Responses to OMB Comments

memorandum

Social and Economic Policy Division

Date	July 5, 2011
To	Regina Gray, HUD
From	Carissa Climaco, Jill Khadduri (Abt Associates Inc.)
Subject	Assessment of the LIHTC Program After 15 Years; 201012-2528-003; OMB PRA Review Response to Questions

We are in receipt of the questions from OMB regarding the OMB PRA submission for the owner survey for the Assessment of the LIHTC Program After 15 Years. Below are our responses.

Question 1: It is not clear how relying on referrals from tax credit syndicators for selection is better than a statistical sample. The referral method will almost surely bias the results and will not be generalizable to the larger universe of LIHTC properties facing the same decision of whether to continue to operate as a LIHTC property. A 1/3 refusal is NOT a reason to abandon sampling. Please explain how this method is superior to a statistical sampling method and add this explanation to part B.

In preparing the OMB PRA submission for this study’s owner survey, it should have been stated and emphasized that this is primarily a qualitative study. Analysis will be completed with data collected mainly through the use of discussion guides, interviews, and open-ended questions. The decision to not rely on a sample of LIHTC property owners was in part based on the lack of data to use as the sampling frame. The data source that represents the universe of rental properties placed into service with tax credits is the HUD National LIHTC Database. Data on properties placed in service since 1995 are most complete. Properties placed in service from 1987 to 1994, the years covered by this study, are largely missing data that would be used in sampling, including project size (number of units, number of bedrooms), presence of a nonprofit sponsor, and construction type (new construction, rehabilitation).

Owner contact information from the HUD National LIHTC Database is also limited for these earliest LIHTC properties. While the some contact data may be the result of annual updates to the database, most owner contact information for the 1987 to 1994 placed in service properties were collected in 1996, when the database was first completed. It is unlikely that these contact data are current and usable to contact owners for the survey.

In finding owners through syndicator referrals we have planned to request specific types of owners and property outcomes so as to be able to gather information on the range of typical property outcomes. These typical property outcomes were learned of through discussions and data collection with tax credit syndicators. The syndicators contacted to participate in the study included a mix of syndicators known in the tax credit industry as being active in the earliest years of the tax credit program. Although none of the syndicators or owners would have been selected through a statistical sampling process, every effort has been and will be made to find the range of properties that represent the most typical outcomes for the LIHTC program’s earliest properties.

Question 2: Please explain in Part B why HUD chose a sample size of 40.

The decision to choose up to 40 owners to survey was based on budget.

Question 3: Why does this PRA request not include the interviews of Tax Credit Syndicators?

The syndicators and brokers contacted for the study had varying business structures, operations, and roles with the LIHTC program. The PRA request does not include the interviews with tax credit syndicators and brokers because those interviews were actually less formal than a structured interview and more like exploratory discussions. Because these were less formal and less structured discussions, we did not believe the interview and discussion guides needed to undertake a PRA review.

Question 4: Please elaborate on how the data will be used. Will it be made public? Used for any reports?

The data from the owner survey will be compiled for analysis and presented in a report. Data that can be tabulated will be compiled into analysis tables. Other data will be analyzed qualitatively, looking for themes and patterns that may be able to be post-coded and tabulated. The data will not be made into a publicly-available data set.

Question 5: Why does HUD believe previous LIHTC surveys had low response rates? Why does HUD believe that the same issue will arise with this survey.

We believe that previous LIHTC surveys of owners had low response rates because owners do not have a relationship with HUD through the LIHTC program. The LIHTC program is funded by the federal government through federal income tax credits. HUD has overall responsibility for housing policy for low income renters, but HUD has limited responsibility with the LIHTC program. For this reason, LIHTC property owners have had no motivation to participate in an owner survey. Also, many LIHTC owners are business entities. Businesses may be less open to respond to a survey to discuss their business operations.

Question 6: Is there any promise of confidentiality to respondents?

Yes, there is a promise of confidentiality to respondents. Potential respondents to the owner survey will be informed that:

• The information they provide will be kept confidential to the extent allowed by law.

