• About Us
  • Mission and Goals
  • Staff
  • Board
  • History
  • Annual Reports
  • Policy Commitees
  • Jobs & Internships
  • Donate
    • CFC Donations
  • Donors
  • Renew Your Membership or Join Today
  • Looking for Housing?
• Issues & Initiatives
  • National Housing Trust Fund
  • Foreclosure Intervention
    • Protecting Tenants at Foreclosure
  • Preservation Initiatives
  • Disaster Recovery
  • Federal Budget & Approps
  • Housing Plus Services
  • Federal Housing Programs
    • Housing Related Issues
    • Other Housing Programs
  • Testimony & Comments
  • Calls to Action
  • NLIHC Fact Sheets
• Research & Publications
  • Memo to Members
  • Out of Reach
  • 2010 Advocates' Guide
  • Research
  • Congressional District Profiles
  • Publications Order Form
• Press
  • Press Releases
  • Press Contact
  • NLIHC Media Awards
    • NLIHC Media Awards Previous Winners
  • State and Local Organizing Award 2010
• State Resources
  • State Partners & State Level Information
  • States Internet Research Repository
  • Contact Elected Officials
  • States' Upcoming Events
  • NLIHC Outreach Team

Federal Data Sources for Housing Advocacy

May 6, 2009
By Keith Wardrip, Senior Research Analyst, National Low Income Housing Coalition

Accurate, reliable demographic and socioeconomic data are crucial to housing researchers and advocates alike from the national scale down to the neighborhood level.  While there are countless resources for both quantitative and qualitative housing data at the state and local level, data collection efforts with national coverage and a housing focus are fewer in number and are in danger of shrinking further as budget cuts loom.

Sources of Housing Data
The Census. As it has been since 1790, the decennial census is the only comprehensive count of the U.S. population (see table).  The Census Bureau attempts to achieve a full count of the population by distributing a questionnaire containing basic demographic questions (e.g., age, sex, race) to all households and individuals living in group quarters (e.g., military barracks, nursing homes, college dormitories, prisons, etc.) every ten years.  Census figures describe the U.S. population at a specific point-in-time (e.g., April 1) during the census year.

While in recent years there have been growing concerns of undercounting in poor and minority urban populations, the decennial census conducted by the Census Bureau is the official source for counts of the number of people and houses in the United States, and it is used to apportion congressional representatives among the states, draw legislative districts, determine the number of electoral votes assigned to each state, and distribute federal funds.

Current issues. Representing the most significant data collection effort on the horizon, Census 2010 will be conducted in April of next year. Budget cuts and the failure of a major contract for handheld computers prevented the full implementation of important field tests in 2008, but President Bush met the Census Bureau’s request for an appropriation of roughly $3 billion in the FY09 Continuing Resolution. In addition, the American Recovery and Reinvestment Act of 2009 (ARRA) included $1 billion for the Census Bureau that will fund measures designed to increase the accuracy of the census, particularly as it relates to hard-to-count populations. These measures include the hiring of additional census takers, increasing the number of community partnership specialists, and stepped-up efforts at outreach and advertising. The omnibus bill that is expected to control appropriations for the remainder of FY09, as well as the FY10 budget, must include sufficient funding to allow the Census Bureau to maximize the accuracy of Census 2010.

Many consumers of federal data have long supported the removal of the Census Bureau from the Department of Commerce, and legislation introduced but not enacted in the 110th Congress would have done just that, making it a stand-alone agency and creating a fixed five-year term for its director. Although the future of this legislation is uncertain, it has recently been reported that the Obama White House will directly partner with the Commerce Secretary in overseeing Census 2010, representing a bigger role for the White House than in recent censuses.

American Community Survey. Historically during the decennial census, one in six households has received an expanded questionnaire, or ‘long form,’ that also includes specific questions regarding a household’s income, education, employment, and other socioeconomic characteristics.  While continuing to conduct a census every ten years through the use of the ‘short form,’ as constitutionally mandated, the Bureau intends to replace the survey component of the decennial census (i.e., the long form) with the American Community Survey (ACS) in 2010.

