![Sample our Health and Social Care journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5938/TOC-Subject-Sample-2021-Health-Social-Care.jpg"})
Abstract
In this paper, we focus on the tax deductibility of mortgage loans, which is a key measure of housing policies in many countries. Several studies have shown that the distribution of the benefits of this measure runs counter to the vertical equity principle. We contribute to the existing literature by addressing the distribution of the tax deductibility of mortgage loans considering the spatial outcomes of this policy. We use Belgian data to measure the impact of this policy on households' income and inequality and poverty in different types of areas. We show that this measure favours suburban households and subsidizes urban sprawl. Exogenous factors such as past housing policies and homeownership determinants play a key role in this phenomenon. We also simulate the distributional impact of three stylized reforms of the deductibility of mortgage loans in Belgium. These simulations show that the distributional outcomes could be improved on both social and spatial sides.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Vertical equity implies the idea of inequality reduction (Lambert, Citation2001), while horizontal equity deals with the process of redistribution more than with the redistribution itself (Verbist, Citation2002).
2 The homeownership bias is an example of horizontal inequity justified by efficiency reasons (see first paragraph); tax relief for dependent children is justified by equity arguments (Verbist & Van Lancker, Citation2016); differences in the ability to evade the tax system concern ‘unavoidability’ (Slemrod & Yitzhaki, Citation2002).
3 This also relates to the fact that housing is both a form of consumption and an asset, which have different consequences from the tax perspective. For a more thorough discussion on this complex issue, we refer to Mirrlees et al. (Citation2011). Regarding the homeownership bias in taxation as compared, Figari et al. (Citation2017), for example, show that in several countries owner-occupiers face lower average tax rates than tenants with the same ability-to-pay.
4 In principle, we could also include net imputed rent as taxable income in order to achieve horizontal equity. Bourassa and Grigsby (Citation2000), however, argue in favour of keeping net imputed rent untaxed. Moreover, quality issues with the imputed rent variable in our underlying data source EU-SILC (see Juntto and Reijo, Citation2010; Törmälehto and Sauli, Citation2013) are an additional (pragmatic) reason why we have not pursued this track.
5 Previous works dealing with the spatial distribution of the deduction of mortgage loans include Brady et al. (Citation2003), Gyourko and Sinai (Citation2003), and Sinai and Gyourko (Citation2004). They concern mainly the differences between administrative divisions of the U.S. territory (States, metropolitan areas…).
6 In 2013, the home mortgages deduction accounted for €2.59 billion of revenue foregone annually (0.7% of GDP) (FPS Finance, 2013).
7 Belgium is a federal state that has three administrative regions: Flanders, the northern, most populated and richest region of the territory; Wallonia, the southern region, and Brussels-Capital, which comprises the Brussels CBD (central business district) and eighteen surrounding urban municipalities.
8 For example, after World War I, grants were allowed to repair the damages but were restricted to the building of houses outside centres and separated by at least 10 meters. In 1935, the government set up a public housing company that facilitated access to ownership of separate houses with a vegetable garden and facilities for keeping small livestock. After WWII, building grants were once again allowed exclusively for single-family freehold houses.
9 Company cars refer to passenger cars that employers provide to employees for professional, commuting and private trips and for whose costs are supported by employers.
10 The average rate for the countries of the Eurozone is 66% (in 2015, according to Eurostat).
11 Significant differences can be noticed between the three Belgian regions. The ratio social housing candidates / social renters is 96% in Flanders, 39% in Wallonia and 108% in Brussels (Anfrie & Gobert, Citation2017; SLRB-BGHM, Citation2017; VMSW, Citation2016).
12 See Section 3.3 for the definition used to distinguish the poor from the non-poor population.
13 Antwerp: 52%, Ghent: 50%, Liege: 49%.
14 Among the total stock of passengers cars in Belgium (5,304,074 in 2012), 15% are company cars, 13.5% of workers benefiting from a company car in their salary package (Macharis & De Witte, Citation2012; May, Citation2017).
15 From 25% to 55%, plus an additional local tax.
16 For various reasons, credit institutions can accept or request other forms of security. In this case, the loans are not eligible for the tax relief.
17 See https://www.euromod.ac.uk/about/what-is-euromod for more information.
18 As a sensitivity check, we have recalculated the poverty line for each scenario. This approach produces very similar results, as these recalculated poverty lines do not differ significantly from the baseline. In order not to overburden the paper these results are not presented.
19 The private sector (banks and insurance companies) is the main provider of mortgage loans. Nevertheless, several public institutions provide also mortgage loans or public guarantee to households that may have difficulties in obtaining an (adequate) mortgage loan from the private sector (low-income workers, large families, etc.).
20 In a first part of their lifecycle, young households have to save money to meet the down-payment criteria imposed by credit institutions to obtain a mortgage loan. In addition, a relative stability in family and professional life may also be required before contracting a long-term loan. As for the elderly, most of them are outright homeowners, or too old to contract a mortgage.
21 From a more international point of view, the positive externalities of homeownership should indeed be balanced by the adverse consequences of this tenure status, e.g. in terms of residential and labour market mobility (see Isebaert et al., 2015, for the case of Belgium).
22 Small variations between income categories come only from the consideration of economies of scale in households in the computation of the allowance: small households are more represented in low-income deciles while families are more represented in high-income deciles, leading to higher and lower benefits per capita, respectively.
23 Except in the (unlikely) case of a perfect equivalence of the mortgage tax relief with the universal allowance.
24 To this aim, we use the INEQDEC0 module for Stata (Jenkins, Citation2015).