6(27)
The present study contributes to this research by empirically investigating
the effects of housing allowance
5
on housing consumption by exploiting a
feature of the 1997 Swedish housing allowance reform that arguably comes
close to being a natural experiment. This study uses the dwelling size
constraint to identify any effect of housing allowance on housing
consumption. This policy reform of the Swedish housing allowance system
arguably produced exogenous variation in housing consumption. Since the
constraint on dwelling size applies to households living in premises of
certain living areas, the effect of reform exposure can be estimated using a
difference-in-difference (DD) strategy. The register-based dataset
encompasses all housing allowance recipients in Sweden in 1996 and 1997.
However, for the DD strategy to be convincing, the treatment and the
control group should be as similar as possible. Therefore, this study is
limited to recipients who are single parents living in rental apartments—a
group of households that is overrepresented in the Swedish housing
allowance system.
Previous studies have found varying degrees of positive effects on housing
consumption accruing from housing allowances. Existing research into
housing allowances draws primarily on the large social experiments that
have taken place in the USA, for example, the US Experimental Housing
Allowance Program (EHAP) and the Housing Voucher and Housing
Certificate programs, which indeed generated a substantial literature on
housing allowances and their effects on housing markets and low-income
households. The first part of EHAP was designed to predict household
responses to housing allowances, i.e., the “demand experiment” (see, e.g.,
Bradbury and Downs, 1981; Department of Housing and Urban
Development, 1980; Mulford et al., 1980; Rosen, 1985). In this
experiment, randomly sampled households were granted housing
allowances and their behavior was compared with that of a control group
without housing allowances. The conclusion was that the housing allowance
did induce households to live in housing that met the minimum standard,
but that the allowance payments had to be tied to the housing standard for
this to be successful.
The Housing Certificate program (see, e.g., Leger and Kennedy, 1990a,b)
determined the amount that a household had to pay out of its own
resources and then made up the difference between this amount and the
gross rent. However, recipient households were not permitted to rent units
with rents exceeding a certain limit. The Housing Voucher Program, in
contrast, used a locally determined payment standard equal to the fair
market rent. The Housing Voucher payments then covered the difference
between this payment standard and 30% of the recipient household’s net
income, regardless of the rent paid for the unit actually chosen by the
family. The results of these two programs indicated that participants in the
Housing Certificate program rented units near the limits allowed by the
program, while Housing Voucher recipients on average selected units with
slightly higher gross rents and also obtained larger units than did Housing
Certificate recipients. However, Susin (2002) showed that low-income
households in metropolitan areas with more vouchers instead experienced
faster rent increases than those areas where vouchers were less abundant.
Almost all countries in Europe as well as the USA and Canada have some
kind of housing allowance system—all welfare states have a degree of
5
Two main housing allowance schemes exist in Sweden: 1) housing allowances for families with
children and for young people (under age 29) without children, and 2) a housing supplement for
pensioners. This study is limited to examining the housing allowance for families with children.
