Twenty-seven months in the Bronx

The New York Housing Conference's 2024 case study laid the bottleneck out in unsparing detail. A new affordable building, financed through the city's Housing New York program, completed construction. The lottery — administered through NYC Housing Connect, with strict income-eligibility documentation requirements — closed roughly nine months after that. The building stood, fully built, fully waitlisted, and entirely vacant for another eighteen months while HPD worked through its applicant-vetting backlog. The total: twenty-seven months between physical completion and the first tenant signing a lease.

This is the gap that doesn't show up in the production totals. The dataset HPD publishes counts the units the moment the financing closes. By that count, the city is producing affordable housing at the headline rate it advertises. Across Housing New York and its successor programs, the cumulative production total looks like this:

0 units across 0 buildings, weighted heavily toward the Bronx and Brooklyn — exactly where need-by-AMI is concentrated. On paper, this is a triumph.

The composition tells you the second story:

Roughly half of the cumulative total comes from preservation — refinancing existing affordable units to extend their affordability covenants for another 30 years. This is real and important; it prevents existing affordable units from converting to market rate when their original tax abatements expire. But it doesn't add net supply. The other half is new construction, and that's where the lease-up bottleneck lives.

The bureaucratic chokepoints

HPD's lottery system requires applicants to document everything: income, employment history, household size, prior-residence stability, income-source verification across an extended look-back window. For households at the lowest AMI bands — extremely low and very low income, the bands the city most aggressively claims to serve — these documentation requirements are catastrophically out of phase with how those households actually live. Pay stubs that don't exist because the work was off-the-books. Bank statements that don't exist because the household isn't banked. Lease histories that don't exist because the household has been moving every six months between informal sublets.

The result is that HPD has to send back applications for resubmission, sometimes multiple times. Each round adds weeks. Each round risks losing the applicant entirely — to a lease expiration that can't wait another month, to a relocation, to discouragement. The applicant-vetting bottleneck is structurally biased against the populations the program was designed to help.

The reform asks

Housing advocates have used this exact data to lobby for three reforms:

• Audit-based compliance instead of step-by-step monitoring — review a sample of applications rigorously rather than every application redundantly.
• Virtual third-party inspections — outsource physical compliance checks to certified inspectors so HPD's caseload doesn't choke on inspection scheduling.
• Algorithmic application screening — automated pre-validation of common documentation patterns so applicants get faster feedback on what's missing.

None of these reforms speed up the building. They speed up the occupation of buildings that are already finished. The 27 months in the Bronx don't represent a city that can't build affordable housing. They represent a city that builds it and then takes two years to put anyone in it. That's a different problem, and it requires a different fix.