Derived table · PPP, COVID EIDL, unemployment insurance and the employee retention credit
Pandemic fraud estimates: what each published figure measures
Published estimates of pandemic relief fraud, improper payment and recovery, each with the issuer's own term for what it counts, the amount exactly as printed, the denominator and period the issuer gives, and the caveats the issuer itself states: 22 estimates from 12 documents by 8 issuers.
Rows22
Columns31
Files1 · 30 KB
FormatCSV
Source vintageInspector general, GAO, Treasury, Justice Department, SBA and congressional documents dated 2022-09-21 to 2026-09-14
Published byPandemic Darlings
Acquired / built2026-09-21
Built from11 more, listed under Related records
SHA-256dd142562c796a9ce0f34dc49d21d83fb048875b5487b028298e63646143b5df5
RightsCompiled from U.S. government and congressional public records; attribution to Pandemic Darlings requested for the compiled table.
Dictionarydictionary.csv
Download CSV 30 KBData dictionary CSVAt a glancePreviewColumnsHow to useBefore you quote a row
At a glance
Measured on the served file on 2026-09-21; the checksum above is the file these numbers describe.
Rows22
Columns31
Key
estimate_id · uniqueCells left blank87 of 682 · 12.8%
Documents12
Issuers8
Definition classes15
Rows with a range rather than a single figure1
What it is
One row is one published estimate, as one document prints it. The row records what the issuer says it is estimating, a definition class derived from the documents themselves, the amount or range exactly as printed with its qualifier, the denominator and share the issuer gives, the period, the method in the issuer's words, and the caveats the issuer states.
Each row carries the path of the archive's copy of the document, that copy's SHA-256 recomputed at build time, the hash the archive recorded when the file was retrieved, the document's public address where one is established, a pin to the place in the document, and the verbatim span that carries the figure.
definition-classes.csv gives the one-line definition of each of the 15 classes. comparability.csv states, for every pair of estimates, whether the two documents measure the same thing, over the same programs, for the same period, and therefore whether the figures may be added.
compiled_label was split out of issuer_term_verbatim before publication. On six rows the delivered term joined two of the document's phrases, elided a quotation, or shortened the issuer's wording; those six strings are now in compiled_label, and issuer_term_verbatim carries the document's own contiguous wording from the same span.
How it was built
Amounts were copied from the held documents. amount_as_printed keeps the issuer's string, amount_qualifier keeps the hedge, and amount_usd_low and amount_usd_high hold the range in dollars, with amount_usd_high blank where the issuer printed a single figure.
Every verbatim span was re-found in the held document at packaging time: 22 of 22. Each of our copies was re-hashed at packaging time; 21 of the 22 rows sit on files whose hash the archive already records, and all 21 matched. The remaining row sits on the House Oversight ID.me staff release, which this archive holds as a text capture it made of the committees' published release rather than as a downloaded file, so no acquisition manifest records a hash for it; the hash in the row was computed from our copy and is proposed for the archive manifest.
The definition classes are derived from the documents, not imposed on them: each class is the set of rows whose issuers describe the same kind of quantity. The class of every row was re-checked against its own span at packaging time.
Preview
| estimate_id | issuer | issuer_type | document_title | document_id | document_date | programmes_covered | issuer_term_verbatim | compiled_label | definition_class | definition_class_meaning | amount_as_printed | amount_qualifier | amount_usd_low | amount_usd_high | unit | denominator_as_printed | share_as_printed | period_covered | method_in_issuer_words | caveats_issuer_states | held_path | sha256_remeasured | sha256_recorded |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SBAOIG2309-01 | SBA Office of Inspector General | inspector general | COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape | SBA OIG Report 23-09 | 2023-06-27 | COVID-19 EIDL; EIDL Targeted Advances; EIDL Supplemental Ta… | potentially fraudulent | POTENTIALLY-FRAUDULENT-DISBURSED | Dollars an oversight body's data analytics flagged as beari… | more than $200 billion | more than | 200000000000 | USD disbursed | approximately $1.2 trillion of COVID-19 Economic Injury Dis… | approximately 17 percent of disbursed COVID-19 EIDLs and PP… | the course of the Coronavirus Disease 2019 (COVID-19) pande… | Using OIG's investigative casework, prior OIG reporting, ad… | Loans identified as potentially fraudulent as part of our r… | held in the archive… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | ||
| SBAOIG2309-02 | SBA Office of Inspector General | inspector general | COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape | SBA OIG Report 23-09 | 2023-06-27 | COVID-19 EIDL (including advances) | potential fraud | POTENTIALLY-FRAUDULENT-DISBURSED | Dollars an oversight body's data analytics flagged as beari… | more than $136 billion | more than | 136000000000 | USD disbursed | total disbursed funds (COVID-19 EIDL programme) | 33 percent | the course of the COVID-19 pandemic | OIG analysis of COVID-19 EIDL data against 11 fraud indicat… | Fraud indicators are not proof of fraud; the report states … | held in the archive… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | ||
| SBAOIG2309-03 | SBA Office of Inspector General | inspector general | COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape | SBA OIG Report 23-09 | 2023-06-27 | PPP | potential fraud | POTENTIALLY-FRAUDULENT-DISBURSED | Dollars an oversight body's data analytics flagged as beari… | $64 billion | 64000000000 | USD disbursed | total disbursed funds (PPP) | 8 percent | the course of the COVID-19 pandemic | OIG analysis of PPP data against fraud indicators | Fraud indicators are not proof of fraud; the report states … | held in the archive… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | |||
| SBAOIG2309-04 | SBA Office of Inspector General | inspector general | COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape | SBA OIG Report 23-09 | 2023-06-27 | COVID-19 EIDL; PPP | seized or returned to SBA | SEIZED-RETURNED | Dollars seized, forfeited, voluntarily turned over or retur… | nearly $30 billion | nearly | 30000000000 | USD seized or returned | as at the report date | OIG collaboration with SBA, the U.S. Secret Service, other … | Due to the informal, ad hoc nature of SBA's tracking, the f… | held in the archive… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c… | ||||
| SBA2023-01 | U.S. Small Business Administration | disbursing agency | Protecting the Integrity of the Pandemic Relief Programs | SBA white paper, June 2023 | 2023-06 | PPP; COVID-EIDL; RRF; SVOG (SBA's four largest pandemic rel… | fraud to date | likely obtained fraudulently / likely fraud | LIKELY-FRAUD-REFERRED | Dollars a completed human-led agency review judged probably… | $36 billion | estimated | 36000000000 | USD disbursed | $1.2 trillion of SBA pandemic aid | applications originating 2020-2022 (86% in the first nine m… | automated screening that flagged over $400 billion, then ad… | the federal government has not developed an accepted method… | SBA agency report — Protecting the Integrity of Pandemic Relief Programs (2023)… | f3786ed8276d55554477f34c9e9685a3665f602a404639e729628a4be7c… | |||
| SBA2023-02 | U.S. Small Business Administration | disbursing agency | Protecting the Integrity of the Pandemic Relief Programs | SBA white paper, June 2023 | 2023-06 | PPP; COVID-EIDL; RRF; SVOG | data anomalies | data anomalies / fraud indicators requiring further review | FLAGGED-FOR-REVIEW | Dollars an automated screen or hold code marked as anomalou… | over $400 billion | over | 400000000000 | USD in applications, loans, grants and awards | $1.2 trillion of SBA pandemic aid | applications 2020-2022 | automated screening process | further investigation of the flagged loans - including over… | SBA agency report — Protecting the Integrity of Pandemic Relief Programs (2023)… | f3786ed8276d55554477f34c9e9685a3665f602a404639e729628a4be7c… | |||
| SBA2023-03 | U.S. Small Business Administration | disbursing agency | Protecting the Integrity of the Pandemic Relief Programs | SBA white paper, June 2023 | 2023-06 | PPP; COVID-EIDL; RRF; SVOG | four largest pandemic relief programs | disbursed (programme size) | PROGRAM-SIZE | The amount a programme disbursed; a denominator, not an est… | the Paycheck Protection Program (PPP) ($792 billion), COVID… | USD delivered | an unprecedented $5 trillion in federal emergency spending | one-fourth | pandemic relief programmes | agency statement of programme size | SBA agency report — Protecting the Integrity of Pandemic Relief Programs (2023)… | f3786ed8276d55554477f34c9e9685a3665f602a404639e729628a4be7c… | |||||
| SBA2023-04 | U.S. Small Business Administration | disbursing agency | Protecting the Integrity of the Pandemic Relief Programs | SBA white paper, June 2023 | 2023-06 | PPP; COVID-EIDL; RRF; SVOG | total recoveries | SEIZED-RETURNED | Dollars seized, forfeited, voluntarily turned over or retur… | $30 billion | total | 30000000000 | USD recovered | to May 2023 | USSS seizures, DOJ-led indictments and convictions, returns… | Many of these funds were collected through voluntary turnov… | SBA agency report — Protecting the Integrity of Pandemic Relief Programs (2023)… | f3786ed8276d55554477f34c9e9685a3665f602a404639e729628a4be7c… | |||||
| GAO106696-01 | U.S. Government Accountability Office | audit agency | Unemployment Insurance: Estimated Amount of Fraud during Pa… | GAO-23-106696 | 2023-09 | ALL unemployment insurance programmes (regular UI and the p… | fraud | FRAUD-ESTIMATE-STATISTICAL | An audit agency's estimate of the actual amount of fraud, p… | between $100 billion and $135 billion | likely between | 100000000000 | 135000000000 | USD of UI benefits | the total amount of UI benefits paid during the pandemic | about 11 percent and about 15 percent, respectively, of the… | April 2020 through May 2023 | Based on statistical sampling and imputation techniques | The full extent of UI fraud during the pandemic will likely… | GAO 23 106696… | d27a191447769930433b1374c788ab8e90361fabc96ea62b34f583aaebc… | d27a191447769930433b1374c788ab8e90361fabc96ea62b34f583aaebc… | |
| DOLOIG191-01 | U.S. Department of Labor Office of Inspector General | inspector general | The Greatest Theft of American Tax Dollars: Unchecked Unemp… | DOL-OIG 19-23-003 | 2023-02-08 | pandemic UI programmes (ETA's rate covers PEUC and FPUC; ap… | improper payments | IMPROPER-PAYMENTS | Payments that should not have been made or were made in the… | at least $191 billion | at least / could have been | 191000000000 | USD of UI payments | the approximate $888 billion in pandemic UI expenditures | 21.52 percent (ETA's estimated improper payment rate) | benefit weeks during the UI pandemic period | Applying the estimated 21.52 percent improper payment rate … | ETA's reported improper payment rate estimate of 21.52 perc… | Dol OIG 19 23 003… | 0b256f5aae55f06d4b84a829e6d38eabdefb9e0d0933086d9a1c23dbf00… | 0b256f5aae55f06d4b84a829e6d38eabdefb9e0d0933086d9a1c23dbf00… |
First 10 of 22 rows · first 24 of 31 columns (the download has all of them) · cells longer than 60 characters are cut with an ellipsis · shown as text, unformatted.
Files
| File | Rows | Size | SHA-256 | |
|---|---|---|---|---|
pandemic-fraud-estimates.csvCSV, UTF-8, header row | 22 | 30,741 B | dd142562c796a9ce0f34dc49d21d83fb048875b5487b028298e63646143b5df5 | Download ↓ |
dictionary.csvdata dictionary, 31 rows | 31 | — | built with this page | Download ↓ |
definition-classes.csvone-line definition of each class | 15 | 2,505 B | f35f845949829652c9fc90dc48f67f5273be1a4428dba527d097905bf20dc6d6 | Download ↓ |
comparability.csvevery pair of estimates: same class, same programs, same period, additive? | 210 | 60,612 B | 346c466bb83901f19cc711718f7e0cc9faf88aedf53dd21445a89f41f12c0f0b | Download ↓ |
README.mdwhat the table is, what it does not measure, and how every number can be checked | — | 5,817 B | 710c955c338c7b711123dc69476ac3fe1d9c3abf9aba4b22897a08a329b35ece | Download ↓ |
SHA256SUMSSHA-256 of every other file on this page | — | 0 B | — | Download ↓ |
Columns
| Column | Type, measured | Blank | Distinct | Range / values | Meaning |
|---|---|---|---|---|---|
estimate_id | text | 0 | 22 | Stable identifier for the published estimate: issuer-document token plus a sequence number. | |
issuer | text | 0 | 8 | SBA Office of Inspector General (5), U.S. Small Business Administration (5), U.S. Department of Labor Office of Inspector General (5), U.S. Government Accountability Office (2), Treasury Inspector General for Tax Administration (2), U.S. Department of Justice, COVID-19 Fraud Enforcement Task Force (1) | The body that published the estimate, as it names itself. |
issuer_type | text | 0 | 6 | inspector general (12), disbursing agency (5), audit agency (2), prosecutor (1), investigator (1), vendor claim reported by congressional committees (1) | What kind of body the issuer is: inspector general, audit agency, disbursing agency, prosecutor, investigator, or a claim reported by a committee. |
document_title | text | 0 | 12 | COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape (4), Protecting the Integrity of the Pandemic Relief Programs (4), The Greatest Theft of American Tax Dollars: Unchecked Unemployment Fraud (2), COVID-19: ETA and States Did Not Protect Pandemic-Related UI Funds from Improper Payments (2), Management Took Actions to Address Erroneous Employee Retention Credit Claims; However, Some Questionable Claims Still Need to Be Addressed (2), COVID-19 Fraud Enforcement Task Force 2024 Report (2) | Title of the document the estimate appears in. |
document_id | text | 0 | 12 | SBA OIG Report 23-09 (4), SBA white paper, June 2023 (4), DOL-OIG 19-23-003 (2), DOL-OIG 19-22-006 (2), TIGTA 2024-40-068 (2), CFETF 2024 Report (2) | Report or release number the issuer uses. |
document_date | text | 0 | 12 | 2023-06-27 (4), 2023-06 (4), 2023-02-08 (2), 2022-09-30 (2), 2024-09-30 (2), 2024-04-09 (2) | Date the document carries. |
programmes_covered | text | 0 | 17 | The programs the estimate covers, as the document states them. | |
issuer_term_verbatim | text | 0 | 20 | The issuer's own words for what is being estimated (for example 'potentially fraudulent', 'improper payments', 'likely fraud'). | |
compiled_label | text | 16 | 6 | likely obtained fraudulently / likely fraud (1), data anomalies / fraud indicators requiring further review (1), disbursed (programme size) (1), potentially fraudulent payments in high-risk areas (1), suspected fraudulent pandemic-era activity (1), lost ... to fraudulent claims (1) | A label this archive compiled when the issuer's own wording was joined, elided or shortened to fit the column. Blank on the rows where the issuer's own term stands unaltered. |
definition_class | text | 0 | 15 | Controlled class derived from the sources in this table; see definition-classes.csv. | |
definition_class_meaning | text | 0 | 15 | One-line definition of that class. | |
amount_as_printed | text | 0 | 22 | The amount or range exactly as the document prints it. | |
amount_qualifier | text | 2 | 12 | more than (4), estimated (4), over (2), at least / could have been (2), nearly (1), total (1) | The qualifier the document attaches to the amount (more than, at least, approximately, estimated). |
amount_usd_low | integer | 2 | 17 | 1.8B – 400B · sum $1888.2B · median $37.5B | Machine-readable lower bound in US dollars; blank where the estimate is a count and not a sum. |
amount_usd_high | integer | 21 | 1 | 135B – 135B · sum $135B · median $135B | Machine-readable upper bound where the document prints a range; blank otherwise. |
unit | text | 0 | 16 | What the amount counts. | |
denominator_as_printed | text | 3 | 17 | The denominator the issuer gives, verbatim; blank where the issuer gives none. | |
share_as_printed | text | 12 | 10 | approximately 17 percent of disbursed COVID-19 EIDLs and PPP funds (1), 33 percent (1), 8 percent (1), one-fourth (1), about 11 percent and about 15 percent, respectively, of the total amount of UI benefits paid during the pandemic (1), 21.52 percent (ETA's estimated improper payment rate) (1) | The share or rate the issuer prints, verbatim; blank where none is printed. |
period_covered | text | 0 | 19 | The period the estimate covers, as the document states it. | |
method_in_issuer_words | text | 0 | 21 | The method, in the issuer's words. | |
caveats_issuer_states | text | 1 | 20 | Caveats the issuer itself states about the figure. | |
held_path | text | 0 | 12 | SBA Oig Report 23 09 (4), SBA agency report — Protecting the Integrity of Pandemic Relief Programs (2023) (4), Dol OIG 19 23 003 (2), Performance Audit Report: COVID-19 — ETA and States Did Not Protect Pandemic-Related UI Funds from Improper Payments Including Fraud or from Payment Delays… (2), TIGTA Report 2024-40-068 — Erroneous ERC Claims: Management Actions Taken (2), DOJ 2024 Cfetf Report (2) | On-site route of the held copy where one exists, otherwise the file name. |
sha256_remeasured | text | 0 | 12 | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c65482 (4), f3786ed8276d55554477f34c9e9685a3665f602a404639e729628a4be7ced1ca (4), 0b256f5aae55f06d4b84a829e6d38eabdefb9e0d0933086d9a1c23dbf0064e52 (2), 81e38c30a733b8f8756ad459ac01365fb3987282e125d1fda47d5000e752037e (2), 71bd822c772135ec17ca26a691b5b80b6191e526201746ed2915a896fc1b0068 (2), 12f8df51e380feb9e203fcdfa9b2ed777cd2c130388e1348f8c845580d51d953 (2) | SHA-256 of our copy, recomputed when this table was built. |
sha256_recorded | text | 5 | 10 | 7041d37d1d71dd84a3bf78a8081810e8d00356c379a76c71b7486c9899c65482 (4), 0b256f5aae55f06d4b84a829e6d38eabdefb9e0d0933086d9a1c23dbf0064e52 (2), 81e38c30a733b8f8756ad459ac01365fb3987282e125d1fda47d5000e752037e (2), 71bd822c772135ec17ca26a691b5b80b6191e526201746ed2915a896fc1b0068 (2), 12f8df51e380feb9e203fcdfa9b2ed777cd2c130388e1348f8c845580d51d953 (2), d27a191447769930433b1374c788ab8e90361fabc96ea62b34f583aaebccecb9 (1) | SHA-256 recorded for that file in the archive manifest or meta file when it was retrieved; blank where none is recorded. |
sha256_match | text | 0 | 3 | yes (17), no recorded sha in archive manifest or meta (4), yes: the archive's pandemic relief integrated ledger (2026-06-22) records this sha256 (1) | Whether the recomputed and recorded hashes agree. |
public_url | text | 4 | 11 | https://www.oversight.gov/sites/default/files/documents/reports/2023-06/SBA-OIG-Report-23-09.pdf (4), https://www.oversight.gov/sites/default/files/documents/reports/2024-04/19-23-003-03-315.pdf (2), https://www.oversight.gov/sites/default/files/documents/reports/2022-10/19-22-006-03-315-COVID-19-ETA-and-States-Did-Not-Protect-Pandemic-Related-UI-Funds-Improper-Payments.pdf (2), https://www.oversight.gov/sites/default/files/documents/reports/2024-10/2024400068fr.pdf (2), https://www.justice.gov/coronavirus/media/1347161/dl (2), https://www.gao.gov/assets/gao-23-106696.pdf (1) | Public address of the document; blank where no public address is established in the archive's metadata. |
pin | text | 0 | 15 | Where in the document the span sits. | |
verbatim_span | text | 0 | 22 | The span from the held document that carries the figure, quoted exactly. | |
verbatim_span_2 | text | 21 | 1 | This represents about 11 percent and about 15 percent, respectively, of the total amount of UI benefits paid during the pandemic. (1) | A second verbatim span where the share or rate is printed separately; blank where not needed. |
verified_how | text | 0 | 5 | held PDF text (pdftotext -layout) (17), held PDF text (pdftotext raw) (2), held PDF text (pdftotext -layout) ; second span: held PDF text (pdftotext -layout) (1), text extraction of the held capture (sba-news-release-26-91-870000-borrower-suspensions-2026-09-14.txt) (1), held file (1) | How the span was checked against our copy. |
measured_at_utc | text | 0 | 1 | 2026-09-19T21:59Z (22) | When the row's measurements were taken. |
Type, blanks, distinct count, and range are measured on the file. The same table as CSV: dictionary.csv.
How to use it
Specific to this file, measured on it: keys, joins, encodings, units, and the values that will trip a naive count.
- Do not add the figures. The rows sit in 15 definition classes. Adding across classes produces a number no issuer published.
comparability.csvstates for each of its 210 pairs why: exactly one pair is additive without double counting. - Improper payment is not fraud.
IMPROPER-PAYMENTSrows count payments that should not have been made; the category includes error and ineligibility as well as fraud, and the issuers say so. - Charged loss is not convicted loss.
CHARGED-LOSSis loss alleged in filed charges. - Each row is fixed to its document's date. Later documents may restate a figure; where a document restates its own earlier figure, both rows are present.
- Two columns, because a quotation and a label are not the same thing.
issuer_term_verbatimholds only what the document prints for what its figure counts; it was re-found in our copy, contiguous, for all 22 rows. Where this table needed a shorter or joined phrase to sit in a column, that phrase is incompiled_labelinstead, filled on six rows and written by this archive, not by the issuer. public_urlis blank where no public address is established. Four rows rest on an SBA white paper the archive holds with no recorded source address; the gap is listed, not guessed.- CRLF line endings, UTF-8, 30,741 bytes.
Before you quote a row
- No row is a finding, and the table draws no conclusion about a true amount. It records what each document says and how it says it.
- It is not complete. It covers only estimates for which this archive holds the document. One estimate was dropped for failing verification, and the reason is recorded in the build receipt.
- It is not the site's calculator. Build your own fraud number takes its own inputs; this table is the record of what the published estimates measure, and the two are not wired together.
- The actions taken against borrowers are a separate table. Pandemic enforcement and suspension events, 2026 records suspensions, referrals and enforcement summaries; it holds no estimate of the size of fraud.
Cite this table. Name the source vintage; the numbers depend on it.
Pandemic fraud estimates: what each published figure measures, compiled from documents dated 2022-09-21 to 2026-09-14. Pandemic Darlings, https://pandemicdarlings.com/data/pandemic-fraud-estimates/, SHA-256 dd142562c796a9ce0f34dc49d21d83fb048875b5487b028298e63646143b5df5