested pension benefits to over 500,000 veterans and 
survivors. GAO was asked to review the management of VA pension 
program. This report assesses (1) the characteristics and trends in 
size of the current pensioner population, (2) the policies and 
procedures VA has in place to ensure that initial pension eligibility 
decisions are well managed, and (3) the procedures VA has in place to 
ensure that pensioners continue to receive the proper benefit payments 
on an ongoing basis. Our study included reviews of agency policies, 
procedures, and internal controls; site visits to 4 of VA’s 57 regional 
offices and all three of its pension maintenance centers; and a 
selected file review of new claims at three locations. 

What GAO Found: 

In 2006, most of the over 500,000 VA pensioners had nonpension incomes 
well below the federal poverty level, were beyond retirement age, and 
had multiple impairments, and the population has been decreasing in 
number. The average annual reported income of these pensioners, 
excluding their VA pensions, was less than $5,000. The average age of 
VA pensioners was 70. More than 80 percent had no spouse or dependent 
children. Three-fourths of veteran pensioners had multiple impairments. 
After reaching a peak of almost 2 million in 1978, the overall size of 
the pensioner population has gradually decreased, although the number 
of pensioners from more recent service periods has been increasing. 

VA policies and procedures are not sufficient to ensure sound decisions 
on new pension claims. Unlike other federal agencies with similar 
income-based programs, VA largely does not independently verify the 
accuracy of financial information provided by claimants to support 
initial pension program eligiblity. In addition, the guidance used by 
staff to make pension eligibility decisions is not always current or 
clear. Further, VA’s quality assurance review process for initial 
claims does not select a sufficient number of pension cases to ensure 
the accuracy of pension claims decisions. Finally, VA does not 
adequately evaluate training for pension staff. 

VA procedures for assessing whether pensioners continue to receive the 
proper benefits have significant limitations. Although the agency 
requires pensioners to report changes that might affect their pensions, 
VA does not require documentation such as bank or asset statements when 
pensioners report financial changes. Also, a key data match operation 
with the Internal Revenue Service is not conducted in a timely or 
efficient manner. Finally, despite millions of dollars in improper 
pension payments made each year, VA lacks a system to monitor and 
analyze their causes. 

Figure: Age Groupings of VA Pensioner Population, 2006: 

This figure is a combination bar chart showing age groupings of VA 
pensioner population, 2006. The X axis represents the age, and the Y 
axis represents the pensioners. 

[See PDF for image] 

Source: Fiscal year 2006-2008 budget submissions. 

[End of figure] 

What GAO Recommends: 

This report contains six recommendations to improve VA’s management of 
the pension program. Specifically, VA should take steps to make more 
accurate and timely pension eligibility and payment decisions, improve 
its quality assurance of initial claims, incorporate feedback in its 
training, improve the effectiveness of pension eligibility processes, 
update its data match threshold, and evaluate the causes of improper 
pension payments. 

In its comments on a draft of this report, VA agreed in part or fully 
with our recommendations, though raised concerns with some options to 
help implement several of them. 

To view the full product, including the scope
and methodology, click on [hyperlink, http://www.GAO-08-112].
For more information, contact Daniel Bertoni at (202) 512-7215 or 
bertonid@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

The VA Pension Population Is Characterized by Low Income, Older Ages, 
Multiple Impairments, and Decreasing Numbers: 

For Initial Pension Claims, VA Policies and Procedures Do Not Always 
Ensure Well-Supported Decisions: 

For Ongoing Pension Cases, VA Procedures and Controls Do Not Adequately 
Ensure That Benefits Are Proper: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and GAO Response: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Pension Eligibility Determination: 

Appendix III: Comments from the Department of Veterans Affairs: 

Appendix IV: GAO Contact and Staff Acknowledgments: 

Related GAO Products: 

Tables: 

Table 1: Estimated Amount of Annual Nonpension Income by Source and 
Age: 

Table 2: Decrease in Number of Veteran and Survivor Cases by Pension 
Program, Fiscal Years 2000-2006: 

Table 3: Percentage Change in the Number of Pensioners by Service 
Period, Fiscal Years 2000-2006: 

Table 4: VA Steps to Process Eligibility Verification Reports: 

Table 5: VA Computer Matching Procedures: 

Table 6: VA Estimates of Pension Improper Payments, Fiscal Years 2003-
2006: 

Table 7: VA Regional Offices Visited and Disposition of Pension Case 
Files Reviewed: 

Figures: 

Figure 1: Age Groupings of Veterans and Survivors in 2006: 

Figure 2: Household Composition of VA Pensioners in Fiscal Year 2006: 

Figure 3: Types of Impairments among Veteran Pensioners by Age: 

Figure 4: Reasons for Program Termination, January 2007: 

Figure 5: Pension Eligibility Criteria for Veterans and Survivors: 

Abbreviations: 

BDN: Benefit Delivery Network: 

BOP: Bureau of Prisons: 

C&P: Compensation and Pensions: 

EVR: Eligibility Verification Report: 

IG: Inspector General: 

IRS: Internal Revenue Service: 

IU: Individual Unemployability: 

IVM: Income Verification Match: 

NDNH: National Directory of New Hires: 

PMC: pension maintenance center: 

RR: Railroad Retirement: 

RRB: Railroad Retirement Board: 

SSA: Social Security Administration: 

SSI: Supplemental Security Insurance: 

SSN: Social Security number: 

VA: Department of Veterans Affairs: 

VBA: Veterans Benefit Administration: 

United States Government Accountability Office: 

Washington, DC 20548: 

February 14, 2008: 

The Honorable Richard Burr: 
Ranking Member: 
Committee on Veterans' Affairs: 
United States Senate: 

The Honorable Larry E. Craig: 
United States Senate: 

The Department of Veterans Affairs (VA) pension program is intended to 
provide a minimum level of economic security for veterans with 
financial need. It is available to those low-income veterans who are at 
least 65 years old or have disabilities unrelated to their military 
service, as well as to their surviving spouses and dependent children. 
In fiscal year 2006, the program paid about $3.5 billion in benefits to 
over 500,000 pensioners. In administering the program, VA is required 
to assess the continued eligibility of pensioners to determine if 
changes in family income and assets, or number of dependents, might 
warrant payment adjustment or termination of the pension. Although the 
VA Pension Benefit program is a relatively small program, it has a 
critical role in supporting disabled and aging veterans and their 
survivors who depend on the program to provide a minimum level of 
economic security. Providing this benefit honors their service to their 
country and provides some financial assurance to those in need. 

The Senate Veterans' Affairs Committee asked GAO to (1) determine the 
characteristics and trends in size of the pensioner population, (2) 
assess the policies and procedures VA has in place to ensure that 
initial pension eligibility decisions are well managed, and (3) assess 
the procedures VA has in place to ensure that pensioners continue to 
receive the proper benefit payments. 

Pensions are one of the many benefits available to qualified veterans 
through VA. Other VA benefits include medical care, educational 
resources, and compensation for disabilities sustained or worsened as a 
result of active duty military service. VA actually administers three 
pension programs,[Footnote 1] and we included pensioners from all three 
programs when analyzing trends in population size, as this allowed for 
a more thorough understanding of changes within different war cohorts 
over time. The remainder of the report, however, focuses on the 
Improved Pension program, which is the largest and the only one still 
open to new members. To determine the characteristics and size of the 
pensioner population, we analyzed data from VA's Benefit Delivery 
Network (BDN), VA budget reports, and other reports through fiscal year 
2006. To address our remaining objectives, we reviewed relevant laws, 
guidance, procedures, and internal controls that VA uses to ensure the 
soundness of pension benefit decisions. We also analyzed VA's internal 
control policies and performance reports. We visited 4 of VA's 57 
regional offices, located in Boston, Milwaukee, Providence, and St. 
Paul. We selected these sites based on variations in size and 
geographic locations. We also visited VA's three pension maintenance 
centers (PMC), located in Philadelphia, Milwaukee, and St. Paul, and 
the Debt Management Center in St. Paul. We interviewed VA officials and 
staff at these sites as well as officials at VA Central Office in 
Washington, D.C. We also conducted case file reviews in three 
locations--Milwaukee, Providence, and St. Paul--to verify the adequacy 
of documentation in support of initial pension decisions. For more 
details on our scope and methodology, see app. I. We conducted our 
review from November 2006 to February 2008, in accordance with 
generally accepted government auditing standards. Those standards 
require that we plan and perform the audit to obtain sufficient, 
appropriate evidence to provide a reasonable basis for our findings and 
conclusions based on our audit objectives. We believe that the evidence 
obtained provides a reasonable basis for our findings and conclusions 
based on our audit objectives. 

Results in Brief: 

In 2006, most of the approximately 500,000 VA pensioners had nonpension 
incomes well below the federal poverty level, were beyond retirement 
age, and had multiple impairments, and the population has been 
decreasing in number. The average annual income of these pensioners, 
excluding their VA pension, was less than $5,000--an amount well below 
the 2006 federal poverty level of $9,800 for an individual, and most 
had few assets and limited education. Additionally, the average age of 
VA pensioners was 70, with approximately 60 percent over age 65 and 
less than 20 percent under age 55. More than 80 percent had no spouse 
or dependent children, and approximately one-third were surviving 
spouses who live alone. Three-fourths of veteran pensioners had 
multiple impairments, and about a third were housebound or needed 
assistance with dressing themselves or other activities of daily 
living. After reaching a peak of almost 2 million in 1978, the overall 
size of the pensioner population has gradually decreased to about 
500,000 today, although the number of pensioners from more recent 
service periods has been increasing. 

VA policies and procedures are not sufficient to ensure sound decisions 
on new pension claims. Unlike other federal agencies with similar 
income-based programs, VA largely does not independently verify the 
accuracy of financial information provided by claimants to support 
initial pension program eligibility, a fact that makes the program 
vulnerable to improper payments. In addition, the guidance used by 
staff to make pension eligibility decisions, which is under revision 
and dispersed across several sources, is not always current or clear. 
Further, VA's quality assurance review process for initial claims does 
not select a sufficient number of pension cases to ensure the accuracy 
of pension claims decisions. Finally, VA does not adequately evaluate 
the training for pension staff. For example, VA does not consistently 
collect feedback from participants at the end of a training course. 

VA procedures for assessing whether pensioners continue to receive the 
proper benefits have significant limitations because VA does not 
require pensioners to submit financial documentation, conducts untimely 
and inefficient verification of pensioners' incomes and assets, and 
lacks a system for identifying and reducing improper pension benefits. 
Although the agency requires all pensioners to submit documentation for 
nonfinancial changes, such as for marriages or deaths, it does not 
require documentation such as bank or asset statements when pensioners 
report financial changes. Also, while the agency does verify certain 
pensioner information by comparing it with data from other federal 
agencies, we found that a key procedure using Social Security 
Administration (SSA) and Internal Revenue Service (IRS) data is not 
conducted in a timely or efficient manner. Finally, despite millions of 
dollars in improper payments made each year, VA does not collect 
sufficient data on causes of improper payments that could be used to 
help it better manage the pension program. 

