BlackLeafwatch the watchmen
Earned Income Tax Credit

The IRS knows the working-family tax credit is wrong 27% of the time. It just cut the audits in half.

Summary

In fiscal 2024 the IRS paid $15.9 billion in erroneous Earned Income Tax Credit claims — a 27.3% error rate the agency has never gotten below 22% since 2003. The same year, EITC examination starts fell 53%, from 183,607 to 86,521, and the staff assigned to audit them fell 59%.

By Locusta · July 9, 2026

The Earned Income Tax Credit is the federal government's flagship anti-poverty tool for working families — 23 million returns claimed $64 billion of it in tax year 2023, per IRS Statistics of Income data. It is also, by the 's own accounting, the single largest source of improper payments the agency reports to Congress. The 's answer in fiscal 2024 was not to fix the error rate. It was to audit the credit half as much.

EITC improper payments, FY 2024
$15.9B
27.3% error rate
EITC examination starts, FY 2024
86,521
−53% vs. FY 2023
Workers & families claiming EITC
23M
tax year 2023

Follow the dollar

Of the $58.4 billion in EITC claims the reported for fiscal 2024 under the Payment Integrity Information Act, $15.9 billion — 27.3 cents of every dollar — was paid in error, per TIGTA's FY2024 improper-payment compliance review. That is the highest dollar total and the highest rate of the four refundable credits the Treasury Department tracks as high-risk for improper payments: worse than the Additional Child Tax Credit (10.7%), the American Opportunity Tax Credit (27.7% but a fifth the dollar volume), and the Net Premium Tax Credit (28.5%, also far smaller). The EITC alone accounts for $15.9 billion of the government's $21.4 billion total improper payments across all four credits combined.

Where the EITC dollar went, FY 2024
Total EITC claims and improper payments, $ billions
EITC claims, FY 2024$58.4BPaid correctly$42.5BPaid in error$15.9B
Source: TIGTA, Report No. 2025-400-025 (May 9, 2025), Figure 1, citing the FY2024 Treasury Department Agency Financial Report
View data as table
EITC claims and improper payments, FY 2024
EITC claims, FY 2024 (total)$58.4BTreasury AFR / PIIA reporting
Paid correctly$42.5B72.7% of claims
Paid in error$15.9B27.3% improper payment rate

This isn't new. 's own trendline shows the EITC's estimated error rate has run above 22% in every year since fiscal 2003, swinging between 23% and 34% over two decades regardless of which administration or which enforcement strategy was in charge. The has told plainly why: EITC eligibility hinges on facts the agency cannot verify at filing time — who a "qualifying child" actually lived with for more than half the year, and how much cash or gig income a filer actually earned. Without independent third-party data to check those claims against, the Treasury Inspector General for Tax Administration found the could deny half of all currently-claimed EITC amounts and still fall short of the 10% improper-payment target Congress set.

The audits that would catch it got cut in half

If the error rate is structural, the obvious lever is verification — examining more of the returns most likely to be wrong before the refund goes out. The pulled the opposite lever. A separate audit found that the Taxpayer Services Division, which runs most EITC enforcement through its Refundable Credits Examination Operations function, cut actual EITC examination starts by 53% between fiscal 2023 and fiscal 2024 — from 183,607 down to 86,521 — while shifting its casework toward non-EITC audits instead, per TIGTA's FY2024 EITC examinations report.

EITC examination starts, FY 2023 vs. FY 2024
Audits opened by the IRS Taxpayer Services Division, count
FY 2023 EITC exam starts
183,607
FY 2024 EITC exam starts
86,521
Source: TIGTA, Report No. 2025-30-802 (May 2025), Figure 1
View data as table
EITC examination starts by fiscal year
FY 2023 EITC exam starts183,607TIGTA Report 2025-30-802, Figure 1
FY 2024 EITC exam starts86,521−53% vs. FY 2023

The people who would have run those audits moved with the workload. Planned staff time (full-time equivalents) assigned to EITC examinations fell from 926 to 380 — a 59% cut — while EITC's share of the examination division's total staff time dropped from 78% to 34% in a single year, per the same TIGTA report. Non-EITC examination staffing rose 185% over the same period. The people didn't disappear from the ; they were reassigned away from the credit with the worst error rate in the federal government's improper-payment inventory.

Staff assigned to EITC audits, FY 2023 vs. FY 2024
Planned full-time-equivalent examiners, Taxpayer Services Division
FY 2023 EITC audit FTEs
926
FY 2024 EITC audit FTEs
380
Source: TIGTA, Report No. 2025-30-802 (May 2025), Figure 2
View data as table
EITC audit staffing by fiscal year
FY 2023 EITC audit FTEs92678% of RCEO planned staff time
FY 2024 EITC audit FTEs38034% of RCEO planned staff time, −59% vs. FY 2023

's stated reason for the pullback is not incompetence — it's a 2022 Treasury directive telling the not to raise audit rates on households earning under $400,000, after a 2023 Stanford study found Black taxpayers were audited at 2.9 to 4.7 times the rate of other taxpayers, driven substantially by EITC case selection. Cutting EITC audits was, in that sense, the path of least resistance to compliance with the directive: rather than build a case-selection model that filters for actual risk without the racial skew, the division simply audited far fewer EITC returns of any kind. The disparity problem may have improved. The 27.3% error rate did not move.

The takeaway

  • The error rate has never been fixed, only tolerated. Above 22% for 22 straight years, per — the EITC is a structurally hard credit to verify, not a program with a one-time bug.
  • Enforcement moved in the opposite direction of the risk. EITC exam starts fell 53% and audit staff fell 59% in the same fiscal year the credit accounted for $15.9 billion of the government's $21.4 billion in reported improper payments across four credits.
  • Fixing the racial-disparity problem and fixing the error rate were treated as substitutes, not both goals to hit. The reduced total EITC scrutiny rather than build a case-selection method that targets risk without the demographic skew — 's report says that better method still doesn't exist.

FY2024 improper-payment figures ($58.4B claims, $15.9B improper, 27.3% rate) come from Treasury's reporting cycle; the 23-million-return, $64-billion recipient figures come from Statistics of Income for tax year 2023 — a different accounting period and methodology, cited here as separate, independently sourced numbers rather than reconciled totals.

Sources

  • Treasury Inspector General for Tax Administration, Assessment of Fiscal Year 2024 Compliance With Improper Payment Reporting Requirements (Report No. 2025-400-025, May 9, 2025) — the FY2024 EITC claims total, improper payment rate, and dollar amount, plus the two-decade error-rate trendline. tigta.gov
  • Treasury Inspector General for Tax Administration, The Reduced Earned Income Tax Credit Examinations in Fiscal Year 2024, but the Process to Mitigate Racial Disparity Needs to Be Defined (Report No. 2025-30-802, May 2025) — EITC examination starts and staffing () figures for FY2023 and FY2024, and the racial-disparity policy context. tigta.gov
  • Internal Revenue Service, Earned Income Tax Credit statistics (Statistics of Income; last reviewed March 26, 2026) — return counts and total dollars claimed by tax year, 1999–2023. irs.gov
  • Hadi Elzayn et al., Measuring and Mitigating Racial Disparities in Tax Audits, Stanford Institute for Economic Policy Research (January 30, 2023) — the finding, cited by as the basis for the 2022 Treasury directive, that Black taxpayers are audited at 2.9–4.7 times the rate of other taxpayers, driven substantially by EITC case selection. dho.stanford.edu
Weekly digest: the most-read systems, in brief. Mondays.

Comments

Always open. Logged-in readers can annotate paragraphs in place.

Loading comments…
or log in to comment under your account