PD&R Edge Banner

October 26, 2021

The American Housing Survey: Then and Now

History of the American Housing Survey

The American Housing Survey (AHS) is largest, most comprehensive national survey of housing in the United States. The survey is sponsored by HUD and is collected by the U.S. Census Bureau (Census). The AHS is a longitudinal survey of housing units, meaning that it follows housing units, not households, over time. The survey collects data on structural housing characteristics, housing costs, housing quality, and demographic characteristics of households living in AHS sample units. The survey has been collected since 1973 and currently is collected every 2 years in odd-numbered years. The most recent data available are from 2019. Census completed data collection for the 2021 AHS at the end of September 2021. HUD’s Housing and Demographic Analysis Division in the Office of Policy Development and Research (PD&R) manages the survey.

Photo of a residential community with various housing.The American Housing Survey provides HUD and the public with data that can be used to analyze changes in housing and households over the past 50 years. Photo credit: shutterstock.com/tokar

Before 1973, the only available information about the U.S. housing inventory came from the decennial censuses, one-time sample surveys, and a survey of vacant housing units. The federal government recognized the need for information on the general condition of the nation’s housing to determine the need for, as well as the effectiveness of, government-sponsored programs. In 1940, Census began collecting the Census of Housing, which the agency conducted with the population census every 10 years until 2000. The idea for the AHS arose in response to turmoil in American cities in the 1960s; the survey was intended to help HUD more frequently assess current housing conditions and create programs to achieve its goals. Title V of the Housing and Urban Development Act of 1970 authorized programs of "research, studies, testing, and demonstrations relating to the missions and programs of the Department.” As a result, HUD received funding that paved the way for the AHS in 1972, and PD&R, the HUD office responsible for the AHS, was established in 1973.

With HUD funding, Census launched the first AHS in 1973 with a sample of 60,000 housing units. The survey, originally named the Annual Housing Survey, was conducted annually from 1973 to 1981. After 1981, the survey became a biennial survey because of budget constraints, and its name was changed to the American Housing Survey. The national sample was redesigned in 1985 based on data from the 1980 decennial census, with a base sample size of approximately 47,000 housing units. In 1997, the AHS moved from paper- to computer-based questionnaires. In 2015, the AHS was redesigned, and a new representative national sample was drawn using the Census Master Address File as a sampling frame with additional oversampling of selected metropolitan areas and HUD-assisted housing units; in addition, the total sample size increased to approximately 115,000 housing units. The AHS will use this sample until Census draws a new sample in 2025 or later. With data from 1973 to the present, the AHS provides HUD and the public with data that can be used to analyze changes in housing and households over the past 50 years.

Then and Now: Housing Quality, Costs, Conditions, and Remodeling in the AHS, 1973 to 2019

The 1970 census collected information on two indicators of housing quality that provided a limited account of housing quality in the United States: the presence of complete plumbing and persons per room. Three years later, the 1973 Annual Housing Survey collected data on a wider range of housing indicators that provided a more complete picture of the state of housing quality. The Lines & Numbers brief in the January 1976 HUD Challenge publication highlighted some of these housing quality indicators.

Although question wording has changed since the 1973 survey, the AHS has continued to collect information on many of these housing quality indicators. Exhibit 1 compares three housing quality indicators highlighted in the 1976 brief with results from the 2019 AHS.

The number of housing units that were occupied the previous winter and had their heating equipment break down decreased by 2.05 million units, and the share of these units decreased 5.8 percent, between 1973 and 2019. However, the percentage of occupied housing units with leaking roofs and housing units with open cracks or holes did not change significantly during this period.

Exhibit 1. Housing Quality, 1973 and 2019 (Units in Thousands)

1973

2019

Difference

Housing units with heating equipment and occupied last winter

59,301

114,553

+55,252*

Heating system broke down during the past winter ^

4,946

2,893

–2,053*

8.3%

2.5%

–5.8%*

Occupied housing units

69,337

124,135

+ 54,798*

With roof leakage^^

5,260

5,511

+251

7.6%

4.4%

–3.1%

With open cracks or holes^^^

4,179

6,761

+2,582*

6.0%

5.4%

–0.6%

^ In 1973, respondents were asked if they had experienced a breakdown in their heating equipment (defined as equipment that was completely unusable for 5 hours or more) at any time during the past winter. In 2019, respondents were asked if their unit was uncomfortably cold for 24 hours or more and if that cold was caused by a breakdown of the main heating equipment.
^^ In 1973, respondents were asked if the roof of their building or house leaked. In 2019, respondents were asked only about water leaking from the outside (for example, through the roof, outside walls, basement, or any closed windows or skylights).
^^^ In 1973, respondents were told not to include hairline cracks. In 2019, respondents were told to include only cracks that were wider than a dime.
* Difference between 1973 and 2019 is statistically significant at the 90 percent confidence level.


