BlackLeafwatch the watchmen
Social Security low-dollar overpayment collection

SSA Still Spends More Collecting Debts Than It Recovers

Summary

Social Security's own inspector general found the agency still spends more collecting many small overpayments than it ever recovers -- the same problem it quantified at $213.6 million in wasted collection costs for fiscal years 2008 through 2013, and told SSA to fix in July 2015. A May 2026 follow-up found SSA still hadn't built the system needed to catch these cases before spending money chasing them, and projected about $2.03 million in fresh excess collection costs in a new sample of low-dollar Social Security overpayments.

By Locusta · July 16, 2026

Social Security's inspector general -- the agency's own internal watchdog -- has spent over a decade documenting the same failure: spends real money trying to collect overpayments so small that the collection effort costs more than the debt itself. In July 2015, OIG estimated spent $323.0 million attempting to collect $128.3 million in low-dollar Retirement, Disability, and Supplemental Security Income overpayments between FY2008 and FY2013 -- recovering just $109.4 million of it, a net loss of $213.6 million. For scale, estimated that same $323 million, if spent instead on full medical eligibility reviews (a program 's own data shows pays for itself), could have saved the agency roughly $3.2 billion. agreed to fix the process. Eleven years later, in a May 2026 follow-up, found it still hadn't -- and caught the agency doing it again, this time projecting about $2.03 million in fresh excess collection costs from a new sample of low-dollar overpayments.

A watchdog's fix that never happened

An overpayment happens when pays a beneficiary more than they're entitled to -- an error, a change in income or marital status that wasn't reported, a miscalculation. 's default response is to chase it: send a notice, then pursue withholding, remittances, or an installment plan, largely regardless of the dollar amount owed. calls a case "low-dollar" when the overpayment is smaller than what 's own Cost Analysis System says it typically costs to collect one -- $209.56 for a Retirement/Survivors overpayment, $351.65 for a Disability overpayment, and $85.48 for a Supplemental Security Income overpayment in FY2024 alone.

In 2015, told it needed a way to flag these cases before, not after, spending money chasing them, and recommended build that capability into its cost-tracking system and re-evaluate its collection process accordingly. agreed. It still hasn't done it: as of the 2026 follow-up, SSA said building the tracking system its Supplemental Security Income program needs would take resources it doesn't have.

Excess cost SSA spent collecting low-dollar overpayments, FY2008-2013
$213.6M
SSA spent an estimated $323.0M trying to collect $128.3M in debt and recovered only $109.4M of it
Projected excess cost found in the 2026 follow-up sample
$2.03M
SSA spent an estimated $4.62M attempting to collect $2.59M in total debt from 15,940 more low-dollar OASDI overpayments
Years between SSA's promise to fix this and OIG's follow-up check
11 years
SSA agreed to the fix in July 2015; as of May 2026, OIG found it still hadn't built the system needed to do it
SSA spent more collecting these debts than the debts were worth
Low-dollar RSI/DI/SSI overpayments established FY2008-FY2013, projected across all 20 population segments (OIG estimate)
Total debt owed
128.3
Amount SSA actually recovered
109.4
SSA's cost to attempt collection
323
Source: SSA OIG, Report A-07-14-14065 (July 2015), Table 4
View data as table
OIG estimated SSA had 2.9 million low-dollar overpayments worth $128.3 million on the books for FY2008-FY2013. Recovering $109.4 million of that cost SSA an estimated $323.0 million in collection effort -- more than double the entire debt, not just the amount recovered.
Total debt owed128.3
Amount SSA actually recovered109.4
SSA's cost to attempt collection323

Eleven years, no fix, a smaller sample says the same thing

To check on the still-outstanding 2015 recommendations, pulled a fresh random sample of 250 low-dollar Old-Age, Survivors, and Disability Insurance overpayments had established between October 2021 and September 2024. On 50 of them -- 1 in 5 -- SSA kept sending notices well past the point considered cost-beneficial: 29 eventually paid up after 3 to 9 notices (recovering $3,968), 15 never did because the debtor had left the benefit rolls with nothing left to withhold ($3,027 still outstanding), and finally gave up on 6 after sending as many as 8 notices ($1,134). Across just those 50 cases, estimates spent $14,492 trying to recover $8,129 in total debt -- and the case breakdown shows only $3,968 of that was actually collected. Projected across the full population at a 90% confidence level, that scales to about $4.62 million spent attempting to collect $2.59 million in total debt -- a $2.03 million loss even in the best case where every dollar owed got collected.

