Housing allowances: still struggling to make ends meet

Abstract

Housing allowances aim at providing adequate and affordable housing. A theoretical discussion and literature review show why it is challenging for housing allowances to actually shield households from financial hardship. Using National French Housing Survey data, an original application with a probit model with a double sample selection (being a tenant, and being eligible for means-tested allowances) follows. Estimation results show that housing allowances help to cope with some life events but that otherwise their recipients remain more exposed to housing financial hardship than their counterparts do. The gap between recipients and non-recipients is larger for the households with children than for those without despite the goal of horizontal equity. A nonlinear decomposition shows that the difference in probability between recipients and non-recipients is mostly explained by the endowment of characteristics in 2001 and by the return to risk in 2013.

Keywords:

• Housing allowances
• housing affordability
• selection effect
• financial difficulties
• unpaid rent

Acknowledgments

I would like to thank the editor and the referees, Françis Calcoen and Stéphane Vigeant. I also thank Centre Maurice Halbwachs (CMH)-ADISP for providing the INSEE National Housing Survey.

Notes

1. http://europa.eu/economy_finance/db_indicators/tab/#.

2. In 2013, the variable rent arrears of at least two months in the last two years is not available anymore. But there is another variable about existing arrears at the moment of the survey. The risk factors for recipients are events A and C, having no education, being a lone parent or being young.

3. In 2013, event B does not jeopardize financial health (recipients or non-recipients of housing allowances).

4. In 2013, only lone parents are more at risk of financial difficulties than couples without children.

5. In 2013, the selection effect for receiving housing allowances is positive suggesting that for non-recipients the estimated probabilities of financial difficulties would have been lower than it is for recipients.

6. In 2013, the probabilities of financial difficulties have increased both for recipients and non-recipients but recipients remain slightly more adversely affected by those difficulties than non-recipients are. Expected probabilities are equal to 0.16 and 0.24 in 2013 vs. 0.12 and 0.19 in 2001 for a couple with secondary school vocational education who, respectively, did not suffer and who suffered from event A (see the second and third rows of Table 6).

7. It is assumed that these variables of location and the dummies indicating occupation status and perception of housing allowances four years before the year under study are not correlated with intrinsic household characteristics influencing financial difficulties four years later: ability to manage one's budget, giving priority to housing over other expenditures and probability of familial or professional shocks.