RAD at Scale: What the Evaluation Evidence Shows About Rent Burden, Exit Rates, and the Conversion Track That Matters

The Rental Assistance Demonstration (RAD) is usually described as a financing tool: public housing authorities convert aging public housing units into Section 8 contracts, bring in private capital, and repair buildings that the public housing capital fund could not sustain. That description is accurate as far as it goes. But for practitioners who administer assisted housing, the more useful question is not whether RAD repairs buildings. It is whether the conversion changes what tenants pay, how they exit, and which households the program actually reaches.

The evaluation evidence assembled so far answers some of those questions and leaves others open. The answers are less dramatic than the program’s critics and defenders tend to suggest, and the open questions are more consequential than the ones that have been settled.

What RAD converts, and into what

RAD enables public housing authorities to convert public housing units into Section 8 housing. The stated goals are to improve physical condition and financial viability while giving tenants the opportunity to choose other affordable housing options (Econometrica, “HUD’s RAD Program”).

Conversion runs through two tracks. Units can become project-based rental assistance (PBRA) properties or project-based voucher (PBV) properties. The distinction is not cosmetic. PBRA and PBV carry different contracts, different oversight arrangements, and, as the evaluation evidence shows, different tenant exit patterns. The RAD Choice Mobility evaluation covered the census of public housing projects that converted to either PBRA or PBV contracts, which means the track-level comparisons rest on the full converted population rather than a sample (Econometrica).

Rent burden: stable shares, not a spike

The most direct question about RAD is whether conversion raises what tenants pay. The Urban Institute’s evaluation, prepared for HUD’s Office of Policy Development and Research, examined the share of income RAD households spent on rent and utilities. It found limited variation in that share and little variation across demographic characteristics of household heads (Urban Institute, “Evaluation of the Rental Assistance Demonstration”).

That finding matters because the rent-burden objection to RAD is often framed as a prediction: conversion will shift costs onto tenants. The evaluation does not support that prediction at the population level. It also does not prove that no household experienced a rent increase. A stable distribution can conceal movement within it, and the evaluation’s own framing is about limited variation, not zero change.

The evaluation also found that incomes increased for households in both PBRA and PBV properties, and that those income changes were closely aligned with changes in rent and utilities (Urban Institute). In practical terms, rents moved with incomes rather than against them. That is consistent with the structure of assisted rent-setting, where tenant contributions are tied to income, but it is worth stating plainly because it undercuts the simpler narrative that conversion is a rent hike in disguise.

The study team’s overall conclusion was that it did not find evidence that RAD conversions led to substantial or rapid changes in the economic circumstances of RAD households (Urban Institute). “Substantial or rapid” is doing real work in that sentence. The finding rules out a shock. It does not rule out slow drift, and it does not address households that left before or during conversion.

Flat rents and the exit-rate gap

The evaluation’s most practitioner-relevant findings concern households that previously paid flat rents, because those households are the ones most exposed to a change in rent-setting method. The results differ by conversion track.

Households in PBV properties exited public housing at rates similar to voucher holders overall. Households in PBRA properties exited at comparatively lower rates (Urban Institute). The evaluation also found that exit rates were low for heads of household who were older or had a disability, and that use of the Choice Mobility option to exit the RAD property with a tenant-based voucher was limited for households previously paying flat rents (Urban Institute).

These findings should be read carefully. An exit is not the same as an eviction. The evaluation reports exit rates, not eviction filings, and the retrieved evidence does not establish what share of exits were voluntary, what share followed lease termination, or how exit rates in RAD properties compare with exit rates in non-converted public housing. The PBRA/PBV gap is a pattern in the data, not an explanation of why the pattern exists.

The question the evidence actually raises

Because rent-to-income shares were stable and incomes tracked rents, the rent-burden question is largely settled at the population level. The more useful question is distributive: which conversion track and which mobility pathway actually reach flat-rent and vulnerable households?