This report contains six recommendations to improve VA's management of 
the pension program. Specifically, VA should take steps to make more 
accurate and timely pension eligibility and payment decisions, improve 
its quality assurance of initial claims, incorporate feedback in its 
training, improve the effectiveness of pension eligibility review and 
verification processes, update its data match threshold, and evaluate 
the causes of improper pension payments. In its comments on a draft of 
this report, VA agreed in part or fully with our recommendations, 
though raised concern with some of the options we present to help 
implement several of the recommendations. 

Background: 

VA's pension program is means-tested and provides a minimum level of 
economic security for veterans with financial need. It is one of two 
cash benefits programs administered by VA. The other is disability 
compensation, which pays benefits to veterans who have disabilities 
related to their military service--often referred to as "service-
connected" disabilities.[Footnote 2] The pension program, on the other 
hand, which is the subject of this report, pays benefits to low-income 
veterans who either are elderly or have disabilities unrelated to their 
military service.[Footnote 3] Each program also provides benefits to 
eligible survivors. A veteran who applies for and meets requirements 
for both disability compensation and pension benefits will receive 
benefits through whichever program provides higher benefits. 

In 2006, VA paid over $34 billion in compensation and pension benefits 
to about 3.5 million veterans and survivors.[Footnote 4] Of this 
amount, $30.9 billion was paid in compensation benefits to 3,014,282 
veterans and their survivors. The remaining $3.5 billion was paid in 
means-tested pension benefits to 535,380 veterans and their survivors. 

The amount of financial assistance provided by the pension program is 
relatively modest and intended to raise pensioners' incomes to a level 
set out in statute. Pensioners are awarded an amount equal to the 
difference between their countable income, as determined by VA, and the 
maximum pension amounts as updated annually by statute.[Footnote 5] The 
maximum pension amount varies according to the pensioner's current 
income and number of dependents. In 2006, veterans with no income and 
no dependents could receive as much as $10,579 annually, while 
survivors with no income and no dependents could receive a maximum of 
$7,094 per year. Pensioners are required to report any changes in 
income, dependency, or other relevant circumstances to VA so that 
benefit levels can be adjusted accordingly.[Footnote 6] Generally, for 
each dollar of income received from other sources, the VA pension is 
reduced by the same amount. 

Eligibility Determinations for VA's Improved Pension Program: 

To determine a veteran's initial eligibility for the pension program, 
VA's regional office staff employ several criteria, including the 
veteran's military status, age or disability, and income. (App. II 
provides a summary of this process.) Eligibility for pension benefits 
is restricted to veterans who are at least 65 or have total and 
permanent disabilities unrelated to their military service. Also, VA 
considers the income of all family members, including spouses and 
children, but excludes the income of other individuals residing in the 
household. Various sources of income are considered when determining 
income eligibility, including employment, interest and dividends, 
retirement, annuities, workers' compensation, Social Security 
retirement and Disability Insurance benefits. Unreimbursed medical 
expenses that exceed 5 percent of the maximum pension amount may be 
deducted from income in determining eligibility. Eligibility for the 
surviving spouses and children of such veterans is based on similar 
factors. Once pensioners have been awarded benefits, VA makes ongoing 
eligibility determinations and adjusts benefit levels as needed. 
Pensioners are required to inform VA of any changes in their 
circumstances--such as hospitalization or incarceration, as well as 
changes in income and assets--that could affect their eligibility or 
benefit levels. To further assess ongoing eligibility and benefit 
levels, VA also requires pensioners who have any income other than 
Social Security to file an annual report with VA.[Footnote 7] Then, VA 
evaluates the information in this report to determine if pensioners 
continue to meet eligibility requirements. 

The VA Pension Population Is Characterized by Low Income, Older Ages, 
Multiple Impairments, and Decreasing Numbers: 

In 2006, most VA pensioners had nonpension incomes well below the 
federal poverty level, were beyond retirement age, and had multiple 
impairments, and the population had been decreasing in number. In 
addition to low incomes, the majority of VA pensioners had few assets 
and limited education. Since 1978, the total pension population has 
been decreasing, although there have been increases in numbers of 
pensioners from more recent service periods, including the Vietnam era 
and the Gulf War. 

VA Pensioners Generally Have Nonpension Incomes Well Below the Poverty 
Level, Few Assets, and Low Levels of Education: 

Most pensioners have very low annual incomes outside of their pension 
benefits. According to our analysis of VA data, veteran pensioners' 
average annual nonpension income was $4,573 in 2006.[Footnote 8] This 
was well below the 2006 federal poverty level of $9,800 for a single 
adult. Survivors had a lower average annual nonpension income of 
$3,046. Both veterans and survivors under 65 had lower average annual 
nonpension incomes than those 65 and older. Social Security benefits 
and non-Social Security retirement income accounted for much of the 
difference, as shown in table 1. When VA pension benefits are included, 
most veterans had annual incomes above the federal poverty level. This 
is not true of survivors, who receive smaller pension awards than 
veterans. Pension benefits in 2006 averaged $8,232 per year for 
veterans and $4,260 per year for survivors, for an average total 
income, respectively, of $12,805 and $7,306. 

Table 1: Estimated Amount of Annual Nonpension Income by Source and 
Age: 

Pensioner age: Under 65; 
Social Security benefits: Income: $2,432; 
Social Security benefits: Percentage of total income: 91%; 
Non-Social Security retirement income: Income: $146; 
Non-Social Security retirement income: Percentage of total income: 5%; 
Other income: Income: $107; 
Other income: Percentage of total income: 4%; 
Total income: $2,685. 

Pensioner age: 65 and over; 
Social Security benefits: Income: $8,194; 
Social Security benefits: Percentage of total income: 87%; 
Non-Social Security retirement income: Income: $1,152; 
Non-Social Security retirement income: Percentage of total income: 12%; 
Other income: Income: $88; 
Other income: Percentage of total income: 1%; 
Total income: $9,434. 

Source: GAO analysis of VA Benefits Delivery Network, October 2006, 5 
percent sample. 

Note: All numerical estimates have margins of error of plus or minus 
17.65 percent or less of the value of those numerical estimates. All 
percentage estimates from the sample file have margins of error of plus 
or minus 1 percentage point or less, unless otherwise noted. When GAO 
analyzed these income amounts, we only included the improved pension 
program pensioners and counted incomes for veterans and veteran spouses 
when the payee was the veteran or surviving spouse. We did not analyze 
children's income or cases for which children were the payee. 
Approximately 5 percent of cases had children or listed children as 
payee. 

[End of table] 

In a 2002 VA study, pensioners reported having few assets and low 
levels of education.[Footnote 9] Less than half of pensioners reported 
owning their own homes, with ownership rates for spouses being lower 
than those for veterans. Even fewer pensioners owned vehicles. About 
one-third of veterans reported owning a car, and only about one-fifth 
of spouses reported owning a car. Moreover, pensioners generally 
reported low levels of education, with those over age 65 reporting less 
education than those under age 65. More than half of veterans and 
spouses over age 65 reported not having a high school diploma, compared 
with 22 percent of veterans under age 65 and 44 percent of spouses 
under age 65. Over a third of pensioners under age 65 reported having a 
high school diploma, and less than 7 percent reported completion of a 
bachelor's or higher degree. 

The Average VA Pensioner Is 70 Years of Age and Has No Spouse or 
Dependents: 

The average age of VA pensioners is approximately 70, with 
approximately 60 percent over age 65 and less than 20 percent age 55 or 
younger. Significant numbers are over age 75, as shown in figure 1. The 
average age is highest for surviving spouses of deceased veterans, who 
constitute approximately one-third of all pensioners. In 2006, their 
average age was 72, while that of veterans was 69. About three-quarters 
of all surviving spouses are over age 65, compared to just over half of 
all veterans. Less than 2 percent of all pensioners are younger than 
age 45. 

Figure 1: Age Groupings of Veterans and Survivors in 2006: 

This figure is a combination bar graph showing age groupings of 
veterans and survivors in 2006. One bar represents veterans, and the 
other represents survivors. The X axis represents the age, and the Y 
axis represents pensioners. 

[See PDF for image] 

Source: Fiscal year 2006-2008 budget submissions. 

[End of figure] 

Most VA pensioners have no spouse or dependent children, according to 
information the pensioners provided to VA in 2006. As shown in figure 
2, about 82 percent of pensioners receive benefits for themselves 
alone, and most of the remaining 18 percent are veterans living with 
dependents. Proportionately few pensioners have dependent children 
eligible for pension benefits: About 22,000 of the half-million 
pensioners receive VA payments for support of their children, in most 
cases for one or two children. 

Figure 2: Household Composition of VA Pensioners in Fiscal Year 2006: 

This figure is a pie chart showing household composition of VA 
pensioners in fiscal year 2006. 

Veterans living alone: 49.1%; 
Spouse living alone: 31.9%; 
Veterans living with dependents: 17.5%; 
Adult children living alone: 1.0%; 
Spouse living with dependents: 0.5%. 

[See PDF for image] 

Source: VA fiscal year 2008 budget submission. 

[End of figure] 

The Majority of Veteran Pensioners Have Multiple Impairments, and About 
One-Third of All Pensioners Require Assistance to Perform Daily 
Activities: 

Most veteran pensioners have multiple disabling conditions, with 
approximately 95 percent reporting at least one impairment and nearly 
75 percent reporting two or more impairments. As shown in figure 3, 
excluding those impairments classified as "other," musculoskeletal 
conditions were the most common type of impairment among veteran 
pensioners. For veterans under age 65, the next most common type of 
impairment was mental, while for veterans age 65 and older it was 
cardiovascular. 

Figure 3: Types of Impairments among Veteran Pensioners by Age: 

This figure is a combination of two pie chart showing types of 
impairments among veteran pensioners by age. 

Types of impairments among veteran pensioners under age 65: 
Cardiovascular: 12%; 
Musculoskeletal: 29%; 
Mental: 15%; 
Neurological: 9%; 
Digestive: 8%; 
Other: 27%. 

Types of impairments among veteran pensioners age 65 and older; 
Cardiovascular: 16%; 
Musculoskeletal: 23%; 
Mental: 9%; 
Neurological: 7%; 
Digestive: 7%; 
Other: 38%. 

[See PDF for image] 

Source: GAO analysis of VA Benefits Delivery October 2006 Mini Master 
file. 

Note: VA's database lists multiple impairments for each beneficiary. 
This figure reflects all impairments listed in the VA database for 
pension beneficiaries. "Other" includes several categories of 
impairments that are individually associated with less than 7 percent 
of veteran pensioners. 

