Tracking Taxpayer Dollars: Fixing a Flawed System for Federal Spending Data

Download

Executive Summary

The federal government spends trillions of taxpayer dollars every year. Agencies across the federal government use contracts, grants, direct payments, loans, and other types of awards to distribute funds and to implement agency plans. Public transparency for that spending creates accountability and helps deter waste, fraud, and abuse.

It is for these reasons that USAspending.gov was created almost 20 years ago: to provide the public an easy-to-use window into the federal government’s spending.¹ The site has become the primary source of data on federal award spending.

Robust transparency enables review and oversight by a broad array of actors: federal agencies, state and local officials, researchers, journalists, and even the general public. Spending data, when complete, accurate, and substantive, allows anyone to examine where federal funds went and what the funds accomplished.

To determine how well current federal award reporting meets those standards, the Project On Government Oversight (POGO) conducted a review of the reported award data for the 60 largest domestic-focused federal assistance programs. These programs account for 92.3% of total federal assistance spending for fiscal year (FY) 2024. This review evaluates several aspects of the quality and usefulness of the reported data.

Specifically, the review examines

• how much of each program’s spending can be tracked to the local level;
• how granularly programs report place of performance location data (county, city, zip code);
• how much subaward reporting helps in tracking federal funds to the local level; and
• how well transaction descriptions explain the exact uses of federal funds.

Highlighted Findings

The vast majority of federal grant spending cannot be tracked to the local level.

• More than 74% of federal grant funding could not be tracked to any local level (zip code, city, or county) in our review.
• That is $712.6 billion in federal spending that the public can’t tell exactly where it went beyond the state locations.

Subaward reporting, required by federal law, should explain where federal funds go after being awarded to intermediary recipients such as state agencies. The subaward data, however, is fundamentally broken and incomplete, offering little to no additional information on the flow of federal funds.

• In our review, prime awards only had local-level data for 20.7% ($198.3 billion) of awarded federal grant funds.
• Subaward data only accounted for an additional 5% ($47.6 billion) of federal grant dollars that could be tracked to the local level.

According to our review, the vast majority of direct payments were trackable to the local level, though a few programs showed signs of significant errors in the data. However, the granularity of the local place of performance for these awards was limited.

• Direct payment awards were trackable to the county level for 96.1% of the federal funds spent through this award type.
• Less than 3% of the prime direct payments were trackable to either a city or zip code level.

Transaction descriptions, which should provide specific details and context for each award transaction, instead are most often uninformative, duplicate entries.

• Just under 97% of the 2.4 million reviewed transactions used duplicate descriptions.
• Direct payment and loan programs reported unique descriptions for just 0.15% and 0.01% of transactions, respectively.

Detailed Findings

USAspending.gov is an important and useful website with valuable data on federal spending that helps inform the public in many ways. But as this review found, the spending data on USAspending.gov also has significant shortcomings that restrict its usefulness:

• majority of federal grant funding cannot be tracked to local level
• limited geographic details for direct payment program spending
• incomplete subaward reporting
• uninformative award descriptions

These reporting issues have real-world consequences. Significant flaws and gaps in spending data create an incomplete picture of how tax dollars were used. That limited understanding makes it more difficult to detect fraud and waste, which leads to less accountability.

Grants: Majority Cannot be Tracked to Local Level

Tracking where federal funds ultimately end up in communities is a critical functionality for public transparency and oversight. If we don’t have a clear understanding of which communities received federal funds and what the funds were used for in those communities, then we cannot evaluate the effectiveness of federal programs and agencies.

POGO’s review identified federal grant spending as the least trackable to local communities. Grants are one of the primary mechanisms through which federal funds flow to states, local governments, school systems, health care providers, and nonprofit organizations. Yet despite the scale and importance of these programs, the overwhelming majority of grant funding cannot be tracked to the local level using publicly available federal spending data.

POGO’s analysis reviewed approximately $958.6 billion in domestic grant spending from FY 2024. Of that amount, more than 74% — approximately $712.6 billion — could not be traced to any specific county, city, or zip code.

Only 20.7% of prime awards included local place of performance data. This means that out of the $958.6 billion expended by these grants during FY 2024, only $198.3 billion could be traced from beginning to end.

