Skip to main content

Changes in the U.S. Housing Stock and Rental Market Dynamics

HUD USER Home > PD&R Edge Home > Trending

Changes in the U.S. Housing Stock and Rental Market Dynamics

Image of Shawn Bucholtz,Former Director of PD&RFormer Director Shawn Bucholtz, Director of PD&R's Housing and Demographic Analysis Division.

Two new reports document the evolution of the housing stock from 2015 to 2017.

From 2015 to 2017, new construction was the primary source of additions to the housing stock. Demolition and fire or natural disasters caused most permanent losses. At the same time, approximately 20 percent of the housing stock changed from owner-occupied to rental units or vice versa.

These findings, and detailed accounts of the changes that occurred in the housing stock between 2015 and 2017, can be found in two new reports featuring data HUD collected in the American Housing Survey (AHS).

The latest Components of Inventory Change (CINCH) and Rental Market Dynamics reports document the evolution of units added and removed from the housing stock between the 2015 and 2017 AHS surveys. Because the reports use AHS data collected from the same housing units every 2 years, researchers can use these studies to track changes over time in the characteristics of individual housing units as well as the households who occupy these units.

"The housing stock is constantly changing in response to consumer demand and external forces," says Shawn Bucholtz, head statistical officer and director of housing and demographic analysis at HUD. David Vandenbroucke, a senior economist at HUD, explains, "Units are added and disappear. They may be split, merged, or converted to nonresidential use and then change back. These studies give us a window into the complexity underlying the housing stock."

The story of the housing stock: Postrecession recovery

According to the national CINCH report, the U.S. housing stock increased by 2.6 million units between 2015 and 2017. This net change resulted from losing approximately 2.1 million units while gaining nearly 3.7 million units, plus a statistical correction of 1.1 million due to changes in survey weights. About half of the losses to the housing stock, 1,009,300 units, were the result of permanent causes, such as destruction by fire. 1,084,400 units were classified as temporary losses, such as a house used as a tax preparer's office that might later return to the housing stock. The new construction of 2,102,700 units accounted for most of the additions to the housing stock. The return of temporary losses added another 1,050,300 units, and 502,800 housing stock additions resulted from other causes, such as older mobile homes that moved and the conversion of structures built for other purposes.

According to Vandenbroucke, "Unlike other data sources, CINCH analysis demonstrates that there's more going on in housing than you can see from just subtracting one total from the previous one. There's a lot of churning of units — large movement into and out of the stock, or between the rental and owned categories, and units becoming more and less affordable. You can see that only through a CINCH kind of analysis."

In addition to describing the housing stock's evolution between 2015 and 2017, the new CINCH study compares the losses and additions that occurred in the most recent 2-year period to the losses and additions experienced in the six preceding 2-year periods between 2001 and 2013. Between 2015 and 2017, the housing stock experienced its third-highest increase of the six 2-year periods since 2001. The data show the impact of the 2006 financial crisis and the economic recession that followed, documenting a notable drop in growth in the 2007 to 2009 and 2009 to 2011 periods. The slowest growth in the housing stock — only 413,000 units — occurred between 2011 and 2013, but the most recent data confirm that the housing market returned to normal growth between 2015 and 2017.

The 2015 to 2017 CINCH report examines how loss and addition rates vary according to the characteristics of the units or their occupants. Owner-occupied units had lower loss and addition rates than all other housing types, whereas rental-occupied units had higher loss and addition rates. Higher-cost units, whether owner occupied or rentals, had a higher addition rate than did lower-cost units. Lower-cost units had a higher loss rate than did higher-cost units, and poorer-quality units had higher loss and addition rates than did higher-quality units. Displacement resulting from losses affected all households equally; loss and addition rates showed little variation based on age, race, ethnicity, household type, or the presence of children in the household.

Geography was not a significant factor in the movement of units into and out of the housing stock. Kathryn Nelson, a consultant working with HUD on affordable housing issues, explains, "There were very few significant differences among the nine Census divisions. However, additions to the stock were highest in the West South-Central area and lowest in New England."

Rental housing: Little growth, big changes in cost and affordability

The latest Rental Market Dynamics report uses AHS data and the CINCH methodology to study how the rental market — including occupied and vacant units — evolved between 2015 and 2017, with a focus on affordable rental housing.

The 2015-2017 Rental Market Dynamics report reveals little change in the number of rental units in the stock between 2015 and 2017. Rental inventory increased by only 2,000 units compared with a total of more than 48 million units. This net change masks the movement of about 7 million units into and out of the housing stock, primarily due to conversion to or from owner-occupied status, new construction, and physical losses.

Although the increase in renter-occupied units was small, rental housing costs and affordability categories changed significantly between 2015 and 2017. The number of renters paying at least $1,250 per month increased by more than 2 million, and median monthly rents rose from $923 to $991. "In terms of affordability," Nelson adds, "we are most concerned about the lowest rent categories." The number of extremely low-rent units increased by 644,000, but very low-rent rentals declined by more than 1.6 million units.

Over a longer period, from 2001 to 2017, the total rental stock increased by 11.3 million units. The number of units affordable to households with incomes below 50 percent of area median income (nonmarket, very low-income, and extremely low-income renters) fell by 6 million units. Over this period, the proportion of such affordable units dropped from 57.8 percent to 35.5 percent of the total rental stock.

New: Metropolitan-area-level analysis

HUD has also released two additional reports using AHS data on the Houston and Detroit metropolitan areas, combining CINCH and rental market analysis in one shorter volume. The 2015-2017 CINCH and Rental Market Dynamics reports are available here on

Published Date: 28 September 2020

The contents of this article are the views of the author(s) and do not necessarily reflect the views or policies of the U.S. Department of Housing and Urban Development or the U.S. Government.