2015 Series • No. 2015–4
Current Policy Perspectives
Household Formation Over Time: Evidence from Two Cohorts of Young Adults 
Residential investment accounts for an important component of U.S. gross domestic product, and traditionally plays a strong role in business cycle expansions. U.S. residential investment has improved slowly during the recovery from the Great Recession, despite a relatively strong national rebound in house prices and record low interest rates. An important determinant of residential investment is the household formation rate, which is largely driven by young adults moving out of their parents' homes after completing high school or college. New household formation can be offset when existing households combine, typically through marriage or by moving in with parents or other relatives for economic reasons. This paper uses National Longitudinal Survey of Youth (NLSY) data from the 1979 and the 1997 cohorts to examine how various demographic, economic, and geographic factors influence the rate of household formation among young adults, both within cohorts and over time across cohorts.
Key Findings
- Comparing parental co-residence rates for young adults between the ages of 23 and 31 years shows that the share of individuals living with parents declines with age, but that the share of those living with parents is higher at nearly every age for the 1997 cohort compared to the 1979 cohort.
- There is important variation in household formation by race both within a given cohort and over time. The share of black youth living with parents is substantially lower at young ages in both cohorts, but after the late teenage years, blacks and Hispanics are more likely to be living with parents than non-black/non-Hispanic youths. In the 1997 cohort, non-black/non-Hispanic and Hispanic youths, regardless of age, are more likely to be living with parents relative to their 1979 counterparts, while the rate of living with parents for blacks is unchanged.
- Overall, housing costs have a meaningful effect on the decision of young adults to live with parents. The share of the 1979 cohort living with parents rose with the cost of housing. Among the 1997 cohort, 23 year-olds living in regions with high housing costs were about 15 percent more likely to be residing with parents than same-age members of the 1979 cohort who were living in areas with low housing costs.
Exhibits
Implications
The almost twenty-year time span separating the two cohorts captures the changing pattern of U.S. household formation over time. The results suggest that failing to account for demographic shifts in the racial composition of the U.S. population could lead to inaccurate predictions about the future number of U.S. households. It also is important to account for local economic conditions when projecting the future number of U.S. households. While differences in housing costs and business cycle conditions explain nearly 45 percent of the variation in household formation rates across the two cohorts, even after controlling for observed differences between the two cohorts, members of the 1997 cohort are more likely to be living with their parents. This finding suggests that there has been a recent decline in the U.S. household formation rate for young adults that cannot be easily measured or accounted for by observable factors. In addition, while the racial composition of the U.S. population appears to be important for predicting household formation rates going forward, it is also true that population demographics are slow to change. So a shifting racial mix in the United States is unlikely to explain the sizeable decline observed in aggregate household formation rates starting around 2006. More work is needed to explain the recent U.S. shift in household formation rates.