[End of figure] 

Many pensioners require aid and attendance in their activities of daily 
living, such as dressing and feeding themselves. According to VA, 
slightly less than one-third of pensioners are housebound or in need of 
aid and attendance, and most of these pensioners are age 65 and 
older.[Footnote 10] While more than 40 percent of pensioners age 65 and 
over require aid and attendance, less than 15 percent of younger 
pensioners do. Only a small number of pensioners reside in a nursing 
home--less than one-half of 1 percent of pensioners under age 65 and 3 
percent of pensioners age 65 and over. 

In a 2002 VA study, about 95 percent of veterans and 87 percent of 
surviving spouses reported having some form of health 
insurance.[Footnote 11] Medicare was a source of coverage for 50 
percent of veterans and 69 percent of surviving spouses. Over half of 
veterans also reported health care coverage through VA or military 
hospitals. Less than 15 percent of veterans relied on Medicaid for 
their health insurance and less than 15 percent of veterans had private 
health insurance. By contrast, 47 percent of surviving spouses relied 
on Medicaid, making it the second most common source of coverage after 
Medicare, while 18 percent reported having private insurance. 

The Size of the Pensioner Population Has Decreased: 

As shown in table 2, the total number of pensioners has been decreasing 
in recent years, although the number of pensioners from more recent 
service periods has been increasing. In 2006, the three pension 
programs served about 535,000 veterans and survivors,[Footnote 12] a 75 
percent decrease from 1978, when participation peaked at almost 2 
million. VA attributes the overall reduction in numbers largely to the 
death of World War II era pensioners and the greater availability of 
Social Security retirement benefits, which often raise veterans' 
incomes above the VA pension program's eligibility levels. 

Table 2: Decrease in Number of Veteran and Survivor Cases by Pension 
Program, Fiscal Years 2000-2006: 

Veterans; 
Program/year: Old Law; 
2000: 304; 
2001: 260; 
2002: 231; 
2003: 191; 
2004: 160; 
2005: 136; 
2006: 110; 
Percentage change 2000-2006: -64%. 

Veterans; 
Program/year: Prior Law; 
2000: 23,604; 
2001: 19,925; 
2002: 16,856; 
2003: 14,288; 
2004: 1,1976; 
2005: 10,100; 
2006: 8,601; 
Percentage change 2000-2006: -64%. 

Veterans; 
Program/year: Improved; 
2000: 340,312; 
2001: 327,867; 
2002: 329,492; 
2003: 332,077; 
2004: 330,767; 
2005: 325,551; 
2006: 321,145; 
Percentage change 2000-2006: -6%. 

Survivors; 
Program/year: Old Law; 
2000: 1,384; 
2001: 1,084; 
2002: 841; 
2003: 664; 
2004: 520; 
2005: 427; 
2006: 345; 
Percentage change 2000-2006: -75%. 

Survivors; 
Program/year: Prior Law; 
2000: 76,112; 
2001: 68,306; 
2002: 60,884; 
2003: 55,076; 
2004: 49,024; 
2005: 43,813; 
2006: 39,341; 
Percentage change 2000-2006: -48%. 

Survivors; 
Program/year: Improved; 
2000: 180,282; 
2001: 172,724; 
2002: 169,151; 
2003: 168,398; 
2004: 166,288; 
2005: 162,898; 
2006: 161,470; 
Percentage change 2000-2006: -10%. 

Source: GAO analysis of VA 2007 and 2008 budget submissions. 

[End of table] 

While total program enrollment has declined, the Vietnam era and Gulf 
War cohorts of pensioners have increased in number in recent years, as 
shown in table 3. VA expects the Vietnam era cohort to continue to 
increase in number as more Vietnam era veterans meet the 65-year age 
requirement for pension eligibility. An estimated 5.2 million of these 
veterans will be age 65 or older in the year 2015. However, VA has no 
estimate of how many also will have qualifying wartime service, meet 
income requirements, and submit pension claims. Further, the number of 
Gulf War veterans receiving pensions increased by nearly 300 percent 
between 2000 and 2006, although they are still less than 5,400 in 
number. Even with increases in these two cohorts, VA estimates that the 
total pensioner population will continue to decline in size through at 
least 2017. 

Table 3: Percentage Change in the Number of Pensioners by Service 
Period, Fiscal Years 2000-2006: 

[See PDF for image] 

Source: GAO analysis of VA fiscal years 2002-2008 budget submissions. 

[End of table] 

The main reason for caseload termination is death, followed by 
increased income, as shown in figure 4. In January 2007, about 70 
percent of veteran pension cases and 50 percent of surviving spouse 
cases were terminated as a result of the death of the pensioner. 
Increased income was the reason for termination in about a fifth of 
veteran cases and in about two-fifths of surviving spouse cases. For 
surviving children, however, about two-thirds of the cases were 
terminated due to the child's age, and one in five was because of a 
death. 

Figure 4: Reasons for Program Termination, January 2007: 

This figure is a bar graph showing reasons for program termination, 
January 2007. The bars represent other, age, income, and death. 

[See PDF for image] 

Source: VA data. 

[End of figure] 

For Initial Pension Claims, VA Policies and Procedures Do Not Always 
Ensure Well-Supported Decisions: 

VA policies and procedures are not sufficient to ensure sound decisions 
on new pension claims. Unlike other federal agencies with similar 
income-based programs, VA largely does not independently verify the 
accuracy of financial information provided by claimants to support 
initial pension program eligibility, a fact that makes the pension 
program vulnerable to improper payments. In addition, the guidance used 
by staff to make pension eligibility decisions, which is under revision 
and dispersed across several sources, is not always current or clear. 
Further, VA's quality assurance review process for initial claims does 
not select a sufficient number of pension cases to examine to ensure 
the accuracy of pension decisions. Finally, VA does not adequately 
evaluate pension training. For example, VA does not systematically 
collect feedback from participants at the end of a training course. 

VA Does Not Verify All Financial Information Provided by Applicants: 

VA does not require new pension applicants to submit documents that 
would support their declarations of income and assets. While the agency 
does corroborate their reported Social Security income with SSA 
records, it does not require claimants to submit evidence for other 
financial resources, such as by asking for copies of bank statements or 
tax returns.[Footnote 13] Furthermore, until recently, there was no 
legislative authority for VA to arrange for the cross-checking of 
claimants' statements of non-Social Security income against the 
Department of Health and Human Service's (HHS) National Directory of 
New Hires (NDNH).[Footnote 14] The NDNH includes quarterly wage data 
for up to eight quarters, which can be compiled into annual data for 
matching purposes. HHS conducts matches using the NDNH for other 
agencies--such as SSA, IRS, and the Department of Housing and Urban 
Development to assist them in improving their enforcement efforts. The 
Dr. James Allen Veteran Vision Equity Act of 2007, which became 
effective December 26, 2007, requires VA to provide applicant 
information to HHS and requires HHS to match it against the NDNH and 
disclose to VA information for verifying applicant employment and 
income.[Footnote 15] 

Furthermore, while VA allows income deductions for certain unreimbursed 
medical expenses, the agency does not always require documentation for 
these expenses. For example, VA requires documentation for the costs of 
nursing home care, but not for the cost of prescription drugs. In 
contrast, VA does require applicants to verify some nonfinancial 
information, such as by submitting an official notice of military 
separation or marriage and divorce records. 

Guidance Used by Staff Is Poorly Organized and Not Necessarily Current 
or Clear: 

We found that staff in the regional offices we visited used guidance 
that was not compiled in a single source; not always current; and, 
according to those we spoke with, unclear. This may be, in part, 
because VA has been in the process, since 2001, of revising various 
sections of its compensation and pension manual to help clarify complex 
regulations. However, staff told us that while the revisions are taking 
place, they must check a variety of sources for updates, including e-
mails and posted memos, to be certain they have the most current 
version for a specific procedure. Staff also said the piecemeal and 
dispersed nature of the guidance can lead to different interpretations 
for pension eligibility decisions. VA expects to have most of the 
revisions implemented by 2009. Finally, staff also expressed a concern 
over the clarity of some pension guidance, which they said both leaves 
too much room for interpretation and can result in inconsistent 
decisions on eligibility. For example, some said they must interpret 
ambiguous guidance when determining how to treat claimants in assisted 
living centers versus nursing homes, and that it is possible for staff 
to reach different conclusions about a claimant's eligibility or proper 
benefit. They told us that when faced with unclear guidance, staff are 
expected to use their own discretion in interpreting the guidance, 
along with the advice of supervisory staff, which they believe can 
vary. 

VA Has a Limited Approach to Evaluating the Quality of Initial Pension 
Decisions: 

The internal controls VA employs to evaluate the quality of initial 
pension decisions are insufficient because VA reviews only a very small 
random sample of initial claims that are selected from compensation and 
pension cases.[Footnote 16] Since pension claims constitute only about 
11 percent of the combined compensation and pension caseload, few are 
likely to be included in the quality assurance review sample. Although 
VA reported about a 12 percent error rate for compensation and pension 
claims combined, a recent VA Inspector General (IG) study found a 
higher incidence of errors in some cases that subsequently required 
pension payment adjustments. Specifically, the IG reported that VA 
procedures did not ensure that the benefits of veterans hospitalized 
for more than 90 days were appropriately adjusted.[Footnote 17] 

VA Does Not Adequately Evaluate the Training for Pension Staff: 

VA's training program does not include a comprehensive evaluation to 
ensure its effectiveness. Although we have found that evaluation is 
essential to performance,[Footnote 18] VA does not systematically 
collect feedback from participants at the end of a training course. For 
example, new staff receiving training at VA's training center in 
Baltimore are required to submit their evaluation of the training. By 
comparison, staff receiving training in the use of an electronic Web 
tool are not required to evaluate the training. 

When we discussed VA's limited evaluation of training with headquarters 
officials, they noted that the agency has made assessments of new 
training materials before they are put into place. However, VA has not 
consistently evaluated all training courses offered at the regional 
offices. Many staffers told us that some of their training is 
repetitive or does not include updates and revisions in procedures. 

For Ongoing Pension Cases, VA Procedures and Controls Do Not Adequately 
Ensure That Benefits Are Proper: 

VA procedures for assessing whether pensioners continue to receive the 
proper benefits have significant limitations because VA does not 
require pensioners to submit financial documentation, conducts untimely 
and inefficient verification of pensioners' incomes and assets, and 
lacks a system for identifying and reducing improper pension benefits. 
Although the agency requires all pensioners to submit documentation for 
nonfinancial changes, such as for marriages or deaths, it does not 
require documentation such as bank or asset statements when pensioners 
report financial changes. Also, while the agency does verify certain 
pensioner information by comparing it with data from other federal 
agencies, we found that a key procedure using SSA and IRS data is not 
conducted in a timely or efficient manner. Finally, despite millions of 
dollars in improper payments made each year, VA does not collect 
sufficient data on causes of improper payments that could be used to 
help it better manage the pension program. 