Under development since the mid-1990s, the ACS has produced annual estimates for every jurisdiction with more than 250,000 residents since 2000. (Full data from the first two years are available through the 2000 and 2001 Supplemental Surveys.)  The sample size was expanded from 800,000 to three million households when data were collected in 2005; as a result, one-year estimates for jurisdictions as small as 65,000 residents in the 50 states, the District of Columbia, and Puerto Rico have been available since the release of the 2005 data. In addition to these one-year estimates, the release of 2007 data in 2008 was the first to include estimates based on three years of data for all areas with a population greater than 20,000. And ACS data released in 2010 will include five-year estimates of population and housing for areas as small as block groups.  In 2006, the survey was expanded to include the population living in group quarters; as a result, 2006 ACS estimates and those that follow are considered more comparable with decennial census estimates.

As is true with all surveys – including the long form component of the decennial census that it will replace – there are margins of error associated with ACS data because estimates depend on the responses of a sample of a population, rather than every member.  Furthermore, since the sample is based on official census population estimates, the decennial census and the Bureau’s Population Estimates Program remain the preferred source for official population counts.  Unlike the point-in-time nature of the decennial census, the ACS produces period estimates and is thus ideally suited for describing the characteristics of a population during the data collection period and for measuring annual differences across geography and through time.

American Housing Survey. The American Housing Survey (AHS) is the only comprehensive national survey specifically focused on housing. The survey is longitudinal in nature, tracking changes in the same housing units over time, and produces national and regional estimates every two years. Historically as many as 47 of the largest metropolitan areas were surveyed on a rolling basis every six years.  However, due to budget constraints, the number of metropolitan areas to be over-sampled as part of the national survey was reduced to 21 in 2007, with seven surveyed every two years; even with the reduction in the number of metropolitan areas, sample sizes have been reduced to meet budget demands. National data were last released in 2008, as were estimates for the following metropolitan areas: Baltimore, Boston, Houston, Miami, Minneapolis, Tampa, and Washington, DC.

Current issues. AHS continues to suffer as a result of cuts to the budget for HUD’s Office of Policy Development and Research (PD&R). Budget cuts dating back to FY06 reduced the survey’s sample size and scaled back the number of metropolitan areas from 47 to 21 (seven surveyed every other year in a six-year cycle). More recent cuts have further reduced the number of metropolitan areas to be surveyed in 2009 from seven to six and call into question the survey’s future. Although PD&R is planning to conduct the AHS in 2011 and has requested the funds to restore the survey to 47 areas, its future is up in the air at the time of publication.

Residential Finance Survey and Home Mortgage Disclosure Act. A focus on housing finance, rather than people or units, sets the Residential Finance Survey (RFS) and Home Mortgage Disclosure Act (HMDA) data apart from the Census Bureau efforts discussed above.  Last conducted in 2001, the former is a decennial investigation of the financial characteristics of residential properties, and the latter is an annual collection of data from disclosure filings made available to advocates to monitor the lending patterns of financial institutions. At publication, the most recent HMDA data available covered mortgage lending that occurred in 2007.

Current issues. Another potential casualty of PD&R’s budgetary woes, the 2011 RFS is currently unfunded and is not expected to be conducted following the decennial census as it has been since 1951. However, recent conversations among staff at HUD, the Federal Housing Finance Agency, the Department of Treasury, and the Federal Reserve suggest that all recognize the importance of collecting financial data in both the single-family and multifamily housing sectors. Although no formal agreement has been reached, the hope is that these groups will collaborate on a survey to replace the RFS and collect this time-sensitive information much more frequently than every 10 years.

Current Population Survey. The last national survey detailed in the following table – the Current Population Survey, or CPS – is distinct in that it does not produce any estimates of housing characteristics.  It is mentioned here because it includes an Annual Social and Economic (ASEC) Supplement, which is the source of official estimates of income and health insurance coverage of the non-institutionalized population (i.e., individuals not considered ‘patients’ or ‘inmates’) and is the primary source of data on the annual poverty status of U.S. residents.  Thus the CPS is a very important source of data for low income housing advocates.

Data on the subsidized rental stock. HUD makes publicly available information on the location and characteristics of a subset of the nation’s federally subsidized rental housing stock. While HUD does not produce a comprehensive, integrated dataset, it does provide project-level files for the following programs: project-based Section 8 and other federal rent subsidies for multifamily housing; FHA insured/subsidized mortgages; Sections 202 and 811; and the Low Income Housing Tax Credit program. In addition to data for these individual programs, HUD produces a dataset called Picture of Subsidized Households, which includes public housing in addition to the previously mentioned files but which currently reflects data collected in 2000.