• Data will only be presented in the aggregate.

• Participants will not be named in the research reports produced by the study, and no specific data will be attributed to specific survey respondents.

Question 7: How does HUD plan to contact/recruit respondents? Will there be any contact with respondents in advance of the survey? Perhaps a fuller explanation of the recruitment process would be useful.

In completing discussions with tax credit syndicators and brokers, we have now determined a range of property types from which we plan to base the selection of owners for the owner survey. The criteria for property types include the nonprofit status of the owner entity, status of ownership transfer (property retained or sold), whether some sort of financial bailout occurred, location and market type (strong market/weak market, non-metro area), whether the property remained affordable or was repositioned, and whether there were extended affordability use restrictions. A summary of the property types and the target number of interviews is presented in Table 1. The target number of interviews was not determined by a statistical means. The intent was to cover the range of property types.

To find and recruit owners, we have reached out to syndicators interviewed for the study. They were asked to please identify projects that met specified project types, and to also signify whether the owner may be willing to participate in the study by responding to an owner survey. With the introduction from the syndicator, we hoped to be able to more readily find survey respondents. In anticipation of conducting the owner survey, we have begun recruiting of some owners through our syndicator contacts. The recruitment has proven to be challenging, and we have not yet been able to confirm an owner is willing to participate in the study.

We do not expect to be able to find owners solely through the syndicators. After reviewing the list of potential owner interviewees found through syndicators, we have begun to look for areas of the country not represented in the syndicator-based list. We intend to review the geographic distribution of early LIHTC properties based on location data in the HUD National LIHTC Database. We will contact state allocating agencies to ask if they would be able to identify specific owners based on property type. Industry contacts have also expressed interest in the study, and if needed, we will contact them for potential owner survey respondents. We are very mindful of the need to cover the range of property types we have identified, and although selection of the owners will not be based on statistical sampling, we do not expect any bias in selecting owners for the survey.

Table 1: LIHTC 15-Year Study Project and Owner Selection Approach
Owner Type when Project was Placed in Service	Status of Ownership Transfer	Did a Bailout Occur?	Market Type	Value after 15 Years (Potential to sell for profit in excess of debt)	New Target Market	Use Restrictions	What We Want to Learn	Target Number of Interviews
These five types represent the greatest proportion of LIHTC properties (retained by General Partner (GP), remaining affordable)
Nonprofit	GP retains	No	Strong market	To be learned	Affordable		Risk of failing; if/how refinanced	3
Nonprofit	GP retains	No	Average, weak	To be learned	Affordable		Risk of failing; if/how refinanced	3
For profit	GP retains	No	Strong	To be learned	Affordable		Risk of failing; if/how refinanced	3
For profit	GP retains	No	Average, weak	To be learned	Affordable	Extended Use	Risk of failing; if/how refinanced	3
For profit	GP retains	No		To be learned	Affordable	No extended use	Risk of failing or being lost as affordable; if/how refinanced	3
These two types are bailouts which may be indicative of the kinds of problems more properties will have over time as they age
Nonprofit		Yes		To be learned			Risk of failing; if/how refinanced	3
For profit		Yes		To be learned			Risk of failing; if/how refinanced	3
These are two kinds of non-metro projects. (All to be non-Farmers Home)
Likely For profit	GP retains	No	Non-metro	To be learned			Risk of failing; if/how refinanced	4
Likely For profit	New GP	No	Non-metro	To be learned		No extended restrictions (if can be found)	Risk of failing; if/how refinanced	3
These are properties with new owners that have been through a sale; almost certain to have value
	New GP	No (or unlikely)	Strong	Yes	Affordable		How refinanced; future expectations	5
	New GP	No (or unlikely)	Average, weak	Yes	Affordable			4
	Will likely have new GP	No (or unlikely)		Yes	Repositioned	No extended use restrictions		3
Note: Blank cells indicate the criteria were not being considered for selecting that property type.