VA Has a Limited Approach to Verifying Financial Information Reported 
by Pensioners: 

Although VA requires all pensioners to report changes that might affect 
their payments or eligibility, it does not require them to submit 
documentation to corroborate changes in financial circumstances. 
Whereas VA does require pensioners to submit documents for such changes 
in circumstances as marriage or spousal deaths, it does not do so for 
reported changes in income or assets. This leaves the agency heavily 
dependent on the pensioner for self-reporting financial updates. The 
exception is Social Security income, which VA staff verify using a 
direct computer link to SSA benefit data. Pensioner reports of other 
financial changes, however, need not be accompanied by such documents 
as pay stubs, bank statements, or tax returns. This is in contrast to 
other income-based federal programs that typically ask for verification 
of key financial information. For example, SSA requires Supplemental 
Security Insurance (SSI) recipients to document their earnings on a 
regular basis.[Footnote 19] 

Similarly, VA does not ask pensioners for financial documentation when 
completing the annual Eligibility Verification Report (EVR), the 
requested update that the agency sends out annually to those who have 
previously reported having income and assets other than their pensions 
and Social Security benefits. VA uses information collected from the 
EVR responses to adjust or, if necessary, terminate VA pension 
benefits. (See table 4 for an overview of the EVR steps.) VA has 
indicated that the pension program is prone to overpayments caused by 
pensioners failing to report income changes as they occur. One agency 
official responsible for the management of the EVR process told us that 
the EVR process provides the agency an opportunity to adjust the 
pension benefits as pensioner status changes, thus preventing higher 
overpayments. 

Table 4: VA Steps to Process Eligibility Verification Reports: 

Step: 1; 
Description: The Hines Information Technology Center prints 
EVR forms and letters to beneficiaries in December; 
Approximately 70,000 EVRs are sent to pensioners annually; 
These EVR forms contain preprinted information regarding Social 
Security and VA benefits based on current VA data. 

Step: 2; 
Description: Pensioners are instructed to verify or correct 
the preprinted information and return the report to VA within 60 days. 

Step: 3; 
Description: The PMC staff conduct an initial review of EVRs 
to; 
* identify incomplete EVRs and return to pensioners for additional 
information (which must be returned within 60 days),; 
* approve and process completed EVRs, and; 
* refer EVRs needing award adjustment for further action. 

Step: 4; 
Description: PMC staff process all completed EVRs. 

Step: 5; 
Description: PMC staff adjust pension payment, create a retroactive 
payment, or terminate from the program, as necessary. 

Source: GAO analysis of VA documents. 

[End of table] 

In addition to VA not asking respondents to submit financial 
documentation, we found a number of other deficiencies in the EVR 
process that may likely limit VA's ability to make timely and accurate 
adjustments to benefits. First, the EVR process only queries pensioners 
who have previously reported income to VA, so the agency fails to reach 
pensioners who may acquire new sources of income, such as earnings from 
new employment.[Footnote 20] Second, during the annual review of EVRs, 
VA does not attempt to corroborate via independent third-party sources 
any information that pensioners report on the EVR beyond SSA benefits. 
Third, the EVR asks pensioners to estimate their income for the year, 
and the agency adjusts their pensions based on these estimates. Yet 
pensioner estimates could prove to be incorrect. For example, income 
estimates for the coming year could change due to the death of a 
spouse. Fourth, the narrow seasonal window within which pension 
maintenance center (PMC) staff attempt to review and process the EVRs-
-the first 3 months of the year--postpones other pension-related 
activities, including data comparisons with other federal agencies that 
could provide third-party verification. This limited approach to 
verification of pensioner-provided updates puts the program at risk of 
issuing improper payments based on pensioners' reports. 

VA's Key Data Match Procedure Is Not Timely or Efficient in Detecting 
Improper Payments: 

In addition to deficiencies in the EVR process, VA's ability to detect 
improper payments is hindered by untimely processing of a key data 
match. Through formal agreements, VA compares its electronic data on 
pensioners' reported income, and other eligibility information, to 
similar information from a number of federal agencies, as shown in 
table 5. VA identifies discrepancies and, after due process, uses this 
information to adjust or terminate pension benefits. VA projects that 
its data-matching activities save the pension program millions of 
dollars annually. 

Table 5: VA Computer Matching Procedures: 

Match: Income Verification Match; 
Frequency: Annual; 
Approximate volume: 36,000-46,000; 
Agency: SSA and IRS; 
Procedure: Compensation and pension records are matched with the SSA 
database for earned income. Then, SSA transmits the file to IRS to 
obtain data on unearned income for the specific year. SSA matches 
against IRS data and sends the combined results back to VA. When VA 
identifies unreported income, it sends letters to employers and 
pensioners to verify the income. No further action is taken until the 
recipients respond, or until 60 days, whichever comes first. After that 
time, VA adjusts the pension based on corrected income data. 

Match: Social Security Verification Match; 
Frequency: Annual; 
Approximate volume: 23,000-28,000; 
Agency: SSA; 
Procedure: VA sends an extract of its pension files to SSA, which then 
runs this extract against the master beneficiary record and inserts SSA 
payment information. The extract tape is then returned and VA master 
records are updated. VA then processes the cases, calculating the 
revised income for VA's purposes, taking into account the SSA cost-of-
living adjustment and Medicare deduction, and adjusts the pension 
accordingly. 

Match: Social Security Death Match; 
Frequency: Monthly; 
Approximate volume: 4,000; 
Agency: SSA; 
Procedure: Compensation and pension records are matched with the SSA 
death records, which include only those deaths reported to SSA since 
the last SSA/VA match. When a match is made between the SSA death 
records and VA records, the individual (payee or spouse) is identified 
as possibly deceased. After 60 days, VA may terminate the award if the 
pensioner is deceased. 

Match: Social Security Prison Match; 
Frequency: Periodically; 
Approximate volume: 900; 
Agency: SSA; 
Procedure: Compensation 
and pension records are matched with the SSA inmate database. VA 
generates a worksheet containing information on the matches. VA staff 
verify the status for each case and terminate pensions as appropriate. 

Match: Bureau of Prisons Match; 
Frequency: Monthly; 
Approximate volume: 30; 
Agency: Department. of Justice, Bureau of Prisons (BOP); 
Procedure: VA master records are matched with the BOP database, 
containing data on incarcerated individuals or fugitive felons. The 
match generates a worksheet if the Social Security number (SSN) in the 
BOP records matches one in the compensation and pension master record. 
VA staff must work each case to suspend or reduce the pension. 

Match: Civil Service Verification Match; 
Frequency: Annual; 
Approximate volume: 5,500; 
Agency: Office of Personnel Management (OPM); 
Procedure: VA master records are matched with OPM records of persons 
receiving Civil Service Annuity (CSA) and Federal Employees Retirement 
System benefits. When the SSN and surname are the same in both records, 
the VA beneficiary is identified as a CSA recipient. 

Match: Railroad Retirement Verification Match; 
Frequency: Annual; 
Approximate volume: 1,200; 
Agency: Railroad Retirement Board (RRB); 
Procedure: VA's master records are matched with RRB records of 
people receiving Railroad Retirement (RR) benefits. If the SSN is the 
same in both records, the VA beneficiary is identified as a recipient 
of RR. 

Source: GAO analysis of VA documents. 

[End of table] 

Nevertheless, VA's key data match, the Income Verification Match 
program (IVM), which uses SSA and IRS data to detect pensioners who 
have failed to report earned or unearned income, is running far behind 
schedule. This match is scheduled to occur on an annual basis. However, 
since 2006, VA has attempted to become current by conducting 2 years' 
worth of IVM data matches simultaneously. Yet combining data match 
records for multiple years has added to the complexity and the length 
of case evaluations, according to officials we interviewed. 

VA estimates that the IVM data match has the potential of saving the 
program over $10 million each year, which suggests that a delay of 2 
years could delay saving as much as $25 million dollars in payment 
errors.[Footnote 21] Such errors include both underpayments, which 
require VA staff to make retroactive payments or benefit adjustments, 
and overpayments, which can be burdensome for VA to recover from such a 
low-income population of pensioners and which threaten the financial 
stability of pensioners. While VA officials told us they plan to have 
all IVM data match cases entered into their system by the end of 
2007,[Footnote 22] they could not provide assurance that these cases 
will be completed in the same year. 

The effectiveness of the IVM is also undermined by the fact that the 
data used are not current. VA uses income data that are about 2 years 
old, despite the fact that the SSA data are available earlier and that 
more recent earned income data are available from another federal 
database, the NDNH.[Footnote 23] Moreover, while VA could make use of 
SSA earned income data as early as September following the end of the 
tax year, it postpones the match of pension benefits and waits for 
unearned income from IRS so that it can simultaneously evaluate 
eligibility for pension and Individual Unemployability (IU) 
benefits.[Footnote 24] We determined that the combination of old earned 
income data, along with the delays noted above, means that individuals 
with unreported earned income could continue to receive benefits for at 
least 2 years before VA can determine that their pension benefits 
should be adjusted or terminated. 

We also found that VA's handling of the IVM data match results is 
inefficient because the dollar threshold used to select pensioners for 
review of ongoing pension eligibility has not kept pace over nearly two 
decades with wages necessary for a person to sustain a living. 
Specifically, the IVM data match threshold in use until December 2006-
-which VA told us was meant to represent wages for marginal employment-
-was set in 1988 (when VA's maximum annual pension rate for a veteran 
with no dependents was $6,463). Yet between 1988 and 2006, the U.S. 
Census Bureau's poverty threshold had increased over 70 
percent.[Footnote 25] As a result of VA's continued reliance upon the 
1988 threshold, PMC staff told us that they manually reviewed many more 
cases than necessary in order to find and delete those cases that did 
not warrant a review. In fact, staff typically eliminated about one-
third of the initial matches because their combined countable income 
and pension did not exceed their maximum allowable pension benefit. VA 
increased the $6,000 threshold to $9,383 in December 2006, but has not 
decided whether it would make such changes to update the threshold on a 
regular basis in the future. 

The timeliness of the IVM process is also hampered by the fact that 
VA's match records are paper based and lack the organizing and 
transmission efficiencies of an electronic database. Therefore, VA's 
follow-up work is slowed by the need to ship boxes of uncollated paper 
records--worksheets, correspondence with pensioners and employers, and 
related documents--from the Hines center to the PMCs for manual 
assembly and analysis. Although software exists to transfer the 
confidential information electronically, VA officials said the need to 
comply with IRS's specific security measures for income-related data 
has precluded VA's use of it. However, we have noted in prior work that 
electronic software exists to transfer confidential information 
securely.[Footnote 26] 

VA Conducts No Analysis of the Causes of Improper Pension Payments: 

Many millions of dollars in improper payments accrue each year before 
they are discovered and corrected, but VA has not taken steps to 
identify the causes of these improper payments. According to VA 
estimates, about 8.4 to 11 percent of pension program payments were 
made in error each year between 2003 and 2006, with most of these being 
overpayments, as shown in table 6. Specifically, VA estimated a total 
of almost $1.2 billion in overpayments and about $44 million in 
underpayments during this period. Such payment errors, particularly 
underpayments, can have a negative impact on pensioners living on very 
low incomes. By contrast, the agency has estimated that only about 1 
percent of disability compensation program payments were made in error, 
with about $730 million in overpayments and $375 million in 
underpayments for the same period. 

