How Government Savings Claims Get Made — What the GAO Found in DOGE's Wall of Receipts

On August 6, 2026, the U.S. Government Accountability Office released a report finding that the Department of Government Efficiency’s public “Wall of Receipts” — a dashboard claiming $110 billion in savings — contained estimates that were “incorrect or lack supporting evidence.” The agency reviewed DOGE’s methodology across contracts, grants, and leases, and found systematic problems with how savings were calculated, attributed, and verified.

The report matters because it touches on a question that goes beyond one agency: how do you measure whether the government actually saved money? The answer turns out to be complicated enough that even a dedicated dashboard can produce numbers that look precise while being largely unverifiable.

What DOGE claimed

DOGE was established by Executive Order 14158 on January 20, 2025, reorganizing the U.S. Digital Service into the U.S. DOGE Service within the Executive Office of the President. Its mandate: modernize federal technology and maximize productivity. The agency operated without statutory authority or permanent staff, with a temporary structure scheduled to end July 4, 2026. By November 2025, reports indicated it had ceased existing as a centralized entity.

The “Wall of Receipts” was DOGE’s public-facing dashboard, displaying estimated savings from terminated contracts, canceled grants, and lease modifications. As of July 7, 2026 — the snapshot date for the GAO review — it claimed $110 billion in cumulative savings. The dashboard presented these figures as concrete achievements: receipts documenting money the government no longer had to spend.

What the GAO found

The GAO’s review — report GAO-26-108615 — examined whether DOGE’s savings estimates were transparent, methodologically sound, and supported by evidence. The findings broke down differently across the three categories of claimed savings.

Contract savings. DOGE reported savings from terminated or modified federal contracts. The GAO found that only 43.2% of these contract savings could be tied to actually terminated contracts. For the rest, the agency either listed contracts that remained active, used savings figures that didn’t match the stated methodology, or provided no documentation linking a dollar amount to a specific contract action.

Grant savings. The results were worse here. The GAO found that 96% of DOGE’s grant savings claims were unverifiable. The dashboard listed dollar amounts attributed to canceled or reduced grants, but in nearly every case the agency could not produce evidence that the grants had been terminated, that the savings matched the grant values, or that the money would not have been spent elsewhere.

Lease savings. DOGE claimed savings from federal lease terminations and modifications. The GAO found that explanations for these savings were largely missing. Without documentation of what leases were affected, under what terms, and how the savings were calculated, the figures on the dashboard had no traceable basis.

The GAO’s overall conclusion: the Wall of Receipts lacked the transparency needed for policymakers to use its data. The numbers were not random — they reflected real actions DOGE took or claimed credit for — but the connection between those actions and the dollar amounts displayed was too thin to support the precision the dashboard implied.

Why measuring government savings is hard

The GAO’s findings are not merely a critique of one agency’s accounting. They expose structural problems that any effort to measure government efficiency will encounter.

Cost avoidance versus actual savings. Government procurement distinguishes between “hard savings” — spending that falls below a previously established baseline, reducing the budget — and “cost avoidance” — preventing future increases so the budget rises less than it otherwise would have. A terminated contract that eliminates $10 million in future payments is a hard saving only if that money was already budgeted. If the contract was up for renewal and the agency simply chose not to renew, the $10 million never entered the budget in the first place. It’s cost avoidance, not savings. DOGE’s dashboard did not consistently distinguish between the two.

Baseline dependence. Every savings figure depends on a baseline — the amount of money that would have been spent without the intervention. But baselines are projections, not facts. A contract scheduled for renegotiation might have been renewed at a lower rate anyway. A grant program might have been funded at a reduced level in the next budget cycle. DOGE’s savings estimates often assumed the full remaining value of a contract or grant would have been spent, without accounting for normal renegotiation, expiration, or budget adjustments.

