We evaluate the long-term effect of housing subsidies on the price, quantity and quality of private rental accommodation in France. The overall upward effect on rents is accompanied by an increase in the total number of dwellings. These effects depend on the market segment. They have not proved long-lasting for small dwellings, which have increased in quantity.
Source: 2013 and 2021 housing accounts, Ministry for the Ecological Transition and Regional Cohesion; 2021 national accounts, INSEE.
Housing subsidies are one of the main tools of housing policy in France and in many developed countries. Introduced in the post-war period, they have gradually been expanded, and in 2021 accounted for total expenditure of EUR 15.7 billion in France, or 0.6% of GDP (2021 housing account, Chart 1). This compares with 0.7% of GDP in Germany and Denmark in 2020, 0.9% in Finland and 1.4% in the United Kingdom (OECD, 2021). The so-called “Barre” reform in 1977 marked a first turning point in the expansion of housing subsidies, with a shift from the funding of construction in the social sector towards the provision of individual benefits, paid primarily to tenants. The main aim of the reform was to limit the share of housing expenditure in household budgets, with a focus on those on low incomes. The second objective, which was more liberal in nature, was to allow market mechanisms to play a greater role in housing by making tenants more solvent and enabling housing supply to adjust to their demand. At the start of the 1990s, the so-called “bouclage” reform extended the payment of subsidies to all low-income households, completing the conversion of this public policy into a redistributive tool.
Source: Annual consumer price index - Base 2015, for actual rents and all items, all French households, INSEE.
In theory, housing subsidies can lead to a rise in overall demand for housing and therefore in rents, at least in the short term, when supply elasticity is considered to be low. Under these conditions, the housing benefit will be partly captured by landlords. In the long term, rents, and more generally the equilibrium, depend on the elasticity of supply. The supply can increase extensively via a rise in the number of dwellings, or intensively via an increase in their quality. In practice, however, rental housing supply can remain partly inelastic.
Several concordant empirical studies conducted in different countries have already highlighted and measured the short-term inflationary impact on rents of housing subsidies targeting tenants, as for example Eriksen and Ross (2015) or Collinson and Ganong (2018) in the United States.
In France, where rents have risen faster than the consumer price index (Chart 2), Laferrère and Le Blanc (2004) and Fack (2006) find a positive effect of housing subsidies on rents in the 1990s. These two studies use the natural experiment provided by the reform of housing subsidies at the start of the 1990s. Laferrère and Le Blanc (2004) show that the significant impact of housing subsidies on rents is only slightly explained by an increase in dwelling quality. Fack (2006) shows that the reform led to a rise in rents that represented 78% of the amount of the subsidies, with no observed improvement in quality.
Our recent study measures the long-term impact of housing subsidies on rents, and on the quantity and quality of private rental dwellings in France. It thus provides an insight into the elasticity of the supply of rental housing.
To measure the long-term impact of housing benefit, we compare similar urban agglomerations receiving higher or lower subsidies since the reform of housing subsidies in the 1990s.
We find no significant impact of housing subsidies on rents in the 1980s, when the expenditure on housing benefit was lower. In contrast, we show that they led to a rise in rents in the two decades following the reform (from 1995 to 2016), with a stronger impact in the short run (in the 1995 to 1999 period).
Between 2000 and 2016, housing subsidies had an overall positive impact on rents, even for tenants that were not subsidised. The impact for non-claimants is large, as in the United States (Susin, 2002). In France, given that 40% of tenants in the private sector are subsidised, a landlord can take housing subsidies into account when setting the rent, before knowing whether or not the tenant is subsidised.
The positive impact on rents holds with a constant magnitude when we divide the study period (2000-2016) into two sub-periods (2000-2008 and 2009-2016). This inflationary impact is accompanied by an increase in the quantity of private rental housing; no effect on quality is detected (Table 1).
Note: The data only enable the effect on quality to be measured in 1999, 2006, and 2016.
We show that this long-term inflationary impact is accompanied by different reactions in the housing market, depending on the market segment. For dwellings with three or more rooms, the rental housing supply remained inelastic in quality and quantity and higher housing subsidies led to a lasting increase in rents between 2000 and 2016. In contrast, for one or two-room dwellings, rents stopped increasing significantly between 2000 and 2016 (after rising between 1995 and 1999), and the quantity of private one-room rentals increased in 1999, 2006 and 2016, driven by new buildings.
Our finding could be due to the entry of a greater number of students into housing markets where subsidies are higher. For small dwellings, the housing market responded to this increase in demand with an increase in supply in terms of quantity. The supply of one-room rentals is probably the most elastic, as landlords who opt for rental investment buy small dwellings and those who are wealthier may prefer to diversify their risks by purchasing several small dwellings.
Any reform of housing subsidies must take account of the hybrid nature of this public policy, which is both a tool to make housing demand more solvent, and a means of redistributing income towards the poor. For example, a cut in housing subsidies could lead to lower rents, as observed in the United Kingdom (Braakmann and MacDonald, 2020). However, the fall may not necessarily be as large as the rise, or even the same size as the subsidy cut (Gibbons and Manning, 2006), which, in this latter case, could lead to a fall in beneficiaries’ living standards.
Several reports in recent years have suggested merging housing subsidies with other benefits, in the form of a negative income tax or a single benefit payment (Bozio et al. (2017), Trannoy and Wasmer (2013)). These proposals are based on the idea that the inflationary impact on rents is due to the identification of housing subsidies as a benefit to be devoted entirely to housing and which can therefore finance a rise in rents. While merging housing subsidies into a single benefit could reduce their inflationary impact, there could still be an upward effect on prices if the number or quality of rental dwellings on offer fail to increase. Indeed, our study confirms that the greater the inelasticity of housing supply, the greater the increase in rents. This underlines the importance of increasing the number of dwellings on offer at the same time as making demand more solvent, to avoid the subsidies being captured by landlords.