Holding Onto the Past: Previous Homes, Post-Move Housing Consumption, and the Great Recession

Abstract

We document that households relocated during the 2007-2009 Great Recession and its aftermath were substantially more likely to hold their previous homes for an extended period of time. We identify two contributing factors to this phenomenon. First, falling house prices pushed many homes into the “negative-equity” and “near-negative-equity” territories, and this made it challenging for owners to sell their homes. Second, we also show that falling home values had a more widespread effect that made all homeowners, regardless of their equity positions, more reluctant to sell. Additionally, we find households without mortgages are more likely to hold previous homes. Overall, we show the relationship between the loan-to-value (LTV) ratio and the likelihood of holding is U-shaped. We further examine the impact of holding previous homes on post-move housing tenure and housing consumption choices. We find that holding previous homes is associated with renting for a longer period. For households that bought new homes after relocation, holding previous homes is associated with the new residences that are less expensive and smaller. Our results suggest that, for households that moved during the housing bust, the Great Recession has a long-lasting effect on their housing consumption choices.

Keywords:

• Housing consumption
• the Great Recession
• loan-to-value ratio

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 $75,000,000\times 4.5\%\times 6=20,250,000.$ To be conservative, we use the annual moving rate of 4.5% in 2011 for our calculation. This is the lowest moving rate of homeowner households during the 2006–2018 period (Frost, 2020).

2 The U.S. Census reports 2.62 per household during the 2015–2019 period. Homeowner households tend to be larger than renter households. Therefore, our estimate is relatively conservative: $20,250,000\times 2.62=53,055,000.$

3 Bian et al. (2018) show that housing prices can be distorted and inflated with mortgage financing, and such distortion is even more severe for subprime mortgages, which lead to the housing bubble eventually resulting in the 2007–2009 housing crash due to borrowers’ defaults for various economic and behavioral reasons (e.g. Green & Wachter, 2005, Seiler, 2015a, Seiler, 2015b, Seiler, 2018).

4 In addition to these effects, other factors linking house prices and household mobility include seasonality (Goodman, 1993) and corporate relocation assistance (Allen et al., 1997).

5 See Ferreira et al. (2010), Schulhofer-Wohl (2011), and Coulson and Grieco (2013).

6 Coulson et al. (2002) provides an excellent review of the social benefits of homeownership and some related issues.

7 The PSID was conducted annually from 1968 to 1997 and biennially after 1997.

8 We eliminate from our sample the small fraction of mobile homeowners due to the idiosyncrasies of their housing arrangements and mobility patterns.

9 Each PSID survey covers two preceding years. Therefore, the 2009 survey covers the 2007–2009 period.

10 We later extend our analysis by substituting $\mathit{\text{Cold}}\_{\mathit{\text{Market}}}_t$ with the state-level FHFA House Price Index to capture the magnitude of house price movements across states.

11 Information on mortgages more than the first two is not collected during most of our sample period. PSID did survey households about whether or not they had a third mortgage in 1999 and 2001. No household in our sample reported having a third mortgage in either year.

12 See Jasso (1992) for a list of validation studies.

13 The focus of our study is the effects of housing equity position and housing market cyclicality on the likelihood of holding previous homes. We acknowledge that there could be many other reasons why people decide to hold their previous homes. In many cases, several motivations may simultaneously drive the decisions to hold. Given the limitations of our data, we are unable to tease out these motivations.

14 Chan et al. (2016) and Choi and Painter (2018) document that underwater homeowners are more likely to overstate home values. In other words, some people are reluctant to admit to being underwater. Those households, which in fact have negative equity, are bunched together with others with thin levels of equity in our near-negative equity group. Therefore, estimates on the near-negative equity group may reflect the combined effect of near-underwaterness and unadmitted-underwaterness. This does not alter our conclusion that elevated LTV ratios raise the likelihood of holding previous homes.

15 Henceforth, we use the term negative equity effect to mean the effect associated with both negative-equity and near-negative-equity homes.

16 As a robustness check, we substitute ${\mathit{\text{HPI}}}_{t-2}$ with HPI_t, the contemporary house price change. The results are similar.

17 While the relative difference of marginal effects between $\Delta \mathit{\text{HPI}}\_\mathit{\text{Up}}$ and $\Delta \mathit{\text{HPI}}\_\mathit{\text{Down}}$ is substantial, the magnitudes of both are rather small. We suspect this is, at least partly, because the HPI, a state-level price index, is unable to capture variations of house price movements across different regions within a state.

18 Benmelech et al. (2022) document that households spend on average $8,000 more on home-related durables and home improvements in the two years following a home purchase.

19 We thank an anonymous reviewer for making this suggestion.