The evidence flags this question without answering it. The PBRA/PBV exit-rate gap and the limited uptake of Choice Mobility among flat-rent households suggest that the administrative design of the mobility option, not the conversion itself, is the lever worth auditing. That is a hypothesis to test, not a finding. Testing it would require data the retrieved evaluation summaries do not provide: the share of eligible flat-rent households who were informed of Choice Mobility, the timing and form of that notice, the share who attempted to use it and were denied or delayed, and the reasons households gave for not moving.

For practitioners, the implication is procedural rather than rhetorical. If Choice Mobility is the mechanism intended to give flat-rent households a path out, then the audit question is whether that mechanism is being administered in a way that reaches them. Low uptake among the households most likely to need the option is a signal worth investigating at the property level, not a conclusion about the program as a whole.

What the evidence does not establish

Several claims that circulate about RAD are not supported by the retrieved evaluation evidence and should be treated as open questions.

Eviction exposure. No retrieved source supports the claim that RAD conversions increased eviction exposure. The evaluation reports exit rates, which are a different measure. Establishing whether eviction filings differ between converted and non-converted properties would require filing-level data matched to conversion status, and that analysis is not in the retrieved material.

Ownership structure. The retrieved sources describe conversion into PBRA and PBV contracts but do not describe how ownership structures changed, including the roles of public housing authorities versus third-party owner-operators. Claims about ownership transformation at scale are not supported by the retrieved text.

Scale-wide rent-burden distribution. The evaluation found limited variation in rent-to-income shares. It does not provide a state-by-state or property-level distribution of rent burden after conversion, and no retrieved source supplies one.

Program scale and statutory mechanics. Specific conversion counts, statutory citations, program caps, and the text of HUD’s RAD notice were not available in the retrieved sources. Practitioners who need those details should consult HUD’s RAD program page and the applicable RAD notice directly rather than relying on secondary summaries.

What to watch as the evidence develops

The RAD Choice Mobility evaluation has produced four reports to date, with a fifth forthcoming. The released reports cover implementation of the Choice Mobility option, public housing authorities’ approaches to asset management under RAD, organizational changes at converting authorities, and a quantitative study of resident outcomes using HUD administrative data for the universe of RAD conversions (Econometrica).

The fifth report, not yet released, examines long-term preservation and financial viability by reviewing project financial statements and HUD data for RAD conversions, non-RAD PBRA projects, and a matched sample of non-RAD public housing projects (Econometrica). That report is the one most likely to speak to the durability question: whether converted properties remain affordable and financially sound over time, and how they compare with the public housing they replaced.

Until it is released, the defensible summary is narrow. RAD conversion, as evaluated, did not produce substantial or rapid changes in tenants’ economic circumstances. Rent-to-income shares were stable, incomes tracked rents, and exit patterns differed by conversion track and by household vulnerability. The eviction and ownership questions remain open, and the mobility option’s reach into flat-rent households is the administrative question the evidence most clearly raises.

FAQ

Did RAD conversions raise rents for tenants?

The Urban Institute evaluation found limited variation in the share of income RAD households spent on rent and utilities, and little variation across demographic characteristics of household heads. It did not find evidence of substantial or rapid changes in households’ economic circumstances. That is a population-level finding, not a guarantee for every household.

Do PBRA and PBV conversions produce different tenant outcomes?

The evaluation found that flat-rent households in PBV properties exited at rates similar to voucher holders overall, while exits for those in PBRA properties were comparatively lower. The evidence does not explain why the gap exists.

Is Choice Mobility widely used by flat-rent households?

The evaluation found that use of the Choice Mobility option to exit the RAD property with a tenant-based voucher was limited for households previously paying flat rents. The retrieved evidence does not establish why uptake was limited.

Did RAD conversions increase evictions?

The retrieved evidence does not support that claim. The evaluation reports exit rates, not eviction filings, and does not compare eviction outcomes between converted and non-converted properties. That question remains open.

Where can practitioners find the underlying program rules?

The retrieved sources do not include HUD’s RAD notice or program page. Practitioners should consult HUD’s RAD program materials and the applicable RAD notice directly for statutory authority, conversion mechanics, and tenant-rights requirements.