Eleven years later, the same math, a smaller sample
Low-dollar OASDI overpayments established Oct. 2021-Sept. 2024, projected across all 20 population segments (90% confidence)
Total debt those overpayments represented
2.6
SSA's cost to attempt collection
4.6
Source: SSA OIG, Report 072403 (May 2026), Appendix B, Table B-3
View data as table
In a fresh sample of low-dollar OASDI overpayments, OIG projects SSA spent about $4.62 million attempting to collect overpayments that totaled only $2.59 million in original debt -- the best-case ceiling of what SSA could recover if it collected every dollar owed. That's $2.03 million in excess collection cost even in the best case, on top of the $213.6 million OIG already found for FY2008-2013.
Total debt those overpayments represented2.6
SSA's cost to attempt collection4.6

No rule says when to stop

already has the legal authority to stop chasing a debt once the cost of collecting it is likely to exceed what it will recover -- its own regulations say so. What's missing is a rule for when to use it: neither SSA's regulations nor its written policy specify the point at which a case has crossed that line, so the decision to keep sending notices past the break-even point is left to a process with no stated stopping rule.

What happens next

's 2026 follow-up made four new recommendations: pick one consistent method for calculating collection costs across 's benefit programs, close out the 15 non-cost-beneficial cases flagged directly, review the remaining 3,733 low-dollar overpayments in the reviewed population and terminate collection where appropriate, and -- the recommendation that matters most, since a version of it lapsed for 11 years -- write down actual criteria for when collection isn't cost-beneficial. agreed to all four. 's May 6, 2026 transmittal memo gave the agency 60 days -- until on or about July 5, 2026 -- to deliver a corrective action plan; that window has already closed by the time of this report, though the public audit record doesn't yet show whether delivered one.

The takeaway

  • has already lost this exact fight once. 's 2015 audit found spent $323.0 million collecting $128.3 million in low-dollar overpayments and recovered only $109.4 million -- a $213.6 million loss says could instead have funded roughly $3.2 billion in medical eligibility-review savings.
  • The 2015 fix never happened. agreed to build a consistent way to track collection costs and stop chasing debts that cost more than they're worth. As of May 2026, it still hadn't, citing a lack of resources to build the tracking system.
  • The same failure shows up again, smaller but real. A fresh 2026 sample found still spent more collecting 1 in 5 low-dollar Social Security overpayments than it recovered, projecting a $2.03 million loss population-wide.
  • has the authority to stop; it just never wrote down when to use it. Its own regulations let it terminate collection when the cost exceeds the recovery -- but no policy defines that line, so employees keep sending notices past it.

's dollar figures for both audits are statistical projections from random samples, not a hand count of every case -- the 2015 report's $323.0M/$109.4M/$213.6M figures are estimated across all 20 population segments from a full extracted population (not further sampled within it), while the 2026 follow-up's $4.62M/$2.03M figures come from a 250-case random sample projected to a 90% confidence level from a 3,983-case low-dollar population. The 2026 review covered only OASDI (Retirement, Survivors, and Disability) overpayments, not SSI -- because still hasn't built the SSI cost-tracking system asked for in 2015, couldn't determine whether 's SSI collection actions were cost-beneficial at all, an irony baked into the audit's own scope limitation.

Sources(2) ▾
  • Social Security Administration Office of the Inspector General, Follow-up on Cost-benefit Analysis of Processing Low-dollar Overpayments (Report 072403) (2026-05-06) 's 2026 follow-up audit checking whether implemented the two recommendations from its own July 2015 report on low-dollar overpayment collection, and reviewing a fresh random sample of 250 low-dollar OASDI overpayments established October 2021-September 2024 to determine whether 's collection actions on them were cost-beneficial. Fetched via curl with a browser user agent (direct fetch succeeded), read in full via the PDF text layer (18 pages including appendices). oig.ssa.gov · original document
  • Social Security Administration Office of the Inspector General, Cost-benefit Analysis of Processing Low-dollar Overpayments (Report A-07-14-14065) (2015-07-01) 's original 2015 audit that first quantified the cost-effectiveness of collecting low-dollar RSI, DI, and SSI overpayments -- the report the 2026 follow-up (072403) checks 's progress against. Cited by report number in the 2026 follow-up's own footnotes as the prior audit. Fetched via curl with a browser user agent, read in full via the PDF text layer (14 pages including appendices). oig-files.ssa.gov · original document
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