Attribution problems. When a federal agency terminates a contract, who gets credit for the savings? If the decision came from the agency’s own budget office, responding to congressional pressure or internal priorities, it’s unclear whether an external efficiency initiative should claim the result. The GAO found cases where DOGE listed contracts as terminated that appeared to follow normal procurement cycles rather than targeted efficiency actions.

The timing mismatch. Government spending operates on multi-year cycles. A contract terminated in 2025 might have savings realized over three years, contingent on the agency not replacing the service with an equivalent purchase. DOGE’s dashboard presented future projected savings as if they were already achieved — a “receipt” for money that might never have been at risk.

What the numbers could mean

The GAO did not conclude that DOGE fabricated its $110 billion figure. The report found estimates that were incorrect, unsupported, or unverifiable — which is a different claim than alleging deliberate deception. Some of the actions DOGE took may have genuinely reduced spending. Some contracts may have been terminated with real fiscal impact.

But the gap between what the dashboard displayed and what the GAO could verify was large enough to raise questions about what the $110 billion actually represented. At minimum, it included:

• Savings from contracts that were terminated and documented (the 43.2% of contract savings the GAO could trace)
• Projected future savings from actions taken, valued at full remaining contract or grant amounts without discounting for renegotiation or replacement spending
• Cost avoidance figures presented alongside actual budget reductions
• Claims the agency could not substantiate with supporting documentation

The GAO’s role is to assess whether federal programs are operating as intended and whether reported results are reliable. Its finding that DOGE’s Wall of Receipts lacked transparency means that Congress, oversight bodies, and the public had no reliable way to tell which parts of the $110 billion reflected real savings and which reflected optimistic projections, accounting conventions, or errors.

The broader problem

Government efficiency initiatives face a measurement challenge that private companies do not. A corporation can compare this quarter’s spending to last quarter’s and attribute the difference to a cost-cutting program. The government operates across hundreds of agencies, thousands of programs, and multi-year appropriations cycles. Savings in one area may simply shift costs to another. Terminating a contract may reduce immediate spending while increasing long-term maintenance costs elsewhere.

The GAO has long warned about the difficulty of measuring federal performance. Its work on DOGE’s Wall of Receipts is a specific instance of a general problem: when an organization claims to have saved money, the claim is only as credible as the baseline, methodology, and documentation behind it. A dashboard that displays dollar amounts without showing how they were calculated produces the appearance of precision without the substance.

DOGE’s temporary structure and lack of statutory authority compounded the problem. An agency without permanent staff or a defined reporting framework has less incentive — and fewer mechanisms — to maintain the documentation trail that makes savings claims auditable. The Wall of Receipts was designed for public communication, not fiscal accountability. Those are different purposes, and the GAO’s report shows what happens when a communications tool is treated as an accounting record.

What would make savings claims credible

A government savings dashboard that could withstand audit would need several things DOGE’s did not provide:

• Clear baselines. Each savings figure should state what spending would have occurred without the intervention, with documentation of why that baseline is reasonable.
• Distinction between hard savings and cost avoidance. Money actually returned to the budget should be separated from money that was never at risk of being spent.
• Traceable documentation. Every dollar on the dashboard should link to a specific contract, grant, or lease — with evidence of termination, modification, or cancellation.
• Realized versus projected savings. Money already saved should be distinguished from future projections, which depend on assumptions that may not hold.
• Independent verification. Savings claims should be reviewed by an independent body before being presented as achieved.

These are not unusual requirements. They are standard accounting practices applied to a context where the stakes — public trust in government efficiency — are high and the incentives to overstate results are structural.

The open question

The GAO report does not tell us how much of DOGE’s $110 billion was real savings, how much was cost avoidance, and how much was unverifiable projection. That information may not exist in a form that can be reconstructed after the fact. DOGE’s temporary structure meant it operated without the documentation infrastructure that makes retrospective audit possible.

The Wall of Receipts is still accessible. The numbers are still displayed. But after the GAO’s review, a viewer needs to know that those figures represent claims — not verified results — and that the gap between the two can be substantial.