The effects of home ownership on post-unemployment wages☆

https://doi.org/10.1016/j.regsciurbeco.2018.10.006Get rights and content

Highlights

  • •

    Home ownership reduces post-unemployment wages.

  • •

    This negative wage effect is robust after using state-level mortgage interest deductions as an instrument for home ownership status.

  • •

    This negative wage effect is particularly evident in a declining housing market or a distressed labor market.

Abstract

This paper studies the effects of home ownership on job search outcomes. In contrast to previous literature focusing mainly on the impact of home ownership on unemployment rate and duration, this paper looks at the relationship between home ownership and post-unemployment wages. Using the Survey of Income and Program Participation 1996–2008 panels, I find that home ownership reduces post-unemployment wages, and this negative wage effect is particularly evident when the unemployed homeowner is located in either a declining housing market or a distressed labor market. These results are robust after using state-level mortgage interest deductions as an instrument for home ownership status.

Introduction

A great deal of research has investigated the labor market consequences of home ownership, testing whether home ownership reduces mobility and leads to inferior job search outcomes. So far, most existing studies have limited their research scope to the effects of home ownership on either unemployment rate or unemployment duration.1 The impact of home ownership on wages has received relatively little attention. To fill this gap, this paper explores wage dynamics during job searches and estimates the effect of home ownership on post-unemployment wages.

From a search model perspective, the overall effect of home ownership on post-unemployment wage is ambiguous, because it has opposite signs in local and non-local labor markets (Munch et al., 2006). Homeowners have higher costs of geographical mobility than do renters, so they are more likely than are renters to set higher reservation wages for accepting job offers in the non-local labor market2 and lower reservation wages for jobs in the local market. Thus, home ownership is likely to decrease the transition rate into employment in the non-local labor market, despite higher post-unemployment wages, but increase the transition rate in the local labor market despite lower post-unemployment wages. The net effect on post-unemployment wages, then, depends on the magnitudes of the local and non-local labor market effects.

Potential employers can also affect post-unemployment wages, which makes theoretical predictions on the wage effect of home ownership even more complicated. On the one hand, according to the wage bargaining theory, unemployed homeowners are reluctant to move and, thus, have fewer outside alternative job offers and lower bargaining power. So, they might be offered lower wages when negotiating with potential employers in the local labor market (Coulson and Fisher, 2009). On the other hand, homeowners are less likely to move, which means they may stay longer in a given job than do renters. Employers may take these facts into account and may prefer to attract homeowners by offering them higher wages (Munch et al., 2008). Thus, depending on different model assumptions, the wage effect of home ownership may have opposite signs.

On the empirical front, to the best of my knowledge, only a small number of studies have examined the wage effects of home ownership, coming to no consensus thus far. Using individual-level data in the U.S., Coulson and Fisher (2009) found that homeowners have lower wages than do renters. In line with these results, Ringo (2014) found that homeowners suffer slower wage growth than renters. Conversely, Munch et al. (2008) used Danish household-level data and found that homeowners have higher wages than do renters.

Motivated by the divergence of theoretical predictions and empirical evidence on the wage effect of home ownership, this paper contributes to the literature from the following perspectives. First, this paper provides a summary of existing theories regarding the wage effect of home ownership. With these theoretical guidelines, my empirical results provide insights into the relative importance of different theories in determining the relationship between home ownership and wages.

Second, most studies that have investigated the wage effect of home ownership (Coulson and Fisher, 2002, 2009; Munch et al., 2008; Ringo, 2014) used the annual wage as a dependent variable, without distinguishing whether home ownership affects wages either through job searches when unemployed or through on-the-job wage growth when employed. In this paper, I focus on wage dynamics during job searches by constructing a re-employment wage change, which is calculated as the difference between post- and pre-unemployment jobs.3 By using the re-employment wage change, this paper focuses its attention on job search outcomes among unemployed workers and builds a closer tie with the large body of literature that has investigated other job search outcomes such as unemployment rates and durations.

Third, I have adopted the Survey of Income and Program Participation (SIPP), which offers several advantages for studying job search behaviors compared with data used in the literature so far. With monthly information on labor market outcomes, the SIPP is particularly suited for tracking unemployment transitions and wage dynamics before and after a spell of unemployment.

Last but not least, the SIPP makes it possible to distinguish outright homeowners and homeowners with mortgage payments as well as private market and subsidized renters. Many studies have recognized that different types of ownership and tenancy play an important role in determining the effects of home ownership on unemployment durations (Flatau et al., 2003; Battu et al., 2008; Baert et al., 2014; Taşkın and Yaman, 2016; Morescalchi, 2016; Laamanen, 2017). For example, mortgage-holders face larger financial responsibilities and, therefore, are more eager to be reemployed than are outright homeowners. In the meantime, subsidized renters may stay unemployed longer, when compared with private market renters, to avoid moving and losing their housing benefits. However, we know little about whether these tenure status could also affect the wage effects of home ownership. This paper fills this gap by studying, in the discussion section, how different housing tenure status affects the post-unemployment wages.