These housing quality indicators and other AHS data have become the backbone of HUD’s reporting on worst-case housing needs. HUD has tracked worst-case needs since 1978. The Worst Case Housing Needs Reports to Congress began in 1991 and provide national data and analysis of critical housing problems facing very low-income renting families. In these reports, renter households with very low incomes who do not receive government housing assistance are defined as having worst-case needs for adequate, affordable rental housing if they pay more than one-half of their income for rent, live in severely inadequate conditions, or both. The Worst Case Housing Needs 2021 Report to Congress was released in July 2021.

The three indicators shown in exhibit 1 are used to measure housing inadequacy. Inadequate housing is housing with moderate or severe physical problems, as defined in the AHS since 1984; these definitions have been modified periodically to reflect changes in the survey. Housing inadequacy is also used to help classify units with worst-case needs, although any of these problems alone would not qualify a unit as having worst-case needs.

Nevertheless, housing quality indicators have changed since 1973. The 1973 AHS asked respondents whether any bedrooms were used for sleeping by three or more people and, if so, how many. The 1976 brief found that 5.7 percent of all occupied housing units had bedrooms occupied by three or more people in 1973. The AHS stopped asking this question in 1978, although a data user could still calculate the percentage of occupied housing units with an average of three or more residents per bedroom.

HUD shifted its focus to overcrowding, defined as more than one person per room. In HUD’s Worst Case Needs Reports, overcrowding is considered a moderate problem rather than a severe problem that constitutes a potential worst-case need. Exhibit 2 shows the rate of occupied housing units with overcrowding from 1973 to 2019. The AHS allows for analyzing overcrowding through other characteristics not captured in other surveys, such as HUD assistance status.

Exhibit 2: Severe Overcrowding in Occupied Housing Units, 1973 to 2019

Line graph showing the rate of overcrowding in occupied housing units from 1973 to 2019.

The AHS adapts to HUD’s needs. In one such example, the AHS stopped asking respondents about rented furniture. From 1973 to 1977, the AHS asked renters if their unit came furnished or unfurnished. If the unit came furnished, respondents were then asked if the cost of the furniture was included in the rent. For unfurnished units, respondents were asked if they rented furniture from another source. Respondents were also asked about how much they paid per month for rented furniture not included in their unit’s rent.

In another example of the AHS adapting to HUD’s needs, the survey asks homeowners about recent home improvement activity. These questions inform HUD about changes to the existing housing stock. In 1973, the AHS asked respondents whether there had been any additions, alterations, and repairs made to their property and, if so, whether these projects cost $100 or more. Since that time, questions on home improvement activity have been altered, added, or removed. These changes include questions on changes to the unit’s square footage, the number of windows repaired or replaced, whether any of the interior had been painted, and whether the unit was combined with other housing units or split into several housing units.

In 2015, the AHS began asking about the reasons for home improvement activities. Exhibit 3 shows the rate of owner-occupied units with recent home improvement activity and the reason for the project. The rate of recent improvement activity in owner-occupied housing has dropped from 59 percent in 2015 to 56 percent in 2019.

Exhibit 3. Owner-Occupied Housing Units

2015

2017

2019

Owner-Occupied Units

74,299

77,567*

79,475*

With home improvement activity in the past 2 years

43,851

43,877

44,477

59.0%

56.6%*

56.0%^^

Reason for home improvement^

Accessibility for elderly or disabled

2,976

2,936

2,253*^^

6.8%

6.7%

5.1%*^^

Energy-efficiency

14,750

13,475*

12,044*^^

33.6%

30.7%*

27.1%*^^

Prepare house for sale

1,537

1,645

1,362*^^

3.5%

3.7%

3.1%

^ Respondents can list more than one reason for home improvement.
* Difference from previous year is statistically different at the 90 percent confidence level.
^^ Difference between 2017 and 2019 is statistically significant at the 90 percent confidence level.
Among households with recent home improvement activity, the number of respondents who reported conducting projects to increase accessibility decreased in 2019, which could influence policies affecting those aging in place.


In addition, the number of occupied housing units that reported completing projects to improve energy efficiency decreased from 2015 to 2019, which may affect energy-efficiency policies and programs at HUD and throughout the federal government.

Data Availability

These analyses only scratch the surface of the wealth of AHS data collected over the past 50 years. AHS data can be accessed in tabular form through the AHS Table Creator and in microdata form through public use files. More information on the AHS can be found on the Census website.

 
 
Home Page Archives