Ideally, subaward reporting would fill significant portions of the location data gap left by prime award reporting and provide additional visibility into where grant funds ultimately go after being distributed by intermediary recipients such as state governments. However, the available subaward data for these grant programs only helped track an additional 5% of federal spending to the local level.

This represents a major spending transparency failure for USAspending.gov. Many large federal grant programs are designed specifically to support local communities , yet the current data leaves the public unable to determine where those funds ultimately reach. This breakdown in transparency occurs primarily after federal agencies distribute funds to states or other intermediaries. Existing subaward reporting systems fail to provide sufficient information about how those funds are distributed to local governments, schools, health care providers, nonprofit organizations, or households.

The grant programs that cannot be tracked to the local level include numerous major programs. One such program is Medicaid, which distributed more than $614 billion in federal funds to support individuals and families with health care costs. Only 0.2% of that can be tracked to the local level. Similarly, of the $19.4 billion spent through the Children’s Health Insurance Program, only 0.1% can be tracked to specific communities. Notable transportation programs also had poor performance in tracking funds. Federal Transit Formula Grants spent $8.2 billion and State of Good Repair Grants Program spent $4.6 billion, but the programs reported local place of performance information for only 1.8% and 0.5% of funds , respectively.

Direct Payments: High Coverage, but Low Detail

Direct payment assistance programs performed substantially better than grant programs in terms of local traceability. As the name of the award type implies, direct payment programs frequently provide benefits directly to households or individuals. Because these programs often award funds to local recipients, the prime award report more regularly contains local place of performance data. The review found that 96.1% of direct payment spending could be tracked to the county level based on prime award data.

However, three major Medicare programs that make direct payment awards had location results that make it obvious the data contains significant errors. The Supplementary Medical Insurance, Prescription Drug Coverage, and Hospital Insurance programs each had several states for which no spending was reported at all, including Mississippi, Vermont, and New Hampshire. Many government sources make it clear that these Medicare programs reach all 50 states. The complete absence of whole states in the award reporting for these three programs can be observed across multiple fiscal years, making it clear that the problem is systemic and not a one-time reporting error.

Additionally, while it is well established that states with higher populations, such as California, New York, Texas, and Florida receive the largest amounts of Medicare spending, the USASpending.gov data had less populous states, such as Minnesota, Indiana, and Kentucky, with the largest amounts from these three programs. None of the other direct payment programs exhibited such obvious systemic errors. These results make it clear that while all the reported awards for these three programs contain county locations, a significant portion of the place of performance data is incorrect. However, POGO is unable to determine how much of the combined $1.03 trillion the three programs spent in FY 2024 correlates to incorrect location data. As such, in this report, the programs receive credit for reporting a local place of performance 100% of the time, though POGO notes on the tables that these programs have serious data quality concerns.

A broader concern for the direct payment programs is that the geographic detail reported for many of the awards was highly limited. Nearly all of the reported geographic information for direct payment programs was restricted to counties. Only around 3% of direct payment awards included location information at the city or zip code level.

County-level reporting provides an important baseline level of transparency. However, counties often encompass large and economically diverse populations. Limiting reporting of the distribution of federal funds for these programs to counties makes it impossible to evaluate if and how resources reach specific communities or neighborhoods within those areas. It makes detecting potentially suspicious or anomalous distribution patterns, which can be early indicators of potential fraud, more difficult.

More detailed reporting on the distribution of these direct payment awards at the city or zip code level — while still protecting individual privacy through aggregation — would significantly improve transparency and accountability. Greater geographic granularity would allow reviewers to better identify unusual payment patterns, evaluate equity in program distribution, and assess whether resources are reaching intended populations at the neighborhood and community level.

The direct payment programs that cannot be tracked below the county level with current available data include programs with widespread impact that would benefit from more granular reporting on where funds reached. For example, three major Medicare programs, including Supplementary Medical Insurance, Prescription Drug Coverage, and Hospital Insurance, spent a combined $1.03 trillion in FY 2024. None of that can be tracked to a city or zip code level. Similarly, the Veterans Compensation for Service-Connected Disability program’s $142.4 billion can only be tracked to the county level and has no city or zip code data reported. The lack of more granular geographic location reporting means that these programs receive less oversight and review. Greater geographic detail can allow reviewers to identify inconsistencies and outliers in the spending patterns, such as communities getting far more or less than expected. Such anomalies can help uncover fraud or significant program failings.