With an understanding of the programs, database skills, and significant effort, advocates can integrate these datasets to create a partial database of the subsidized housing in a particular geography. And thanks to data released by HUD in 2008, the database can include the three most recent Real Estate Assessment Center (REAC) scores that quantify the properties’ physical conditions. Information for projects receiving subsidies from the following programs will be omitted, however, because HUD does not make it available to the public: HOME; tax-exempt multifamily housing bonds; Section 8 Mod Rehab; project-based vouchers; HOPWA; and McKinney-Vento permanent housing. Data on subsidized rural housing are also not publicly available.

Fair Market Rents. HUD also annually updates Fair Market Rents (FMRs) for every metropolitan area and rural county in the U.S. Although it is primarily an administrative dataset used to determine the ‘payment standard amount’ for the Section 8 voucher program, it is of interest to housing advocates given its frequency and comprehensive geographic coverage. Commonly set at a community’s 40th percentile gross rent, FMRs reflect HUD’s best estimate of the cost of a decent, modest apartment and are published for various unit sizes.

Additional Surveys. Other surveys not described in the table but nonetheless of importance to housing advocates and researchers include:
Housing Vacancy Survey, a Census Bureau survey that quantifies rental and homeowner vacancy rates, the characteristics of vacant units, and the overall homeownership rate on a quarterly (nation, regions) and annual (states, 75 largest metropolitan areas) basis. Data collected for the Housing Vacancy Survey are also used to produce the annual CPS estimates.
Survey of Construction, a Census Bureau product that tracks the number and value of residential units permitted, constructed, sold, and improved for the nation and select metropolitan areas.
Survey of Market Absorption, a HUD-sponsored survey of the absorption rate of newly constructed multifamily units conducted by the Census Bureau.
Survey of Income and Program Participation, a Census Bureau survey that tracks families for between two and four years, investigating household members’ sources of income, participation in and effectiveness of government transfer programs, and basic demographic characteristics.

Federal preservation data legislation. Advocates can also use data from HUD on the location and characteristics of certain subsidized properties to develop a database of assisted housing in a particular geography. This kind of database can be used to preserve affordable housing by raising awareness of projects at-risk of leaving the subsidized housing stock. (See NLIHC’s work summarizing the available data and how it can be used at http://nlihc.org/template/page.cfm?id=133.)

However, data-driven preservation efforts today cannot be maximized because HUD does not publish data on all of its programs in one easy-to-use database. To the contrary, the data that HUD currently makes available to the public are not complete, not always updated in a timely fashion, and require a certain level of database expertise to utilize fully. Legislation is expected to be introduced in the House that would require HUD to improve upon the quantity, quality, and usability of the subsidized housing data that the agency provides to the public.

Tips for Local Utilization and Success
First and foremost, housing advocates should encourage everyone to fully participate in Census 2010 and to respond to other federal housing surveys. The research conducted with these datasets can only fully capture the housing experiences of the nation if everyone is counted.

Advocates can also be end-users of the vast array of survey and census data. Research produced by advocates both clearly illustrates the depth and breadth of the affordable housing crisis and also demonstrates the importance of these federal data collection efforts. Quantifying the problem by calculating the scarcity of units affordable to the lowest income families, for example, can make it easier to set specific and defensible goals for expanding the affordable housing stock.

What to Say to Legislators
Call your Members of Congress and ask to speak to the person who deals with appropriations  with the message that funding for the collection and analysis of housing data is vital to understanding the breadth and depth of the nation’s affordable housing crisis. Informed and effective housing policy is possible only with a concrete understanding of today’s housing issues. Ask the Member to support the collection and analysis of housing data in the appropriations process by:

Meeting the Census Bureau’s request for funding to prepare for and carryout an accurate Census 2010.
Restoring years of cuts to HUD’s Office of Policy Development & Research’s budget, at least to its pre-2007 level, that would allow it to once again fully implement the American Housing Survey.
Continuing to fully fund the American Community Survey.

For More Information
Many organizations that understand the importance of federal statistics have formed coalitions and membership groups that track federal data collection efforts, advocate for their continued funding, and provide members with an opportunity to communicate directly with the federal agencies collecting the data. A few of these groups include the Council of Professional Associations on Federal Statistics (http://www.copafs.org/), The Census Project (http://www.thecensusproject.org), and the Housing Statistics Users Group (http://groups.google.com/group/housing-statistics-users-group).

National Low Income Housing Coalition • 202-662-1530 • www.nlihc.org