Table 6: VA Estimates of Pension Improper Payments, Fiscal Years 2003-
2006: 

Fiscal years: 2003; 
Total outlays: 3,219; 
Overpayments: 269; 
Underpayments: 15; 
Improper payments (percentage of total pension outlays): 8.8%. 

Fiscal years: 2004; 
Total outlays: 3,326; 
Overpayments: 265; 
Underpayments: 15; 
Improper payments (percentage of total pension outlays): 8.4%. 

Fiscal years: 2005; 
Total outlays: 3,383; 
Overpayments: 361; 
Underpayments: 10; 
Improper payments (percentage of total pension outlays): 11%. 

Fiscal years: 2006; 
Total outlays: 3,525; 
Overpayments: 300; 
Underpayments: 4; 
Improper payments (percentage of total pension outlays): 8.6%. 

Source: GAO analysis of VA documents and VA Performance and 
Accountability Report, 2006. 

Note: Overpayment and underpayments rounded to the nearest million. 

[End of table] 

Furthermore, according to VA officials, less than 30 percent of the 
overpayments for compensation and pension are recovered. We found that 
VA has not been successful in collecting overpayments. In fact, a 
significant portion of overpayments are written off because of 
pensioners' death or bankruptcy, or because the overpayment is 
considered uncollectible. For example, in cases where VA's Committee on 
Waivers concludes that repayment of the overpayment debt would cause 
undue financial hardship for the pensioner, the debt may be 
waived.[Footnote 27] Most of the remaining overpayments remain on VA's 
financial account for up to 10 years, at which time VA discontinues its 
collection efforts. 

Despite VA's estimates of relatively high amounts in improper payments 
for the pension program, the agency lacks a process for determining the 
nature and actual extent of payment errors. For example, VA currently 
does not have the ability to identify the dollar amount of overpayments 
generated because of a failure to report income, gambling winnings, or 
death of a spouse. Likewise, VA cannot identify the amount of improper 
payments caused by administrative factors, such as those associated 
with VA delays in conducting the IVM data matches. Although the agency 
has identified several causes for improper payments, such as inaccurate 
reporting of Social Security benefits, remarriage of surviving spouse, 
or change in dependents, it has not developed a system to assess the 
degree to which they occur, and therefore cannot develop strategies to 
target problems and take steps to correct them. 

Conclusions: 

VA's pension eligibility operations currently provide a limited return 
on investment when it comes to making appropriate payments. The 
agency's financial verification processes are heavily weighted toward 
detection of payment errors late in the process rather than up-front 
prevention of errors. By not requiring supporting financial 
documentation from individuals at the time of application, and relying 
much later in the process on untimely data matching for this 
information, VA increases the likelihood of making improper payments. 
Additionally, VA concentrates staff resources in the first 3 months of 
each year on a process for updating eligibility that is time-consuming 
and still lacks for financial documentation or independent 
verifications. Meanwhile, the data-matching operations that do offer 
third-party verifications on financial status are delayed until after 
the EVR process is completed. A key data match procedure, the IVM, has 
been delayed in the past, and the agency continues to be at risk of 
making improper payments as it tries to rectify the situation. Even 
when the IVM data match becomes current, VA's ability to use it in a 
timely way will be limited if processing of the information remains 
grounded in a manual handling of paper documents. Moreover, if VA does 
not regularly update its IVM data match threshold in its incoming 
financial information system, the agency will continue to have an 
inefficient procedure for selecting cases for IVM review. 

In view of the fact that the program incurs a proportionately high 
level of improper payments--about 8.4 to 11 percent of annual program 
benefits--VA's investment in the prevention of errors seems too modest. 
VA could readily justify increasing its investment in the error 
prevention process. The agency does not have a quality assurance 
process robust enough to ensure with a high degree of confidence the 
accuracy of decisions made in initial pension cases, it has not 
developed a consistently rigorous approach to the evaluation of 
training, and it does little to identify and analyze the causes of 
improper payments. 

Taken together, VA forfeits the opportunity to address administrative 
weaknesses and prevent payment errors. Unless these weaknesses are 
addressed, VA pension benefit payments will remain vulnerable to a 
relatively high rate of errors. Reconciling improper payments draws 
resources from the agency that could be better utilized elsewhere, and 
to the extent overpayments are never collected, they undermine the 
agency's financial stewardship over public funds. Moreover, 
underpayments can lead to hardships among this low-income segment of 
the veteran population. Certainly, the program is not too small to 
ensure that appropriate levels of assistance are provided to a 
vulnerable and deserving population. This can and should be done with a 
level of efficiency that minimally burdens all concerned. 

Recommendations for Executive Action: 

In order to enhance VA's management of the pension eligibility process, 
we recommend that the Secretary of Veterans Affairs direct the Under 
Secretary for Benefits take the following actions. 

1. Take steps to make more accurate and timely decisions about initial 
and ongoing pension eligibility and payment levels. Such actions might 
include requiring pensioners to submit supporting documentation of 
their income and assets, conducting a more robust independent 
verification with third parties, or maximizing process automation. 

2. Take steps to improve its quality assurance review of initial 
claims, which could include reviewing a larger sample of pension 
claims. 

3. Incorporate evaluative feedback more consistently into the planning, 
design, and implementation of its training efforts. 

4. Evaluate options for improving the effectiveness of its annual 
eligibility verification review. This effort could involve 
reformulating the EVR process by (a) surveying every pensioner rather 
than a selected subset; (b) performing reviews on a rolling basis, such 
as on an individual's anniversary date, rather than diverting staff for 
this activity for a 3-month period; (c) reviewing pensioners' 
eligibility once every few years rather than every year; or (d) 
focusing on verifying pensioners' income and assets. 

5. Update the IVM data match threshold level to be in line with the 
U.S. Census Bureau poverty threshold or a comparable measure. 

6. Evaluate the causes of improper payments and use the results to 
develop and implement an action plan to prevent them from occurring. 

Agency Comments and GAO Response: 

We provided a draft of this report to the Secretary of Veterans Affairs 
for review and comment. In its written comments on a draft of this 
report (see app. III), VA agreed in part or fully with our 
recommendations. Although VA raised concern with some of the options we 
present to help implement several of the recommendations, the agency's 
comments indicate it will take steps to implement these 
recommendations. 

Regarding our recommendation that VA should take steps to make more 
accurate and timely pension eligibility and payment decisions, the 
agency agreed in part but took exception with several of the actions we 
suggested. VA noted that requiring pensioners to provide documentation 
of income and net worth could be onerous to individuals and possibly 
diminish the timeliness of initial pension eligibility decisions. 
However, VA also stated elsewhere in its comments that pensioners use 
the end-of-year tax statements (Form 1099) to accurately report income 
from all sources. This indicates that it may not be an added burden to 
pensioners to include a Form 1099 in their initial claim. But to the 
extent that burden is added to preparing the initial application, the 
possible gains in accuracy and avoidance of corrective actions later 
taken to address improper payments need to be considered. Additionally, 
VA cited its data-matching process with SSA as an existing mechanism to 
independently verify income. However, as we point out in this report, 
other federal agencies with similar income-based programs independently 
verify self-reported information to support initial eligibility 
decisions. By verifying claims up front, rather than years after the 
claim is established, the agency can effectively save the program 
millions of dollars that it might never recover otherwise. Moreover, VA 
does not always require documentation of unreimbursed medical expenses, 
which can be deducted from income for pension eligibility. It is 
important for VA to gather supporting documentation in this area as 
well. The agency may consider piloting the feasibility of requiring 
additional documentation of key financial information. 

Regarding our recommendation that VA take steps to improve its quality 
assurance review of initial claims, which could include reviewing a 
larger sample of pension claims, the agency agreed with this 
recommendation and indicated it will double the entire rating sample 
size in 2008. While we commend VA's effort to increase the overall 
number of claims reviewed in its quality assurance review of rating-
related decisions, we remain concerned that this approach will not 
ensure that enough initial pension claims are reviewed for quality 
assurance. As we point out in this report, VA samples 10 claims from 
most regional offices' caseload of compensation and initial pension 
claims. Given that initial pension claims constitute about 11 percent 
of the caseload, a regional office, on average, would likely have only 
1 pension claim selected for review. Doubling the sample size would 
increase the expected number of claims to 2, which we believe is too 
few. Alternatively, VA might consider increasing the number of pension 
claims in the overall sample, such as weighing the sample size to 
include more pension claims or conducting stand-alone reviews of 
pension claims. 

The agency agreed with our recommendation to incorporate evaluative 
feedback more consistently into the planning, design, and 
implementation of its training efforts. 

The agency agreed with our recommendation that VA evaluate options for 
improving the effectiveness of its annual eligibility verification 
review. The agency stated there are inconsistencies between the 
possibilities we present, though it did not elaborate. Our intent in 
presenting these options is to stimulate actions that VA may take to 
improve the effectiveness of the EVR process. The agency discussed 
issues--such as veterans' ease in providing corroborating information-
-that need to be considered as it moves forward. 

The agency agreed with our recommendation that it update the IVM data 
match threshold level to be in line with the U. S. Census Bureau 
poverty threshold or a comparable measure. 

The agency agreed with our recommendation that it evaluate the causes 
of improper payments and use the results to develop and implement an 
action plan to prevent them from occurring. 

In the version of the draft report sent to VA for review, we 
recommended that VA seek legislative authority to use the NDNH in its 
enforcement efforts. However, we have withdrawn this draft 
recommendation because VA has been mandated to use this database 
effective December 2007. In its comments on this draft recommendation, 
VA indicated that it has initiated the process necessary to gain access 
to the NDNH earnings database. Given this development, we encourage VA 
to act swiftly to position itself to fully utilize the NDNH in a timely 
manner. 

We will send copies of this report to the Secretary of Veterans 
Affairs, appropriate congressional committees, and other interested 
parties. The report will also be available at no charge on GAO's Web 
site at [hyperlink, http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-7215 or bertonid@gao.gov. Contact points for 
our Offices of Congressional Relations and Public Affairs may be found 
on the last page of this report. Other contacts and staff 
acknowledgments are listed in appendix IV. 

