Housing and Health

• Published date: 01 September 2019
• Author: Stefan Angel,Benjamin Bittschi
• Date: 01 September 2019
• DOI: http://doi.org/10.1111/roiw.12341

HOUSING AND HEALTH

by Stefan Angel

WU Vienna University of Economics and Business

and

Benjamin Bittschi*

Karlsruhe Institute of Technology

Deprived housing is recognized as a source of poor health, but there is still little evidence of a causal

relationship between housing and health. While existing literature identifies neighborhood effects and

the individual dwelling as factorswhich affect health, it does not offer a joint examination of these fac-

tors. Moreover, endogeneity is a concern in analyses of both problems. Thus far, studies addressing

endogeneity have done so through experimental design or instrumental variables. The first approach

suffers from problemsof external validity and the latter from the lack of reliable instruments.We there-

fore adopt an alternative strategy which considers both sourcesof endogeneity in order to identify the

effects of housing on health by estimating fixed-effect models. We reveal how housing problems affect

health depending on living conditions and socioeconomic status.Our results therefore indicate that liv-

ing in poor housing is an important short-term socioeconomic determinant that directly affects health.

JEL Codes: I14, I18, R20

Keywords: Europe, EU-SILC data, fixed effects, health, housing

1. Introduction

Housing is an important aspect of an individuals standard of living, and

poor housing conditions present a considerable challenge to social inclusion.

Housing costs also represent a significant share of total household expenditures.

In 2007, approximately 12.5 percent of the EU27 population lived in households

that spent 40 percent or more of their equivalized disposable income on housing

(Eurostat, 2010). To address these challenges, modern welfare states have devel-

oped a multitude of policies to mitigate the financial burden and the threat of

social exclusion posed by poor housing conditions. As well as limiting the cost of

living, these expenditures are also justified by their positive spillover effects on a

variety of desired socioeconomic outcomes.

Note: The authors would like to thank the editor, Conchita DAmbrosio, two anonymous refer-

ees, Melissa Engel, Friedrich Heinemann, Thomas Stratmann, Mustafa Yeter, and numerous semi-

nar and conference participants for helpful comments and discussions. Stefan Angel gratefully

acknowledges financial support by the Theodor K€

orner Fonds.

*Correspondence to: Benjamin Bittschi, KIT Karlsruhe, Chair of Public Finance and Public

Management, Kronenstrasse 34, 76133 Karlsruhe, Germany (benjamin.bittschi@kit.edu).

V

C2017 International Association for Research in Income and Wealth

1

DOI: 10.1111/roiw.12341

bs_bs_banner

Review of Inc ome and Wealth

Series 65, Numb er 3, September 2019

495

Indeed, many cross-sectional studies document the spillover effects of hous-

ing assistance programs on desirable outcomes such as improved health, fewer

behavioral problems, greater educational attainment, and increased labor force

participation. However, evidence for a causal link between poor housing condi-

tions and socioeconomic outcomes remains sparse (Newman, 2008). The main

problem is that it is difficult for empirical research using observational data to

accurately identify the effect of poor housing conditions on the respective out-

come. Individuals choose neighborhoods and dwellings for reasons that are diffi-

cult to measure. Consequently, empirical analyses suffer from selection bias if

they fail to consider individual characteristics that influence the outcome variable

and the sorting process into neighborhoods and dwellings (Ludwig et al., 2008).

Many authors devote substantial effort to identifying separate effects that may

jointly influence health and neighborhood choice (e.g. Katz et al., 2001; Ludwig

et al., 2001; Kling et al., 2007; Bilger and Carrieri, 2013), yet neglect them when

considering the choice of individual dwellings. Failing to control for all sources of

endogeneity may result in inconsistency of all parameters in the estimated

equations.

The aim of this paper is therefore to account for both dimensions of deprived

housing—that is, poor neighborhoods and poor dwellings—and to identify direct

effects of bad housing on health. We analyze four waves of EU-SILC (European

Union Survey on Income and Living Conditions, Eurostat User Database) panel

data (2005–8) for 21 European countries. EU-SILC data provide a massive sample

and thus greater test power compared to existing studies. The data allow us to use

time and person fixed effects to control for unobserved heterogeneity in the selec-

tion of dwellings and neighborhoods. The use of panel data also makes it possible

to address problems of potential simultaneity between health and socioeconomic

control variables, such as income, wealth, or education. Finally, having large-scale

panel data facilitates the identification of effects over a short time period, which

are expected to be potentially small in substantive terms. We can thus reveal how

housing problems affect health differently, depending on living conditions and

socioeconomic status. In summary, the overall aims of our research are to improve

the empirical estimation strategy which is applied to observational data in order

to reveal causal pathways between housing deprivation and health, and to offer

insights on the heterogeneous effects of deprived housing conditions on health.

2. Housing, Neighborhoods, and Health

The existing literature on the relationship between poor housing and health

can be classified into two strands: one concerned with the influence of neighbor-

hood effects on health and a second exploring how poor dwelling conditions and

insufficient basic facilities affect health. Unfortunately, there is little overlap, little

cross-referencing, and there are few integrated analyses of the two issues. In this

section, we will review both strands.

1

1

We concentrate on studies applyingmethods that are able to establish a causal link between bad

housing and health. Wealso concentrate on papers from developed countries, leavingaside the issue of

the lack of basic facilities in developing countries.

Review of Income and Wealth, Series 6 5, Number 3, Se pte mber 2019

V

C2017 International Association for Research in Income and Wealth

2

496