The main challenge in estimating the impact of home ownership on job search outcomes is overcoming selection bias, which can arise for two reasons. First, homeowners may be more likely to prefer stability, that is, they may be less willing to move for a job because of their preference for stability rather than the moving costs associated with home ownership. Second, home ownership usually requires a substantial down payment and a long-term financial commitment, which require homeowners to have greater abilities and, thus, better performance in the labor market to qualify for mortgage loans. Though one can include pre-unemployment wage and relevant demographics as control variables, it is hard to measure preferences and abilities perfectly, so the Ordinary Least Squares (OLS) regression is usually biased.

The traditional approach to addressing this problem is to use instruments or exclusion restrictions, that is, variables that influence housing tenure but not labor market outcomes. Several instruments have been adopted in the literature. For example, Coulson and Fisher (2002) used the regional share of homeowners; Flatau et al. (2003) used age dummies; Munch et al. (2006) and Ringo (2014) used the proportion of homeowners in the municipality where the individual was born; Coulson and Fisher (2009) used the state-level mortgage interest deductions (MID), the proportion of multifamily housing stocks, and the gender composition of the first two children; Baert et al. (2014) used regional level home-owners rate and relative cost of owing to renting, and Laamanen (2017) used the rental housing market deregulation reform.4

In this paper, I follow the literature and conduct a Two-Stage Least-Squares (2SLS) estimation using state-level mortgage interest deductions (MID) as an instrumental variable. The tax benefits associated with MID can affect costs of owning and, in turn, the probability of becoming a homeowner in that state. The state-level tax benefits are calculated by a National Bureau of Economic Research (NBER) program, TAXSIM.5

The 2SLS estimation results show that being a homeowner reduces re-employment wage change by 0.131 dollars per hour. This negative wage effect is robust to several alternative specifications. In the meantime, I find that home ownership has no significant effect on unemployment duration. These results suggest that home ownership leads to labor market inflexibility, which causes lower wages, but does not cause longer durations of unemployment. This negative wage effect is consistent with two theoretical explanations. First, to avoid moving, homeowners are more likely to accept lower wages in the local labor market. It is also possible that homeowners have lower bargaining powers and are offered lower wages by their potential employers.

Additional sub-samples results indicate that the negative wage effect of home ownership is more evident in area where either local housing prices have declined, or local labor markets are distressed. Also, I find that, compared with renters, homeowners are less likely to experience an interstate move but are more likely to experience either an occupation or an industry change. These results are consistent with the notion that home ownership reduces post-unemployment wages by preventing efficient labor market reallocations across regions and sectors.

Further analysis reveals that homeowners are not a homogenous group as the negative wage effect is greater among homeowners with mortgages than among outright homeowners, suggesting that the financial obligation associated with a mortgage may force homeowners to be less selective when searching for a job. In the meantime, subsidized renters experience a greater wage loss during a job search than do private renters, which is in line with the explanation that they tend to give up job opportunities for below market housing costs.

The rest of this paper is organized as follows. Section 2 summarizes the literature. Section 3 outlines the theoretical framework and derives empirical implications. Section 4 describes the SIPP data and provides descriptive evidence regarding unemployment spells and wage dynamics for homeowners and renters. Section 5 presents the econometric model and identification strategy. Section 6 presents empirical findings and evidence for the robustness of these findings. Section 7 discusses the role of local housing and labor market conditions and the effects of more detailed housing tenure status. Section 7 concludes.

Section snippets

Literature review

The impacts of housing tenure on labor market outcomes have received great attention ever since Oswald (1996, 1997, 1999), who showed a positive and significant correlation between the unemployment rate and the proportion of homeowners for several OECD countries and regions between 1960 and 1990. The proposed mechanism in these papers suggested that homeowners were more likely to be unemployed because the greater moving costs involved in buying and selling their homes made it less likely for

Conceptual model

To set the stage for the empirical analysis, this section presents a few theoretical explanations as to why homeowners are different from renters regarding job search behaviors and, thus, their post-unemployment wages. First, homeowners face higher moving costs than do renters because their moves involve selling a house, and usually, buying a new one. Due to the higher moving costs, unemployed homeowners are more likely to set higher reservation wages for accepting job offers outside of

Data and summary statistics

I use data from the Survey of Income and Program Participation (SIPP) 1996, 2001, 2004, and 2008 panels. Each SIPP panel surveys approximately 40,000 households over 3–4 years and provides monthly information on employment status and wages. This longitudinal design makes it possible to construct complete unemployment spells with pre- and post-spell wages and a rich set of pre-spell characteristics, which are essential for understanding the effects of being a homeowner on wage dynamics during a

