Housing Performance and the Electorate

Abstract

The electorate is concerned with personal financial and macroeconomic conditions in addition to policy issues and tends to hold the incumbent party accountable when voting. The performance of the United States’ largest asset class, residential real estate, should influence individual voter behavior. According to economic voting theory and the “homevoter” hypothesis, homeowners will be more supportive of policies that are perceived as beneficial to their property value. We investigate this relationship in the US residential real estate market by evaluating the effects of heterogenous county-level housing market performance on voter behavior in national presidential elections and find counties with superior house price performance in the four years preceding an election are more likely to “vote-switch” to the incumbent party. Counties with relatively inferior house price performance in the four years leading up to the election are more likely to switch their vote from the incumbent to the challenging party. The relationship is strongest in the years closest to an election and in counties that rank in the higher quartile of housing price performance. Election outcomes in “swing” counties are particularly vulnerable the local real estate economy. To our knowledge, this is the first study to link heterogeneous local residential real estate performance over a series of national election outcomes.

Keywords:

• Homevoter
• Housing
• Economy
• Elections
• Electorate
• Voting Theory, Real Estate

Notes

1 For literature review, see Anzalone (2016), Kiewiet and Rivers (1984), Lewis-Beck and Stegmaier’s (2000), Monroe (1984), Norpoth (1996), Norpoth (1985), Lewis-Beck (1990: chs. 2 and 3), Nannestad and Paldam (1994), and Frey (1978).

2 See, for example, Sirri and Tufano (1998), and Ivkovic and Weisbenner (2009) for discussion of return chasing behavior.

3 The 2000 election results serve as the prior election vote for the 2004;county-level HPI holding period returns are calculated in the one, two, three, and four years before each election beginning in 2004.

4 $\text{Total flips}=(336\mathrm{*}1+216\mathrm{*}2+88\mathrm{*}3+1\mathrm{*}4)=1,036$

5 FHFA House Price Index | Federal Housing Finance Agency. (n.d.). Retrieved December 18, 2022, from https://www.fhfa.gov/DataTools/Downloads/Pages/House-Price-Index.aspx

6 County's education level data is not available for each year. To proxy education attainment for each county, we use the average county education level in 2000 with the 5 years average from 2015 to 2019.

7 GDP, per capita income, per capita dividends interest and rent, unemployment insurance compensation, retirement and other (retirement and disability insurance benefits, medical benefits, veterans' benefits, education and training assistance, other transfer receipts of individuals from governments, current transfer receipts of nonprofit institutions, current transfer receipts of individuals from businesses), employer contributions for government social insurance, data are collected from the U.S. Bureau of Economic Analysis (BEA).

8 In terms of political party, we do not find significant differences between parties related to HPI in the analyses measuring counties likelihood of voting for the incumbent party, or vote-switching to the nonincumbent party; however, we find some evidence that HPI matters less for counties that vote-switched to the incumbent party when a Republican was in office. However, this was only measurable in 3 elections during our sample period thus reducing the sample size, and only 23% of counties actually switched their party vote at least once during the last five elections.

9 2004 election, George W. Bush, Republican party; 2012 election, Barack Obama, Democratic party; 2020 election, Donald Trump, Republican party.

10 We also interact unemployment with HPI index in the full models. We find that in higher unemployment times, incumbent voters’ response to house price appreciation lessens.