WEDNESDAY, MAY 22, 1974.

Mr. BOLAND. The committee will come to order. We will continue with the general questions. I think Mr. Tiernan was questioning when we closed yesterday.

Mr. TIERNAN. Thank you.

GAO REPORT ON HOUSING ALLOWANCES

Mr. Secretary, we had asked some questions and the chairman had asked you about the relative cost. I guess this table was supplied yesterday to the committee. I am not quite clear as to all the different footnotes and how they all tie in. Was this table prepared since the study made by the Controller General of March 28, 1974?

Secretary LYNN. Yes, sir.

Mr. TIERNAN. Now, in that report they seem to indicate that the estimated costs that have been placed on it by you and the Department are not-well, in their opinion they think it depends on how far you are going to go on the income allowances and others.

Secretary LYNN. Pardon me, I want to be sure we are talking about the same report. Is this the report with respect to the direct cash assistance, housing allowances?

Mr. TIERNAN. Observations on housing allowances and the experimental housing allowance program.

Secretary LYNN. Our tables are comparisons of section 23 with 235 and 236 and public housing and did not involve the direct cash assistance program.

Mr. TIERNAN. So then, do you have anything that indicates as in that report that estimated cost might not be that much different than what you would have had if you continued the present subsidized housing program?

Secretary LYNN. We disagree very strongly with the analysis of that report as to computations and the comparisons. We made our disagreements known at the time the report was being prepared. Some of our suggestions found their way into the report, but a number of very important ones did not.

HUD RESPONSE TO GAO REPORT

We would be happy to furnish this committee with a formal response. A response was written for me because I wanted to be satisfied that our position on this was accurate. We would be happy to

( 101 )

furnish the committee with a response to that report insofar as it does make cost analyses and, frankly, insofar as it covers some other points, like our experimental areas not being the appropriate ones to use for purposes of the experiment.

[The response follows:]

HUD RESPONSE TO GAO REPORT: "OBSERVATIONS ON HOUSING ALLOWANCE AND THE EXPERIMENTAL HOUSING ALLOWANCE PROGRAM"

Fro completeness, HUD's response to the GAO report: "Observations on Housing Allowances and the Experimental Housing Allowance Program" is presented below in two parts. The first is HUD's letter to the GAO on the draft submitted to HUD for their review and comment. Agency comments are usually published in the final version of the report. Portions of the HUD response which were deleted by the GAO in the GAO report to the Congress are underscored with a dashed line. The second part is a HUD letter to the GAO on the published version of the GAO report to the Congress.

This letter is in response to the GAO Report to the Congress entitled: Observations on Housing Allowances and the Experimental Housing Allowance Program (B-171630).

Earlier, we had formally provided our comments on a draft of this report (my letter to you dated January 18, 1974). Since it was expected that the letter would be printed (in its entirety) in the final version of the GAO Report, I did not anticipate the necessity of further comment. However, because fully relevant and extremely important portions of my earlier letter were omitted in the report, it is imperative that we make the comments contained herein.

So as not to be repetitive, our comments this time will address those items which relate either to the deletions from my earlier letter or to pertinent differences between the final report and the draft we reviewed.

Page 2, Impact of Housing Allowance. This section discusses the possibility that housing allowances will result in higher rents for not only those households who are eligible for the program but also for higher income groups who are not eligible.

For both of these groups the GAO Report concludes that significant rent increases will follow the implementation of housing allowances. Results from two housing market models (one prepared by a GAO consultant and the other by the Urban Institute) suggested this conclusion.

As was noted in our January 18 letter, the GAO commissioned model
was not adequately constructed to address this question. It is
unfortunate that the description of this model was not included in
the final report so that readers would be able to evaluate the model
themselves. Because this model did not permit landlords or builders
to expand the supply of standard housing (which is a very important
impact of the housing allowance program) the outputs of this model
are meaningless in addressing the possible results of this program.
The preliminary results of the Urban Institute's model which were
quoted are just that, preliminary. When these results were generated,
as was noted by the author, the model was not complete and had not
undergone any tests for validity. The Urban Institute, as well as
the National Bureau of Economic Research, are still under contract to

develop the models further. The latter is developing another simulation model of urban housing markets. When both are complete, we expect to test their results against actual data, and if these tests are satisfactory, we shall then utilize these models to project the market impact of various types of housing allowance arrangements.

Finally, it is important to note that while a short-run rise in the rent of program-defined standard housing units may be anticipated, it is just such a rise which could stimulate landlords of substandard units to upgrade their units to qualify for these increased rents, stimulate new construction and provide the means of better preserving the existing housing stock. Therefore, it is the long run (four to five years) change in rents which are most important. It is possible that there will be little or no change in rents (induced by the program) over the long run that are not associated with commensurate increases in the quality of housing and services obtained.

Page 4, Comparison of Housing Allowance Costs. The GAO claims in this section that a housing allowance program would be more costly per unit than the production subsidy program.

Two extremely important points of our January 18 reply to this section were deleted from our reply as printed in Appendix 1 of the GAO Report. Not only were these points deleted, but no reference was made to them in the final report, and the final report was not changed in any noticeable way to acknowledge these points. This is in direct conflict with the GAO note at the end of Appendix 1.

The first point which was deleted reads as follows:

-

"First -- as the GAO notes the comparison includes only
costs to the Federal Government. Costs to local governments
attributable to foregone tax revenues and the cost associated
with the acquiring, holding and disposing of mortgages in
default have been excluded. In any meaningful consideration
of program alternatives, however, full costs should be
considered. This was the approach followed by HUD in its
most recent National Housing Policy Review. It yielded
higher yearly unit cost estimates for public housing,
235 and 236, than did the GAO approach. (For public housing:
$1,650 vs. $968; for Section 236: $1,100 vs. $874 for
Section 235: $1,097 vs. $725.)"

As can be noted, it is our contention that the GAO procedure of computing the costs of other programs was not done properly. Thus, these comparisons do not provide "a reasonable indicator of comparative cost of housing allowances" as is claimed on Page 4.

The second point which was excluded is:

"Third, for a comparison of costs to be meaningful, the
comparison should be controlled for differing income
levels of the households served. Average income levels
served by the Section 235 and 236 programs considerably
exceed those of the public housing programs and those of
estimated housing allowance recipients, and as a result,
the subsidies under 235 and 236 would have to be raised
greatly if they were to be used to serve the same income
distribution that is served by the allowance program.
The amount of the increase depends on the future design
of the housing allowance program, but the GAO analysis
sheds little light on this issue. As a result, we believe
that it significantly understates the cost of using existing
programs to serve the same lower income groups that would
be served by an allowance program."

The thrust of this paragraph is that a direct comparison of the Section 235 and 236 programs to a housing allowance program is artificial because Section 235 and 236 serves only a fraction of the population that would be reached by a housing allowance program. As such, the comparison made of these programs is analogous to comparing apples and oranges. As the GAO notes, the validity of the cost comparisons depends on assumptions. Different assumptions would produce different conclusions and we seriously question the validity of the ones used. In fact, we believe they are in serious error.

Page 4, Need to Expand EHAP. In this section the GAO contends that the sites selected for EHAP are not fully representative of urban housing markets in that they will not provide information on markets with either low-quality housing or low-vacancy rates. Thus, the GAO believes EHAP should be expanded to include some sites with low housing quality and some with low vacancy rates.

In our January 18 letter we raised a number of points appositive to this conclusion, some of which the GAO final report responded to and some of which it did not. The GAO report's conclusions do not seem to us to be well supported. For example:

-- Of the 10 operating sites for which we have data, six were below the average 1970 U.S. vacancy rate, one was at the average and