• homeownersip
• job search

Housing tenure structures differ considerably across OECD countries. Typically, the share of owner-occupied housing is very high in Southern European countries (Spain, Italy, Greece), while relatively low in Austria, Germany, France, the Netherlands and in some Nordic countries such as Sweden, Finland and Denmark (see Catte et al. [2004]). This share has increased in most OECD countries during last two decades. Italy, Spain, Portugal, the Netherlands and Belgium registered the largest increases from the eighties to the mid of the first decade of the century. In the UK, the homeonwership rate increased steadily up to 2005, then dropped sharply in the last few years.
High homeownership rates and their increments have been in part determined by pro-ownership policies implemented throughout most OECD countries. Such policies can take the vest of tax reliefs for mortgage interest payments, exemptions for capital gains taxation applied to residential property, or generous subsidies for low-income families to reduce the costs of homeownership.
Although the orientation of several governments, particularly in Europe, has been to promote homeownership, economists have raised several concerns about the consequences of large homeownership shares on the functioning of the labour market. The analysis of the relation between the housing tenure and the labour market dates back to mid-nineties, when Andrew Oswald pinned on high homeownership rates the blame for the high unemployment rates in Europe (Oswald [1996], Oswald [1997], Oswald [1999]). Making use of aggregate data, Oswald found a strong positive statistical association between the two rates at the international and regional level. Thus, his receipt to reduce unemployment was strikingly at odds with the prevailing political wisdom: “We can put Europe back to work . . . by reducing homeownership” (Oswald [1999])1.
Oswald identified five mechanisms that drive the positive impact on unemployment. These are mostly related to the idea that the mobility constraints imposed by homeownership introduce frictions in the labour market which harm its efficiency. The first and most important, is a direct effect. Owner occupiers incur higher transaction costs for selling home and moving to a new accommodation, which can be even higher whenever capital gains for houses are taxed and there are no exemptions. The other four mechanisms can be thought as externalities of homeownership. Second, where the rental market is thin, everyone, regardless of the housing tenure, can have difficulties in finding the accommodation close to the job they aspire to, so that efficient job matches are hindered. This is the case even for mobile people who would be happy to move for job reasons, such as young still living in the family-home in a kind of free-rent status. Third, owning the accommodation hampers also job-to-job changes and prolong poor job matches hence harming the whole economic system efficiency. Fourth, owner-occupiers are more likely than renters to lobby for deterring entrepreneurs from investing in their residential area. Fifth, since owners commute much more than renters, and over longer distances, higher homeownership rates in the area lead to higher transport costs and possibly to congestion costs, which make getting to work more costly for everyone and act as a sort of unemployment benefits in increasing the relative attractiveness of not working.
Although the argument underlying a positive relationship between homeownership and unemployment originates from macroeconomic evidence, its theoretical foundations are mainly microeconomic and a vast part of the empirical research carried out subsequently has been based on micro data. The typical test of the so-called “Oswald thesis” consists in estimating the homeownership effect on the duration of the unemployment spell or, less often, on the probability to be unemployed. The expected sign is positive, along the main idea that higher mobility costs due to property holding reduce the willingness to accept job offers which require a residential move. Fewer studies, and only recently, have focused on the employment duration, testing whether homeownership can reduce job finding rates also for people already employed.
Generally, empirical tests on micro data find no support for the Oswald hypothesis, and in most cases even that homeowners have shorter unemployment spells and lower likelihood to be unemployed, which are exactly counter-Oswald results. However, homeownership seems to increase job stability by hampering job-to-job changes for people already employed.
As regards unemployment outcomes, the existing literature has put forward two possible reasons for the falsification of the Oswald hypothesis. The first one looks at the different effect of mobility costs on job search behavior in the local and in the non-local labour market. In fact, homeowners may have higher reservation wages for jobs which require a residential move (nonlocal market), but also lower reservation wages for jobs which do not (local market), so that job finding rates for the latter may be as high as to offset the lower rates for jobs in a distant area. Thus, whether the total job finding rate is lower for homeowners or for renters is just an empirical matter which depends on the magnitude of these two opposite effects.
The second explanation points out the need for a refinement in the definition of the residential status. On the one hand, one should distinguish between owners who have to comply with mortgage payments and outright owners, as housing financial commitments can bear higher pressure to return to work or to keep on with the current job. For mortgage-holders, these financial constraints can counteract the effect of the reduced mobility due to ownership. On the other hand, unemployment outcomes can be different also for private and social renters. In fact, below-market rent, long waiting lists, security of tenure and the restricted transferability within social housing, can harm the relative performance of social renters.
In this work, we aim at investigating the validity of the “Oswald thesis” taking into account the recent refinements by following both a theoretical and an empirical approach. The focus will be on UK micro data. In particular, we will make use of the two leading UK Surveys, the UK Quarterly Labour Force Survey and the British Household Panel Survey.
In the first chapter, we investigate the relation between job search effort of unemployed and housing tenure. We test this relation focusing on the impact of the UK Jobseeker’s Allowance reform introduced in the UK in October 1996, whose main aspect was a strengthening of search requirements for eligibility to the unemployment benefit. We revisit a simple model of search in which we introduce moving costs and housing costs to capture the two channels through which the degree of attachment to the accommodation influences search behaviour. Our theory suggests that a tightening in job search requirements, as implied by this reform, raises movements off benefit of non-employed with low search intensity and that this effect should adjust in size depending on the different housing tenure. We draw a dataset from the Labour Force Survey for the period 1995-1997, and by means of a Difference-in-Differences approach we analyze the impact of the reform on the claimant outflow. Average Treatment Effect estimates suggest that the impact of the reform is related to housing tenure. Specifically, renters account for a major portion of claimants who were crowded out of the benefit without finding a job, while the effect on outright owners and mortgagers is lower. Empirical evidence from our dataset clearly confirms that mortgagers search for a job more intensively than renters, as our model predicts. This latter finding is consistent with a higher estimated treatment effect for renters, since a high initial search intensity seems the key to insulate oneself from the impact of the tightening of search requirements.
In the second chapter we attempt to reconcile the empirical evidence with the argument in favour of a negative effect of homeownership on exit rates from unemployment. Taking into account some likely reasons for the falsification of the Oswald’s thesis, we provide evidence which supports it. At first, in a theoretical model of endogenous job search adapted to distinguish between local and non-local labour markets, we show that homeowners higher moving costs imply unambiguously lower search and lower job finding rates, even though an opposite effect works for jobs which do not require a move. Then, in the empirical analysis we make use of data drawn from the British Household Panel Survey to compare job search intensity measures by housing tenure. In defining the residential status, we distinguish between outright owners and mortgage-holders, and between social and private renters. We find that, controlling for housing costs and for the four-fold tenure definition, non-employed outright owners have definitely a lower attachment to the labour market than renters, and that this effect is even more evident when we compare them to private renters.
In the third chapter, we analyse the impact of the housing tenure on labour market outcomes using individual data from the UK labour Force Survey. We estimate both a binary model for the probability to be unemployed and a hazard model for exits out of unemployment. In both models
we test for endogeneity of housing tenure. In the binary model, exogeneity is rejected so we perform endogenous multinomial treatment effects estimates. In the hazard model, we find no evidence of unobserved heterogeneity thus estimates are performed assuming exogeneity. Results show that mortgagers have the lowest probability to be unemployed and the highest job finding rates, while social renters exhibit the worst performance. Whether private renters perform better than outright owners is a matter of debate: while we have no evidence in favour of this claim, the evidence in favour of the opposite is only modest.