Signed by: 

Daniel Bertoni: 

Director, Education, Workforce, and Income Security Issues: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The Senate Veterans' Affairs Committee asked GAO to (1) determine the 
characteristics and trends in size of the pensioner population, (2) 
assess the policies and procedures the Department of Veterans Affairs 
(VA) has in place to ensure that initial pension eligibility decisions 
are well managed, and (3) assess the procedures VA has in place to 
ensure that pensioners continue to receive the proper benefit payments. 

To determine the characteristics and size of the pensioner population, 
we analyzed data from VA's Benefit Delivery Network (BDN), VA budget 
reports, and other reports through fiscal year 2006. To address our 
remaining objectives, we reviewed relevant laws, guidance, procedures, 
and internal controls that VA uses to ensure the soundness of pension 
benefit decisions. We also analyzed VA's internal control policies and 
performance reports. We visited 4 of VA's 57 regional offices, located 
in Boston, Milwaukee, Providence, and St. Paul. We selected these sites 
based on variations in size and geographic locations. We also visited 
VA's three pension maintenance centers (PMC), located in Philadelphia, 
Milwaukee, and St. Paul, and the Debt Management Center in St. Paul. We 
interviewed VA officials and staff at these sites as well as officials 
at VA Central Office in Washington, D.C. We also conducted case file 
reviews in three locations--Milwaukee, Providence, and St. Paul--to 
verify the adequacy of documentation in support of initial pension 
decisions. We conducted our review from November 2006 to February 2008, 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe that 
the evidence obtained provides a reasonable basis for our findings and 
conclusions based on our audit objectives. 

Analysis of Pensioners' Characteristics: 

To analyze the characteristics of pensioners, we extracted data on 
basic characteristics and enrollment trends from VA's Benefit Delivery 
Network and routine annual VA budget submissions. We supplemented our 
analysis with additional data from a study contracted by VA and 
conducted by ORC Macro. 

BDN: 

VA maintains data on pensioners in its Benefits Delivery Network at the 
Hines Information Technology Center in Illinois. Hines issues routine 
reports to various units in VA, including budget and policy, for use in 
program management. Most of these reports provide characteristics for 
two or three factors, for example, for veterans by age grouping and by 
state of residence. We based our analysis on year-end reports through 
fiscal year 2006. 

To obtain data not available from published VA reports, we analyzed 
additional BDN data as of September 2006. VA regularly provides GAO 
with two sources of BDN data. The first provides data on a few 
characteristics of the entire universe of pensioners. The other 
provides data on a broader range of characteristics for a random sample 
of 5 percent of all pensioners. We used data from the entire universe 
of cases when they were available and supplemented our analysis with 
additional data from the random sample. When reporting data from the 5 
percent sample, we report confidence intervals around point estimates. 

Budget Data: 

We also analyzed data extracted from tables in the fiscal year 2002 
through 2008 budget submissions. 

ORC Macro Study Data: 

In 2002, VA contracted with ORC Macro to survey a sample of pensioners 
that included veterans and surviving spouses. For more detailed data 
not available from VA's BDN, we used data reported in the ORC Macro 
report. However, data are not generalizable to current pensioners 
because of the time frame when the survey was conducted. The ORC Macro 
study was a telephone-administered survey. The survey sample generally 
mirrored the pensioner population at the time. However, for the sample 
of veterans, nursing home residents were underrepresented. In addition, 
the spouse sample and spouse recently enrolling sample underrepresented 
spouses receiving aid or attendance services. In all three cases, ORC 
Macro conducted a sensitivity analysis to find out whether these 
underrepresented groups would have changed key question response 
frequencies if they were correctly represented. In each of the three 
cases, ORC found they would not significantly change response 
frequencies. 

GAO's applied research and methods group evaluated the BDN and ORC 
Macro survey data and found the information to be sufficiently reliable 
for our purposes. 

Case File Review: 

To verify the adequacy of documentation in support of initial pension 
decisions, we reviewed case files on a random sample of Improved 
Pension claims and Death Pension claims that had been completed by 
three VA regional offices, in Milwaukee, St. Paul, and Providence. We 
chose Milwaukee and St. Paul because they were associated with a PMC. 
Because the Milwaukee site processes new claims in both the regional 
office and the PMC, we included files that originated at both locations 
in our review. Additionally, because the St. Paul office completed 
pension cases that had been transferred from another regional office, 
we included those cases in our review. We chose the Providence regional 
office because it was not associated with a PMC and because of its 
geographic location. 

At the three sites we reviewed a total of 72 case files for completed 
initial pension claims. In order to select case files to review, VA 
provided a list of all initial (or original) claims for Improved 
Pension (for veterans) and Death Pension (for survivors) considered and 
determined between February 1, 2007, and February 28, 2007.[Footnote 
28] To select from this list, which included both approved and denied 
cases, we used random number assignment procedures to help ensure a 
broad range of these types of cases were included in our sample. 
Specifically, at St. Paul, we used a randomized ordering process to 
select and review 32 case files. On subsequent site visits, we 
randomized all EP 180 and EP 190 claims and reviewed the first 
available 14 EP 180 and 6 EP 190 case files. In some instances, case 
files we selected were unavailable because they were being reviewed at 
the regional office or were categorized as pension claims but over the 
course of VA developing the claim no longer met our selection criteria 
of initial pension claims. We attempted to review a 2-to-1 ratio of EP 
180 to EP 190 cases, which is approximately the ratio that exists in 
the pension claims population. The number of case files we reviewed at 
each site is given in table 7. 

Table 7: VA Regional Offices Visited and Disposition of Pension Case 
Files Reviewed: 

Regional office: Milwaukee; 
Pension awards: 14; 
Pension denials: 6; 
EP 180 (Veterans' claims): 14; 
EP 190 (Survivors' claims): 6. 

Regional office: St. Paul; 
Pension awards: 24; 
Pension denials: 8; 
EP 180 (Veterans' claims): 23; 
EP 190 (Survivors' claims): 9. 

Regional office: Providence; 
Pension awards: 15; 
Pension denials: 5; 
EP 180 (Veterans' claims): 14; 
EP 190 (Survivors' claims): 6. 

Regional office: Total; 
Pension awards: 53; 
Pension denials: 19; 
EP 180 (Veterans' claims): 51; 
EP 190 (Survivors' claims): 21. 

Source: GAO. 

[End of table] 

To ensure consistency, we used a standardized checklist to examine case 
files at each location, and the same individuals conducted the reviews 
at all three sites. The checklist, which was developed by examining the 
procedural guidance and case files at a regional office, included 
information about each of the major eligibility criteria: military 
service, disability, net worth, and income. In addition, we reviewed 
demographic information about the claimants. We also looked for 
evidence that VA had followed guidance on procedures, for example, 
whether letters were sent to the claimant appropriately. At each of the 
sites we communicated with VA office management when questions about 
the files we were reviewing arose. In some cases these conversations 
helped us better understand why a case had been handled in a particular 
way, while in other cases management acknowledged that errors had been 
made. 

The results of visits to the sample of regional offices are not 
generalizable to all 57 regional offices; similarly, results of case 
file reviews are not generalizable to all pension files. The 
testimonial evidence, such as that gathered during interviews with 
staff, also is not generalizable. 

[End of section] 

Appendix II: Pension Eligibility Determination: 

Administrative Structure: 

VA's Compensation and Pensions (C&P) Service administers the pension 
program and oversees the operation of 57 regional offices and three 
pension maintenance centers (PMC). The C&P central office has 
responsibility for managing policy and procedures, guidance, quality 
assurance, and general operations. The regional offices process initial 
pension applications and determine eligibility. The offices have 
jurisdiction in the geographic area where a veteran or survivor lives. 
At least one regional office is located in every state except Wyoming, 
as well as the District of Columbia, Puerto Rico, and the Philippines. 

The PMCs are responsible for conducting annual reviews of eligibility 
and adjusting benefit levels when a pensioner's circumstances change. 
As part of the annual review, staff at the PMCs examine annual reports 
from pensioners, which contain information on, for example, income and 
medical expenses. Staff also examine the results of a variety of 
computerized data matches with other government agencies to determine 
whether adjustments are warranted. Additionally, VA's Hines Information 
Technology Center handles all computer transactions that affect benefit 
payment levels, and VA's Debt Management Center is responsible for 
collecting overpayments that occur. 

Initial Eligibility Determinations for VA's Improved Pension Program: 

To determine a veteran's initial eligibility for the pension program, 
VA's regional office staff review several eligibility criteria, 
including the veteran's military status, age or disability, and income. 
For the surviving spouses and children of such veterans, VA uses 
similar factors to determine eligibility, as shown in figure 5. All 
survivors are subject to income and asset limits, but only surviving 
children must meet disability or age requirements. 

Figure 5: Pension Eligibility Criteria for Veterans and Survivors: 

This figure is a flowchart showing pension eligibility criteria for 
veterans survivors. 

[See PDF for image] 

Source: Title 38, U.S. Code. 

[End of figure] 

Military Service Requirement: 

To qualify for the pension program, veterans must meet one of four 
military service criteria set out in statute, but in practice they have 
been collapsed into a single requirement.[Footnote 29] The VA 
characterizes these criteria as generally requiring that veterans have 
served on active duty in the military, naval, or air forces for 90 days 
or more, with at least one of those days being during a period of 
war,[Footnote 30] and been discharged under conditions other than 
dishonorable.[Footnote 31] In addition, veterans who enlisted after 
September 7, 1980, generally must have served at least 24 months or the 
full period for which called or ordered to active duty to 
qualify.[Footnote 32] 

Disability or Age Requirement: 

Eligibility for pension benefits is restricted to veterans who are 
either totally and permanently disabled due to circumstances unrelated 
to their military service or to willful misconduct, or who are at least 
age 65. Veterans under age 65 must be considered totally and 
permanently disabled, which means that the veteran is unable to pursue 
substantially gainful employment due to a disabling condition and that 
the condition is reasonably certain to continue throughout the 
veteran's life.[Footnote 33] Veterans who receive Social Security 
disability benefits or long-term care in nursing homes are presumed to 
be totally and permanently disabled. 

Income and Asset Limits: 

In determining eligibility, VA considers the income of all family 
members, including spouses and children, but excludes other individuals 
residing in the household. Pensioners meet income eligibility 
requirements if their family incomes are less than the maximum annual 
pension amounts established annually by Congress. Various sources of 
income are considered when determining eligibility, including 
employment, interest and dividends, retirement, annuities, workers' 
compensation, and Social Security Disability Insurance benefits. Income 
from other federal or state means-tested programs--such as Temporary 
Assistance to Needy Families, food stamps, housing assistance, or 
Supplemental Security Income--is not counted toward family income. 
Pensioners may deduct unreimbursed medical expenses that exceed 5 
percent of the maximum annual pension amount from their income for the 
purposes of determining eligibility. 