Subaward Reporting: Fundamentally Broken

POGO’s review indicates that subaward reporting on USAspending.gov currently adds very little meaningful transparency to federal spending data. In theory, subaward data should help create a far more complete picture of federal spending and provide meaningful visibility into the flow of taxpayer dollars from federal agencies to local recipients and communities. In practice, however, subaward reporting is fundamentally broken and incomplete.

The subaward problem is particularly evident among the reviewed federal grant programs. Since grants often pass through intermediary recipients, robust subaward reporting would be expected. However, as detailed in the grants section, subaward reporting for the grant programs only provided local-level tracking for an additional 5% of reviewed grant dollars (about $47.6 billion).

Subaward reporting problems were also evident outside of the grant programs. The Supplemental Nutrition Assistance Program (SNAP), a direct payment program that provides funds to low-income families to help them buy groceries, spent $84.4 billion; none of it was trackable to the local level. The vast majority of prime awards were paid to state agencies that administer SNAP within their states. Those state agencies did not report any subawards detailing how the funds were distributed across the states.

There were also two programs that made direct payments to companies to improve broadband access. The Universal Service Fund – High Cost program and the Affordable Connectivity Program spent a combined $12.5 billion, none of which was trackable to the local level. The prime awards appeared to often support broadband activities across wide areas, even multistate regions, which prevented the reporting of specific local place of performance. However, neither program had any subaward reporting at all, meaning there was no breakdown as to how much work was done in which communities under these larger awards.

Importantly, these weaknesses are not new. Problems with incomplete and inconsistent subaward reporting have existed within USAspending.gov for years.

• In a 2023 report, the Government Accountability Office (GAO) noted that they and other federal agencies “have identified challenges with the completeness and accuracy of subaward data displayed on USAspending.gov.”² The provided examples of failures to report subaward data included the inability to determine how much federal money was being subawarded to entities in China and the inability to follow COVID-19 funding to the subrecipient level.
• Another 2023 GAO report reviewed award reporting for two small business assistance programs and found that while an estimated 70% of prime award recipients used subawards, only about 10% reported subawards into USAspending.gov.³
• In March 2025, a GAO study of 3,680 single audit results found that 36% of the audits uncovered difficulties in properly reporting subawards.⁴
• The inspector general for Housing and Urban Development (HUD) recently found that out of 140 prime awards across two HUD programs, 35 had at least one subaward that went unreported to USAspending.gov.⁵

Despite longstanding awareness of these issues, the federal government has failed to implement sufficient corrective measures to ensure reliable reporting across agencies and recipient entities.

Award Descriptions: Mostly Uninformative Duplicate Entries

In addition to significant gaps in tracking federal award spending to the local level, USAspending.gov data suffers from shockingly poor quality of award descriptions. Ideally, award or transaction descriptions would provide details on the exact use of the federal funds, location of work being performed, scope of services or products being paid for, and potentially even the number of people being assisted. But if you look at USAspending,gov right now, you’ll find, more often than not, vague award descriptions.

Given the shortcomings observed in the subaward data and how difficult it is to track federal assistance funds to their use in specific communities, strong descriptions of the spending transactions could help users better understand how their tax dollars were used. Unfortunately, POGO’s review of transaction descriptions found excessive use of repetitive and uninformative descriptions. Often, descriptions would include a repeated title or vague government terms, offering little sense as to the specific scope and purpose of each award. This problem was especially severe for direct payment and loan programs. These programs overwhelmingly relied on standardized or duplicate transaction descriptions that provided little to no distinction between individual awards.

Across the 2.4 million prime transactions reviewed from USAspending.gov, only 75,577 unique transaction descriptions were identified. Almost 97% of all reviewed transactions used duplicate descriptions. Of the 60 programs reviewed, 15 programs had a single unique transaction description that was duplicated for every award record reported.

The problem was particularly acute for direct payment and loan programs, which used unique descriptions for just 0.15% and 0.01% of transactions, respectively. By contrast, cooperative agreements and grants demonstrated somewhat stronger use of unique descriptions, using unique descriptions for 34.7% and 29% of transactions, respectively. However, these rates still indicate that the majority of awards reported in these grant and cooperative agreement programs were duplicates.