Empirical strategy

To investigate how home ownership affects job search outcomes, I estimate models of the formYi=β0+β1Hi+β2Xi+εi,where i is the unemployment spell index, and Yi represents job search outcome variables, including the re-employment wage change and unemployment duration. The re-employment wage change is calculated as the difference between the post- and the pre-unemployment jobs, and the unemployment duration is calculated as the number of weeks between two consecutive employment spells.18

Main results

Columns 2 and 3 in Table 4 report the OLS estimation results. Column 2 shows that home ownership is associated with a significant decrease of 0.03 dollars per hour in the re-employment wage change. Considering that the hourly wage has decreased by 1.11 dollars on average for renters, home ownership is associated with a three percent higher decrease when comparing with renters. Column 3 shows that homeowners experience approximately 0.367 weeks longer unemployment duration compared with renters,

Discussion

The results so far indicate that home ownership has a negative wage effect. Will this wage effect vary with local housing and labor market conditions? Dose home ownership affect wages through geographic, occupation, or industry mobility? Dose being either an outright homeowner or a subsidized renter have an impact on this wage effect? This section attempts to answer these questions by discussing the role of these factors empirically.

Conclusion

This paper studies the labor market consequences of being a homeowner. In contrast to previous literature that has focused on unemployment rate and duration, this paper highlights a negative wage effect of home ownership. By focusing on unemployed workers, my findings indicate that home ownership reduces post-unemployment wages during a job search, particularly in areas where either the housing prices are in decline or the labor market is distressed. A further analysis distinguishes between

References (57)

  • John M. Abowd et al.

    Did the housing price bubble clobber local labor market job and worker flows when it burst?

    Am. Econ. Rev.

    (2012)
  • Joshua D. Angrist et al.

    Children and their parents' labor supply: evidence from exogenous variation in family size

    Am. Econ. Rev.

    (1998)
  • Stijn Baert et al.

    Does homeownership lead to longer unemployment spells? The role of mortgage payments

    Economist

    (2014)
  • Javier A. Barrios Garcia et al.

    User cost changes, unemployment and home-ownership: evidence from Spain

    Urban Stud.

    (2004)
  • Harminder Battu et al.

    Housing tenure, job mobility and unemployment in the UK

    Econ. J.

    (2008)
  • Christopher F. Baum et al.

    Enhanced routines for instrumental variables/GMM estimation and testing

    STATA J.

    (2007)
  • Patrick Bayer et al.

    Place of work and place of residence: informal hiring networks and labor market outcomes

    J. Polit. Econ.

    (2008)
  • David G. Blanchflower et al.

    Does High Home-ownership Impair the Labor Market?

    (2013)
  • Audra J. Bowlus

    Matching workers and jobs: cyclical fluctuations in match quality

    J. Labor Econ.

    (1995)
  • Raj Chetty

    Moral hazard versus liquidity and optimal unemployment insurance

    J. Polit. Econ.

    (2008)
  • Kathleen P. Classen

    The effect of unemployment insurance on the duration of unemployment and subsequent earnings

    Ind. Labor Relat. Rev.

    (1977)
  • Robert Collinson et al.

    Low-income housing policy

  • Edward N. Coulson et al.

    Tenure choice and labour market outcomes

    Hous. Stud.

    (2002)
  • Edward N. Coulson et al.

    Housing tenure and labor market impacts: the search goes on

    J. Urban Econ.

    (2009)
  • Ronald G. Ehrenberg et al.

    Unemployment insurance, duration of unemployment, and subsequent wage gain

    Am. Econ. Rev.

    (1976)
  • Henry S. Farber

    Unemployment in the great recession: did the housing market crisis prevent the unemployed from moving to take jobs?

    Am. Econ. Rev.

    (2012)
  • Daniel Feenberg et al.

    An Introduction to the TAXSIM model

    J. Pol. Anal. Manag.

    (1993)
  • Paul Flatau et al.

    Home Ownership and Unemployment: Does the Oswald Thesis Hold for Australian Regions?

    (2002)

    • Cited by (8)

    View all citing articles on Scopus
    ☆

    I am grateful to Robert Moffitt for guidance and support. I also thank Yuya Sasaki, Yingyao Hu, Hülya Eraslan, Richard Green, Gary Painter, Junfu Zhang, Lynn Fisher, Shimeng Liu, Chun Kuang, and participants at the JHU Applied Micro Seminar, the USC Lusk Center Seminar, the AREUEA National Conference, and the Annual Meeting of the Urban Economics Association for helpful comments. Finally, I am very grateful to Dan Feenberg for his help in using the micro tax return data and the TAXSIM calculator.

    View full text
    © 2018 Elsevier B.V. All rights reserved.