VA does not apply specific limits on the net worth of assets when 
determining eligibility for pensions. However, veterans will not be 
awarded a pension if VA determines that they have sufficient assets to 
live on for a reasonable period of time. To make this determination, VA 
guidance calls for net worth in excess of $80,000 to be reviewed during 
the initial eligibility determination process, taking into 
consideration such factors as the veteran's life expectancy and the 
convertibility of the assets into cash. 

Net worth includes all personal property and real estate owned by 
veterans and their families, excluding the value of their home and the 
land on which their homes are located. 

Survivor Eligibility: 

When a veteran with qualifying wartime service dies, his survivors may 
be entitled to a pension. The veteran does not have to be receiving a 
pension at the time of his death in order for his survivors to be 
eligible for benefits. Survivors are not required to have a disability 
in order to qualify, but they must meet income and asset requirements. 
For spouses, there is no age requirement, but they lose eligibility if 
they remarry. Generally children under age 18 are eligible, as are 
those under 23 who are in school. Older unmarried children may also be 
eligible, but only if they are incapable of self-support and if this 
incapacity occurred prior to their reaching age 18. 

Ongoing Eligibility Determinations for VA's Improved Pension Program: 

Once pensioners have been awarded benefits, VA makes ongoing 
eligibility determinations and adjusts benefit levels as needed. 
Pensioners are required to inform VA of any changes in entitlement 
factors that could affect their eligibility or benefit levels as soon 
as they occur. Hospitalization or incarceration, as well as changes in 
income, assets, or marital status, can affect the continued eligibility 
of pensioners or result in adjustments to the amount of payments that 
they receive. Any changes in circumstances must be reported to VA in 
writing as soon as they occur, and VA is required to make any necessary 
adjustments. To further assess ongoing eligibility and benefit levels, 
VA also requires pensioners who have any income other than Social 
Security to file an annual report with VA.[Footnote 34] Then, VA 
evaluates the information in this report to determine if pensioners 
continue to meet income eligibility requirements. 

[End of section] 

Appendix III: Comments from the Department of Veterans Affairs: 

The Secretary Of Veterans Affairs: 
Washington: 

January 31, 2008: 

Mr. Daniel Bertoni: 
Director, Education, Workforce, and Income Security Issues: 

U. S. Government Accountability Office: 
441 G Street, NW:
Washington, DC 20548: 

Dear Mr. Bertoni: 

The Department of Veterans Affairs (VA) has reviewed the Government 
Accountability Office's (GAO) draft report, Veterans Benefits: Improved 
Management Would Enhance VA's Pension Program (GAO-08-112) and 
partially agrees with GAO's conclusions and concurs with GAO's 
recommendations. 

The Department of Veterans Affairs (VA) concurs in part with GAO's 
recommendation that VA should take steps to make more accurate and 
timely pension eligibility and payment decisions. VA also concurs with 
recommendations to improve its quality assurance of initial claims; 
incorporate feedback in its training; improve the effectiveness of 
pension eligibility review and verification process; and evaluate the 
causes of improper pension payments. 

The enclosure addresses each of GAO's recommendation and provides 
additional discussion and comments to the draft report. VA appreciates 
the opportunity to comment on your draft report.

Sincerely, 

Signed by: 

James B. Peake, M.D.: 

Enclosure: 
 
Department of Veterans Affairs (VA) Comments to Government 
Accountability Office (GAO) Draft Report Veterans' Benefits: Improved 
Management Would Enhance VA's Pension Program (GAO-08-112): 

In order to enhance VA's management of the pension eligibility process, 
we recommend that the Secretary of Veterans Affairs direct the Under 
Secretary for benefits to take the following actions: 

* Take steps to make more accurate and timely decisions about 
initial and ongoing pension eligibility and payment levels. Such 
actions might include requiring pensioners to submit supporting 
documentation of their income and assets, conducting a more 
robust independent verification with third parties, or maximizing 
process automation.

Concur in part: VA agrees that additional steps are appropriate to 
improve the timeliness of pension claims processing and has already 
taken the following actions: 

* VBA increased the number of resources dedicated to claims processing 
in 2007 and further increases are expected in 2008. 

* VBA will complete the consolidation of all pension claims processing 
to the Pension Maintenance Centers (PMC) in 2008. 

* VA exceeded the FY 2007 goal of 45 days to complete eligibility 
verification reports. Actual performance was 34 days. 

VA does not agree that new, regular, and perhaps onerous submissions of 
income and net worth documents by pension beneficiaries, the average 
age of whom is 70 years, would yield substantive improvements in the 
pension program. GAO submitted no evidence to support the need for the 
suggested actions. Further, requesting and evaluating such documents 
from claimants and/or beneficiaries may diminish timeliness. 

VA already has many means in place to independently verify the income 
of pension beneficiaries. An original pension decision is based on 
estimated income that the claimant reports he or she expects to receive 
during the first year of eligibility. When VA needs information about 
the Social Security Administration (SSA) benefits, VA can obtain the 
information directly from SSA through an interface with our VETSNET 
system's Share application. Decision-makers have this interface at 
their disposal, as well as the output of matching agreements with SSA 
when making subsequent income adjustments. The matching programs act in 
two ways to ensure accuracy: first, the SSA verified rate is inserted 
in the VA master record; and second, information is sent to the PMCs 
when verified Social Security income differs from that upon which 
benefits are being paid. Additionally, VA obtains civil service and 
railroad retirement annuity information through inter-agency matching 
agreements. These matches are effective in obtaining accurate income 
data. On page 7 of the draft report, table 1 shows that pensioners 
under age 65 receive 91 percent of their total income from SSA, and 
those age 65 and older receive 87 percent of total income from SSA, 
which, as stated above, is a verifiable source. 

We agree that maximizing the use of automation is essential to 
achieving additional process improvements. In August 2008, VBA will 
release VETSNET enhancements that will make this system considerably 
easier to use. Enhancements include capturing income data and 
displaying it from year to year. When entitlement to both has been 
established, automatically change payment to the veteran to either 
service-connected disability compensation or nonservice-connected 
pension whichever is greater.

* Take steps to improve its quality assurance review of initial 
claims, which could include reviewing a larger sample of pension 
claims. 

Concur – VBA's quality assurance program will double the case review 
for the entire rating sample during FY 2008, including initial pension 
claims. 

* Incorporate evaluative feedback more consistently into the 
planning, design, and implementation of its training efforts. 

Concur–VA has developed standardized pension training based on M21-1 MR 
materials posted on the Compensation and Pension Service (C&P) training 
Web site and locally developed training aids. PMC review exercises also 
measure the success of the training. To ensure a consistent means of 
providing evaluative feedback to local training, VA will post the 
general evaluation forms for C&P training to the Web site by the end of 
June 2008. 

Additionally, all field personnel have standardized Training and 
Performance Support System (TPSS) modules with pension segments. The 
TPSS training system allows local training coordinators to measure the 
quality of training and success of each veterans service representative 
enrolled in the system. 

* Evaluate options for improving the effectiveness of its annual 
eligibility verification review. This effort could involve 
reformulating the EVR process by (a) surveying every pensioner 
rather than a selected subset; (b) performing reviews on a rolling 
basis, such as on an individuals' anniversary date, rather than 
diverting staff for this activity for a 3-month period; (c) reviewing 
pensioners' eligibility once every few years rather than every 
year; and (d) focusing on verifying pensioners' income and 
assets. 

Concur – Although there are inconsistencies between the options 
presented, VBA will consider ways to improve the effectiveness of the 
eligibility verification review. Originally, VBA issued EVRs throughout 
the year, or whenever income information was needed from a beneficiary. 
Public Law 103-271 gave the Secretary discretion in requiring annual 
EVRs. As a result of that legislation, VA eliminated annual EVRs for 
most beneficiaries who have no income or only Social Security income. 
This reduced the EVR workload to a level that allows staff to issue and 
work them all seasonally. 

At the end of 1995, VA converted the staggered (rolling) reporting 
period to a calendar year reporting period for all beneficiaries 
required to file an EVR. Before then, VBA based staggered reporting 
period on the anniversary date of when entitlement to pension was first 
established. One important reason for this change was to make it easier 
for veterans, whose income information (e.g., Form 1099) was sent to 
them at the end of the calendar year for tax purposes, to accurately 
report income from all sources. Most beneficiaries also have official 
unreimbursed medical expense reports provided to them on a calendar 
year basis. Unreimbursed medical expenses are used to reduce countable 
income and increase pension payments. Staggering reporting periods 
would result in a substantial increase in workload because most 
beneficiaries would still file their medical expenses at the end of the 
year. 

* Seek legislative authority to facilitate use of NDNH earnings data 
for pension eligibility purposes. 

Concur - The draft GAO report misstates VA's position about accessing 
the National Directory of New Hires (NDNH) database. VBA has initiated 
the process necessary to gain access to the NDNH earnings database. 
Proposed legislation was submitted in both the 109th and 110th Congress 
to obtain access to the NDNH database, and Congress recently authorized 
access to the database in Public Law 110-157. VA is initiating contact 
with the Department of Health and Human Services (HHS) to gain access. 
Before VA can begin using the system, the following two key steps must 
be completed. 

* VBA will enter into a memorandum of understanding (MOU) with the HHS, 
Office of Child Support Enforcement (OCSE), which maintains the NDNH 
database. The MOU will describe the purpose, legal authority, 
justification, and expected results of the data matches and the records 
to be matched. In addition, retention and disposition of data, as well 
as reporting requirements, will be addressed. VBA will meet with OCSE 
on January 31, 2008, to discuss the MOU. 

* Because the enabling legislation contains a provision that requires 
VBA to obtain the consent of the individual pension applicant prior to 
using the NDNH database, it will be necessary to modify language in the 
pension applications and other forms to implement this requirement. 

VA anticipates completing these steps during FY 2009. 

* Update the IVM data match threshold level to be in line with the 
U.S. Census Bureau poverty threshold or a comparable measure. 

Concur — VBA will adjust the IVM data match threshold level to conform 
to the updated setting of the U.S. Census Bureau poverty threshold 
level. We will submit the project initiation request to adjust the 
threshold by April 2008. 

* Evaluate the causes of improper payments and use the results to 
develop and implement an action plan to prevent them from 
occurring. 

Concur - VBA will review and analyze all available information, 
including Systematic Technical Accuracy Review (STAR) error reports and 
debt management information, concerning the causes of improper pension 
payments. Based on the outcome of that analysis, VBA will identify and 
implement feasible system and/or process interventions to minimize the 
occurrence of improper payments. We anticipate completion of an action 
plan to implement the interventions by the end of June 2008. 

