Skip Header

Working Paper Number SEHSD-WP2017-44
W. Ward Kingkade
Component ID: #ti1114366263


Housing is one of the most basic necessities in life, and housing costs usually figure as a fundamental element of a household’s expenditures.  Enabling low-income households to defray their housing costs is an important component of the social safety net.  The ability of such households to meet their housing needs is taken into account in measures of poverty, such as the Census Bureau’s Supplemental Poverty Measure (Renwick, 2011; Short, 2014; Renwick and Fox, 2016).

A question on monthly rent is included in the nationally representative American Community Survey (ACS), whose size and design would facilitate the calculation of subnational and regional measures of rent.  How to interpret responses to this item, in particular, whether they indicate the out-of-pocket expenditure of the household for rent or the total cost for rent of that unit, has been a matter of some uncertainty for ACS recipients of housing subsidies.  This uncertainty is increased by the fact that respondents in different modes received alternative instructions on what to include.  The question is whether, and for whom, the responses can be interpreted as indicative of the rental value of the dwelling unit on the market.  In the case of residents of subsidized housing, the question becomes whether the ACS item represents the entire rental cost of the unit, or only the amount the household contributes in payment, excluding the subsidy also received by the housing provider.

Administrative datasets maintained by the Department of Housing and Urban Development (HUD), the agency of the government providing support towards housing costs to low-income households, provide a separate source of data on housing support recipients’ expenditures on housing that can be helpful in establishing what the ACS rent item is capturing for these households.  These are used to address the above question in the analysis described below.

Research Questions

In the analysis that follows, we will attempt to address the following research questions:

  1. Are the ACS respondents who receive housing subsidies more likely to report their out of pocket rent payments or the full amount including the subsidy provided to the property owner?

  2. Does this likelihood vary by ACS response mode (mail self-response, computer-assisted personal interview, computer-assisted telephone interview)?

  3. Does this likelihood vary by type of housing assistance?

The data to be analyzed consist of a dataset of merged records in which HUD administrative data for household heads in 2013 and 2012 have been matched to corresponding records for the same individuals in the annual ACS housing files for those years. The results from this analysis will hopefully be able to inform subsequent research on how best to incorporate the reporting of rent from the ACS as part of the Census Bureau’s Supplemental Poverty Measure research. 

Back to Header