Many of the repeated descriptions were vague and uninformative. For example, the Mortgage Insurance Homes program, which helps people buy homes by insuring lenders against losses in mortgage loans, had more than 766,000 award records covering $231.3 billion in loans, but every award included the exact same transaction description: “Single Family Forward Loan.” No details on loan rate, length of loan, or the type of home — just the general type of loan repeated for every record.

Similarly, the Veterans Compensation for Service-Connected Disability program, which provides financial assistance to veterans with disabilities that were incurred or became more severe during military service, had almost 38,000 awards reported, covering $142.4 billion. These awards were aggregated totals of payments made in a county during a particular time period. All of the prime awards reported the same transaction description: “SVCOMP.” This uninformative acronym likely stands for service veterans compensation, but the description doesn’t even include the words. The descriptions could provide information on how many veterans are receiving benefits under the payment, the range of payments being made, the time period, and geographic region covered. Such details would help those reviewing the spending to better evaluate how the money was spent and compare it to other programs.

Recommendations for Reform

The findings of this review demonstrate that current federal spending reporting systems do not provide sufficient visibility into where large portions of federal assistance funding are going and how those funds are being distributed within local communities. Major weaknesses in grant tracking, geographic specificity, subaward reporting, and transaction descriptions significantly limit public oversight and reduce the effectiveness of fraud prevention efforts.

Addressing these problems will require coordinated action by Congress, the Office of Management and Budget (OMB), the Department of the Treasury, and federal agencies responsible for administering and reporting federal assistance programs.

Fix Subaward Reporting System

Subaward reporting represents one of the biggest weaknesses in the federal spending reporting framework. Complete and reliable subaward reporting is essential to understanding how federal dollars ultimately reach communities.

Federal guidance on subaward reporting: Federal guidance should be issued or updated to clarify reporting requirements for agencies and prime award recipients. Federal agencies should be required to regularly contact prime award recipients to inform them of their reporting responsibilities for subawards and to prompt them to check the completeness and accuracy of the subaward data they have reported. Agencies should also conduct periodic reviews of subaward data to identify potential gaps in expected reporting patterns. When such potential gaps are identified, agencies should reach out once again to relevant prime recipients and work to ensure the reported data is complete and accurate.

Ensure reporting of aggregated subawards: Agencies should take steps to ensure that subawards to households and individuals can be reported in aggregated totals for geographic areas — counties, cities, and zip codes — which improves oversight while still protecting privacy. While such aggregated reporting is fairly common on USAspending.gov for prime awards going directly to households, the practice appears to be wholly missing for subawards that state agencies are charged with distributing to households. Any technical or administrative barriers to aggregated reporting of subawards should be identified and removed to ensure that federal funds going to households and individuals by way of state agencies or other intermediaries can be properly reported and tracked on USAspending.gov.

Establish enforcement mechanisms: Stronger enforcement mechanisms should also be established to ensure improvements in compliance with subaward reporting requirements. Agencies should incorporate performance reviews of subaward reporting into routine grant oversight and monitoring activities. Persistent reporting failures should trigger the development of corrective action plans. Oversight bodies, such as agency inspectors general, should also periodically audit or evaluate agency performance in subaward reporting.

Improve Geographic Granularity

Federal spending should include granular and specific data on the use of federal awards to maximize oversight capabilities. This includes the level of geographic detail provided in prime and subaward reporting. Although some reviewed programs, mostly direct payment and loan programs, provide respectable county-level reporting, the review found that city- and zip code-level information remains limited.

Identify and correct erroneous place of performance data: Inspectors General at major agencies should review the accuracy of place of performance data for each agency’s largest assistance programs. The cause of any errors in reporting, such as those identified here for the three direct-payment Medicare programs, should be explored and fixes should be implemented to ensure consistently accurate place of performance data.

Guidance on greater geographic specificity: OMB and the Department of the Treasury should issue guidance requiring agencies to report spending data with greater geographic specificity whenever feasible. Programs reporting aggregate spending information to protect personal privacy should be instructed to report aggregate amounts for more specific geographic levels (cities and zip codes) when feasible. Only in cases where the award data cannot be narrowed to more granular locations or when doing so would risk disclosing private details due to a small sample size should agencies be allowed to limit place of performance information to larger geographic locations.