In addition, VA would like to note that we have been reporting on 
improper payments as required annually in the Annual Performance and 
Accountability Report (PAR). For FY 2008, OMB requires all agencies to 
report causes of improper payments. On page 4, GAO states "VA does not 
collect sufficient data on causes of improper payments that could be 
used to help it better manage the pension program." VA's Office of 
Finance has drafted a report format with instructions for the Veterans 
Health Administration and the Veterans Benefits Administration to 
report the causes of errors, dollar amounts, and percentages to 
the Office of Financial Policy, which compiles data for reporting in 
the PAR and routine reporting under the Improper Payments Information 
Act of 2002. 

[End of section] 

Appendix IV: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

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

Staff Acknowledgments: 

The following individuals made important contributions to the report: 
Brett Fallavollita, Assistant Director; Anna Kelley; Scott Purdy; 
Shannon Diamant; as well as Susan Bernstein; Pat Elston; Lara Laufer; 
Wayne Turowski; Walter Vance; Joan Vogel; and Craig Winslow. 

[End of section] 

Related GAO Products: 

Veterans Affairs: Continued Focus on Critical Success Factors Is 
Essential to Achieving Information Technology Realignment. GAO-07-844. 
Washington, D.C.: June 15, 2007. 

Veterans Benefits Administration: Progress Made in Long-Term Effort to 
Replace Benefits Payment System, but Challenges Persist. GAO-07-614. 
Washington, D.C.: April 27, 2007. 

Veterans' Disability Benefits: Processing of Claims Continues to 
Present Challenges. GAO-07-562T. Washington, D.C.: March 13, 2007. 

Veterans' Disability Benefits: VA Can Improve Its Procedures for 
Obtaining Military Service Records. GAO-07-98. Washington, D.C.: 
December 12, 2006. 

Veterans' Disability Benefits: VA Should Improve Its Management of 
Individual Unemployability Benefits by Strengthening Criteria, 
Guidance, and Procedures. GAO-06-309. Washington, D.C.: May 30, 2006. 

Veterans' Benefits: Further Changes in VBA's Field Office Structure 
Could Help Improve Disability Claims Processing. GAO-06-149. 
Washington, D.C.: December 9, 2005. 

Veterans' Benefits: VA Needs Plan for Assessing Consistency of 
Decisions. GAO-05-99. Washington, D.C.: November 19, 2004. 

Veterans' Benefits: More Transparency Needed to Improve Oversight of 
VBA's Compensation and Pension Staffing Levels. GAO-05-47. Washington, 
D.C.: November 15, 2004. 

Veterans' Benefits: Improvements Needed in the Reporting and Use of 
Data on the Accuracy of Disability Claims Decisions. GAO-03-1045. 
Washington, D.C.: September 30, 2003. 

Veterans Benefits Administration: Process for Preventing Improper 
Payments to Deceased Veterans Can Be Improved. GAO-03-906. Washington, 
D.C.: July 24, 2003. 

Major Management Challenges and Program Risks: Department of Veterans 
Affairs. GAO-03-110. Washington, D.C.: January 1, 2003. 

Veterans' Benefits: Claims Processing Timeliness Performance Measures 
Could Be Improved. GAO-03-282. Washington, D.C.: December 19, 2002. 

Veterans' Benefits: Quality Assurance for Disability Claims and Appeals 
Processing Can Be Further Improved. GAO-02-806. Washington, D.C.: 
August 16, 2002. 

Veterans' Benefits: Despite Recent Improvements, Meeting Claims 
Processing Goals Will Be Challenging. GAO-02-645T. Washington, D.C.: 
April 26, 2002. 

[End of section] 

Footnotes: 

[1] Pension benefits are administered through three programs commonly 
referred to as the Old Law (Act of Mar. 20, 1933, ch. 3, 48 Stat. 8), 
Prior Law (Veterans' Pension Act of 1959, Pub. L. No. 86211, 73 Stat. 
432), and Improved Pension (Veterans' and Survivors' Pension 
Improvement Act of 1978, Pub. L. No. 95-588, 93 Stat. 2497) programs. 
The Improved Pension program is the largest of VA's three current 
pension programs. We focus on the Improved Pension program in this 
report. Since it was established, it is also the only pension program 
that has remained open to new claimants. In fiscal year 2006, the 
Improved Pension program represented over 90 percent of the $3.5 
billion in pension benefits administered by VA. 

[2] 38 U.S.C. § 1110. 

[3] 38 U.S.C. §§ 1513 and 1521. 

[4] VA uses pension cases as its unit of analysis in all budget data. A 
small percentage of cases include a pensioner's dependents. However, VA 
generally discusses data on pension cases in terms of "veterans" and 
"survivors." Since a case may include dependents in addition to a 
qualifying veteran or surviving spouse, more individuals are served by 
the program than the number of cases shown here. 

[5] 38 U.S.C. § 5312(a). 

[6] 38 U.S.C. § 1506(3). 

[7] 38 C.F.R. § 3.256(b) (2006). The required report is termed the 
Eligibility Verification Report (EVR). 

[8] Income information is estimated from GAO's analysis of the VA 
Benefits Delivery Network October 2006 5 percent sample, which provides 
income amounts as calculated by VA for pension eligibility purposes. 
All numerical estimates other than percentages have margins of error of 
plus or minus 17.65 percent or less of the value of those numerical 
estimates, unless otherwise noted. 

[9] ORC Macro, Economic Systems Inc., and Hay Group, Evaluation of VA 
Pension and Parents' DIC Programs: VA Pension Program Final Report 
(Washington, D.C.: Dec. 22, 2004). 

[10] Aid and attendance data does not include adult children receiving 
pension benefits. 

[11] ORC Macro, Economic Systems Inc., and Hay Group, Evaluation of VA 
Pension and Parents' DIC Programs: VA Pension Program Final Report, 
(Washington, D.C.: Dec. 22, 2004). pp. 46-47. 

[12] The numbers served represent the average number of cases during 
the year. Since a case may include dependents in addition to a 
qualifying veteran or surviving spouse, more individuals are served by 
the program than the number of cases shown here. 

[13] By comparison, SSA requires applicants for Disability Insurance or 
Supplemental Security Income benefits to submit supporting evidence of 
their income and assets, such as pay stubs, W2 forms, tax returns, bank 
statements, and other documents verifying the income numbers provided. 

[14] In 2006, while examining VA's Individual Unemployability benefits, 
GAO recommended that VA "seek legislative authority to use earnings 
data from the National Directory of New Hires." See GAO, Veterans' 
Disability Benefits: VA Should Improve Its Management of Individual 
Unemployability Benefits by Strengthening Criteria, Guidance, and 
Procedures, GAO-06-309 (Washington, D.C.: May 2006). 

[15] Pub. L. No. 110-157, § 301,121 Stat 1831. 

[16] VA reviews a total of 10 "rating-related" claims from each of its 
57 regional offices, doubling this amount for the largest and poorly 
performing offices. These cases include compensation, pension, appeals, 
and other types of cases. 

[17] VA Office of Inspector General, Audit of Adjustments of 
Hospitalized Veterans' Compensation and Pension Benefits, Adjustments 
of Compensation and Pension Awards for Hospitalized Veterans Needed to 
be Improved, 05-01143-71, Feb. 1, 2007. 

[18] GAO, Human Capital: A Guide for Assessing Strategic Training and 
Development Efforts in the Federal Government, GAO-04-546G (Washington, 
D.C.: Mar. 1, 2004). 

[19] According to Social Security Administration Publication No. 05-
11011, March 2006, ICN 480265, SSI beneficiaries are required to report 
and document their earnings on a regular basis. 

[20] EVRs are only sent to pensioners who previously reported some 
source of household income beyond their SSA or VA benefit levels. 

[21] VA identified a savings of $11,611,609 from the IVM matches with 
SSA and IRS during fiscal year 2004 and a savings of $13,499,698 during 
fiscal year 2005. 

[22] In 2006, VA began processing 2 tax years' worth of information. 
However, because IRS data is 2 years behind, this means that it 
processed information for the tax years 2002 and 2003 during that year. 
In 2007, VA began to process tax years 2004 and 2005 while it was still 
working on some of the earlier cases. 

[23] NDNH can provide more current earnings data than SSA, but, until 
December 2007, VA did not have the statutory authority to access this 
database. 

[24] The IU program provides disability compensation benefits to 
certain veterans whose service-connected disabilities have caused them 
to be unemployable. VA places an earnings limit on the continued 
receipt of IU benefits, and uses the IVM to ensure this requirement is 
met. 

[25] VA's IVM data match threshold was set at $6,000 in 1988. VA told 
us that the $6,000 threshold reflected at the time a cutoff for 
marginal employment, which the agency partially defined as employment 
insufficient to sustain a living. VA told us poverty-level wages were 
also used in its calculation. The U.S. Census Bureau's poverty 
threshold for one individual increased from $6,022 in 1988 to $10,294 
in 2006. 

[26] GAO-06-309. 

[27] 38 U.S.C. § 5302. 

[28] For productivity purposes, VA refers to these claims as End 
Product Code (EP) 180 and 190, respectively. 

[29] 38 U.S.C. § 1521(j). Veterans with active service generally have 
significantly more than 90 days of such service, and all six periods of 
war applicable under the statute to the pension program, 38 U.S.C. § 
1501(4), exceed 90 days, 38 U.S.C. § 101(7), (8), (9), (29), (30), and 
(33). 

[30] Under the statute, the period of the Persian Gulf War began August 
2, 1990, and will end "on the date thereafter prescribed by 
presidential proclamation or by law." 38 U.S.C. § 101(33). Since no 
such proclamation has been issued or law enacted, this period of war 
has never ended. The statute also provides for additional periods of 
war commencing with the declaration of war by Congress and terminating 
by presidential proclamation or concurrent resolution of Congress. 38 
U.S.C. § 101(11). 

[31] Veterans with dishonorable discharges may obtain an upgrade to 
their discharge status in certain cases, however, and thus become 
eligible for the pension program. 38 U.S.C. § 5303(e). 

[32] 38 U.S.C. § 5303A. The requirement applies to officers only if 
they entered (or enter) active duty after October 16, 1981. 

[33] VA defines "substantially gainful employment" for both pensions 
and disability compensation as employment that is "ordinarily followed 
by [individuals without disabilities] to earn their livelihood with 
earnings common to the particular occupation in the community where the 
veteran resides." Web Automated Reference Materials System M21-1MR, pt. 
V, subpt. II, ch. 1, § B. In a 2002 survey, 1 percent of veterans and 
survivors under age 65 said they were working, while 4 percent of 
veterans and 10 percent of survivors said they were seeking work. Most 
of the remaining pensioners under of the age of 65 said they were not 
working or seeking work because of health problems or disabilities. 
(ORC Macro, Economic Systems Inc., and Hay Group, Evaluation of VA 
Pension and Parents' DIC Programs: VA Pension Program Final Report 
[Washington, D.C.: Dec. 22, 2004]). 

[34] The required report is termed the Eligibility Verification Report 
(EVR). 

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: