Auditors: $495M homebuyer aid lacked clear math in 67 awards

On September 13, 2026, the primary post-María housing document is the HUD OIG Audit Report 2024-FW-1003, dated March 19, 2024: the Puerto Rico Housing Finance Authority (PRHFA) Homebuyer Assistance Program (HBA) positively impacted participants, but may not have distributed disaster recovery funds equitably. HUD awarded PRDOH —and PRDOH designated PRHFA as subrecipient— a program budget of $495 million in CDBG-DR during the reviewed period. Budget ≠ allocated ≠ disbursed.

Table 1 of the report (DRGR data as of July 13, 2023) separates: HBA low-moderate income Budget $344,721,869, Allocated $339,959,407, Disbursed $138,970,260; HBA urgent need Budget $150,278,131, Allocated $149,080,812, Disbursed $99,242,170; Total Budget $495,000,000, Allocated $489,040,219, Disbursed $238,212,430. The budget started at $295 million in July 2020 and rose to $495 million in February 2023; the primary notes that in October 2023—after the fieldwork—PRDOH amended the action plan to bring HBA to $695 million, a figure outside the field audit period.

Operationally, HBA offers downpayment and closing costs: up to $45,000 for LMI households or urgent need, $55,000 if there is a Critical Recovery Workforce member, plus $5,000 if they buy in a designated urban center. A five-year lien is recorded as a second mortgage. PRHFA started the program in July 2020. The report found that the program was on track in the subrecipient agreement outcomes: more than 434 applications monthly on average and more than $5.2 million disbursed monthly (targets 400 and $5 million); 63 MOUs with lenders (target 50). As of July 21, 2023, the public dashboard reported more than 6,700 families assisted.

The finding of equity is harsh: in the 67 closed files sampled (universe 4,170 closures Jul 2020–Nov 2022), PRHFA could not demonstrate how it calculated the financial assistance award in any case. The underwriters did not perform the analysis required by the guidelines; the calculation was delegated to the lending institutions. Example of the primary (Table 2): two CRW participants with similar profiles —incomes $42,556 vs $37,432; prices $130,000 vs $112,000— received $55,000 versus $8,500. PRHFA did not explain the difference. Without a verifiable method, there is no evidence of equitable distribution.

Separately, PRDOH could not reliably measure the impact of HBA on the island's home ownership rate because it used PRCS/ACS and OCIF data, influenced by factors unrelated to the program. The OIG recommends updating guidelines and the action plan, and establishing a verifiable method of awards with training for lenders. The primary does not report theft nor quantify questioned costs of inequitable awards; it documents the absence of calculation and risk of inequity.

Editorial use: cite 2024-FW-1003, $495 million budget, $238.2 million disbursed, 6,700 families, sample 67/67 without calculation documentation, and the contrast $55,000 vs $8,500. Do not confuse with 2024-FW-1001 (Ad Hoc fraud) or with CEWRI. Cousin: none. Topic: economy.

The primary also documents satisfaction of interviewed participants and barriers (agents, sellers, lenders). That does not cancel the equity finding: a program can be popular and, at the same time, opaque in the award formula. The statistical projection on perceived discrimination appears in the appendices; this note focuses on the 67/67 documentation finding and the CRW example.

PRHFA alleged that SSP calculates automatically according to guidelines; the OIG requested the system parameters and did not receive them in a way that would allow recreating the awards. Without auditable parameters, the software's “black box” does not replace the recordkeeping required by 24 CFR 570.490 and the subrecipient agreement.

Intersection with the lot: HBA is CDBG-DR for unmet housing needs, not the electric grant or HMGP. Maintaining that boundary avoids adding $495 million to the $1.932 billion electric or to COR3's PA. Budget $495M / disbursed $238.2M is the correct pair for the DRGR report cut.