Subaward reporting requirements: The subaward reporting system needs significant fixes. One of those fixes should be OMB and the Treasury Department establishing clear guidance requiring prime award recipients to report subawards with as much geographic detail as is feasible. This especially applies to state agencies that are allocating federal program funds to households and individuals. Similar to direct payment awards made by federal agencies, these should be reported for granular geographic levels so long as privacy is still protected by such reporting.

Review of reporting greater geographic specificity: OMB and the Treasury Department should require agencies to review reporting of prime awards and subawards for their largest programs, evaluating the level of geographic specificity being achieved and identifying any barriers to granular location reporting. Agencies should publicly report the results of these reviews, any identified barriers, and recommended changes that could help make granular reporting of federal funds easier and more reliable.

Enhance Award Descriptions

Poor transaction descriptions substantially limit the usefulness of federal spending data for oversight and public understanding. Descriptions should shed light on awards and provide details about what the funds were used for and where. The excessive use of vague and repetitive descriptions offers no real additional information.

Federal guidance on description reporting: OMB and the Department of the Treasury should issue guidance establishing reporting standards for award and transaction descriptions. The standards should require agencies to report descriptions that include key specifics and meaningful context regarding the purpose and intended use of federal funds. Descriptions should include summaries of any activities, quantities of materials or services provided, and outcome or purpose of the individual award. The guidance should clearly prohibit the repetitive use of uninformative descriptions that fail to distinguish one transaction from another.

For aggregated reporting of direct payment programs, agencies should be required to provide contextual information for the geographic totals. This information should include the number of individual awards represented within the aggregation, the range of award amounts, and the average or median payment values. Providing these details would allow the public to better understand the scale and distribution of payments while continuing to protect the privacy of individual recipients.

Methodology

This review examined the 60 federal assistance programs with the largest levels of domestic spending in FY 2024. Programs with international focus were excluded from the analysis because comparable local geographic reporting standards do not always exist for international locations. For example, geographic fields such as counties and zip codes, which are central to this review, are not typically reported outside the United States.

For loan programs, reviewers used the face value of loans when determining total program spending levels. For all other programs, reviewers used reported federal obligations to determine total program spending levels.

Prime Award Analysis

Data covering the FY 2024 spending for each program was downloaded from USAspending.gov and analyzed to determine the extent of funding that could be tracked to a “local” level.

The review examined whether prime awards included information in any of the following “place of performance” fields:

• place of performance county
• place of performance city
• place of performance zip code

Awards did not have to have all three fields completed. Reported data in any of these fields qualified the award to be considered “trackable to the local level.” The spending level for each place of performance field was also calculated.

The review did not independently verify the accuracy of the reported geographic information. As a result, erroneous or invalid entries — such as incorrect zip codes or nonexistent place names — would still qualify an award as locally trackable for purposes of this analysis.

Subaward Analysis

The review also analyzed subaward data associated with each program to evaluate whether subaward reporting provided additional visibility into the local distribution of federal funds.

Subaward records on USAspending.gov include certain details from the originating prime award. If the associated prime award already included local place of performance information at the county, city, or zip code level, the related subaward records were considered already locally trackable and were not further evaluated for additional geographic detail.

For subawards associated with prime awards that lacked local place of performance information, the review examined whether the subaward records themselves included local geographic information.

The review examined whether subawards included information in either of the following subaward place of performance fields:

• subaward place of performance city
• subaward place of performance zip code

If a subaward record listed no local place of performance data for the prime award but had data for a subaward location reported in either of these fields, then the award was considered to have additional funding that could be tracked to the local level.

As with prime award data, the accuracy of reported subaward geographic information was not independently verified.

Transaction Description Analysis

The review also evaluated the uniqueness of award transaction descriptions reported on USAspending.gov.

Transaction descriptions for each reviewed program were isolated and analyzed to identify how many exact duplicates were reported.

It should be noted that minor changes such as punctuation, spacing, or small word changes were enough of a difference for descriptions to be counted as unique. As a result, the analysis may classify some descriptions as unique even when they are substantively similar or functionally duplicative. Accordingly, the findings regarding duplicate transaction descriptions may somewhat understate the overall level of repetitive or uninformative reporting present in the data.

Appendix: Data Tables for Program Results