The thing parodies itself
If Section 8 is under continuing funding pressure (which it is) that is hurting communities and families nationwide (which it is), we are certainly motivated to find ways to improve it, with lower costs or better results. And if we think that means-testing housing assistance might create a poverty trap (certainly a reasonable hypothesis), and we have been running a Moving To Work (MTW) demonstration program for eight years (which we have), wouldn’t we want to know how it had performed?
Senator Blutarsky wants to know …
What has worked and what hasn’t?
On behalf of policymakers and taxpayers, HUD’s Office of Inspector General did:
We reviewed (link in PDF) the U.S. Department of Housing and Urban Development’s (HUD) design and implementation of the Public Housing/Section 8 Moving to Work Demonstration program. We wanted to know whether:
1. The program tested ways to provide and administer housing assistance that reduced costs, promoted self-sufficiency, and increased housing choices, and
2. HUD had the authority to approve housing authority requests to make tenants enter new contracts with time-limited housing assistance.
HUD struggled to balance flexibility and accountability in the design and implementation …
Alert Secret Service agents seized the data and wrestled it to the ground
… of the Public Housing/Section 8 Moving to Work Demonstration program and relied on an existing system to collect tenant information. The existing system could not accept tenant information …
… and was not adapted in time to support the interim evaluation, and as a result, HUD was not able to collect tenant information needed to measure interim program impacts on costs, family self-sufficiency, and housing choices as planned. (Pages 2-3)
Without the tenant information, HUD was not able to measure interim program impacts on costs, family self-sufficiency, and housing choices as planned. (Page 6)
“We’re extremely proud to report HUD failed utterly.”
The dog ate my data. How much data did the dog eat?
HUD entered Moving to Work Demonstration Agreements with 26 housing authorities (now 32+) between 1998 and 2004. These agreements (for examples, click here) were for 5 to 12 years.
What then did HUD do?
As a result of the delay, HUD had to redesign the evaluation methodology…. The redesigned methodology recognized that outcome information would not be available to evaluate program accomplishment and instead focused on the types of demonstration activities and experience of participating housing agencies.
Translation: Since they couldn’t track results, they documented approaches.
With the Multifamily Tenant Characteristics System to collect information, HUD was not able to provide the measurable and comparable results initially planned …. HUD’s evaluation could not cite:
1. Statistics showing [whether and to what extent] Moving To Work demonstration activities could be considered models for reducing costs and achieving greater cost-effectiveness, promoting resident employment and self-sufficiency, and increasing housing choices for low-income households; and
2. Comparative analyses intended to show the impact of program activities and importance of individual policy changes. (Page 7)
Translation: They ran many pilots but have no evidence to show which experiments worked and which did not.
Why, you may wonder, was HUD unable to collect needed information? Because Gulliver tied his shoelaces together:
As you can plainly see, Mr. Gulliver, under 5 CFR 1320.5(a) and 1320.3(c) ….
Under various paperwork reduction act provisions (specifically 5 CFR 1320.5(as) and 1320.3(c)), requests for “collection of information” may not be sent out unless OMB approval is given, and a “collection of information” is defined as “reporting, recordkeeping, or disclosure requirements imposed on ten or more persons …” Guess what?
HUD selected Moving To Work demonstration housing authorities under two requests for applications, spaced about 4 years apart. [The first netted seven, but after the second], 14 were required to submit.
Accordingly, when HUD decided to make a second selection, it did not have a mechanism to monitor [Moving To Work performance]. (Page 9)
What does HUD say about all this?
“The missing data isn’t important.”
We do not believe that the absence of data had as big an impact on evaluation outcomes as indicated by the finding…. (Page 15)
“Data, data, we don’ need no steenkin’ data!”
“More data won’t make any difference — we know the answer without bothering to experiment.”
An increase in the amount of tenant characteristic data is not going to really change the outcome or our ability to conclude which flexibilities should be (or could be) replicated. … Given the design constraints, the Department will still not be able to conclude with any greater certainty than in its prior evaluation, which flexibilities were “successful” and could or should be replicated. (Pages 15-16)
The French have a phrase, “l’affaire s’arrange,” the thing arranges itself. Perhaps HUD should have a phrase, the thing parodies itself.
