What is the response of household expenditure to a change in property taxes? And does it depend on the household balance sheet position? While a large body of research (among others, Johnson et al. 2006, Parker et al. 2013, and Agarwal and Qian 2015) has made notable progress to quantify the effects of (income) tax rebates, little is known about the impact of (housing) tax hikes on consumer spending and whether some groups of society disproportionally bear the costs of any possible adjustment in standards of living following the tax change.
Property taxes and household behaviour: New evidence
In a recent working paper (Surico and Trezzi 2015) we try to fill this important gap in the literature by exploiting the 2011 changes in the Italian Imposta Municipale Unica (IMU tax). A newly appointed central government swiftly legislated and implemented a law which re-designed significantly the municipal system on property taxes, raising around 4 billion euros from taxes on the main dwelling and an additional 10.1 billion euros on other residential properties for a total revenue increase of 0.90% of GDP. The IMU affected 25.8 millions of tax payers (or around 70% of households) with an average contribution per tax-paying household around 357 euros on the main dwelling, and about 905 euros on other residential properties.
Using a new set of questions (on the amount of IMU paid) appositely added to the bi-yearly Survey of Household Income and Wealth (SHIW) conducted by the Bank of Italy, we identify the effects of property taxes on household behaviour. We do so by comparing the change in expenditure for IMU payers between 2010 and 2012 (i.e. before and after the tax reform) to the change in expenditure for non-IMU payers over the same period, netting out the effect of demographics, the change in house value, property characteristics, expectations on household income and on local house prices as well as regional fixed effects. In the most restrictive specification, we look only at home-owners and thus focus exclusively on variation in the amount of property taxes paid.
Our identification strategy builds upon four features of the 2011 changes in the municipal system of housing taxation in Italy. First, the national government introduced a new tax on the main dwelling and increased by an exogenous factor the (by then obsolete) land registry estimates of the rental values used to calculate the tax base for the main dwelling and other residential properties. Second, the timing and depth of the legislated changes were largely unanticipated. Third, municipal governments were allowed to unilaterally modify the rates proposed by Monti's government and the geographical variation in property tax rates appears unrelated to past local economic conditions. Fourth, the IMU tax changes were announced as an experiment (whose possible extension would have been subject to government revision) and most respondents in the 2012 wave of the SHIW did not expect the property tax changes to persist longer than five years.
The empirical analysis isolates five major empirical regularities.
- First, the marginal propensity to consume (MPC) non-durable goods and services out of the IMU tax is around 0.05 whereas the marginal propensity to spend (MPS) on durable goods is about 0.43.
- Second, these average effects mask pervasive heterogeneity across residential properties, with a large and significant marginal propensity to spend associated with the taxes paid on the main dwelling and a small and insignificant propensity to spend out of taxes on other residential properties.
In contrast, the marginal propensity to consume is always statistically indistinguishable from zero.
- Third, the significant response to the main dwelling IMU tax is far more pronounced among home-owners with mortgage debt and with low liquid wealth-to-income ratio.
- Fourth, debtors concentrated their expenditure cut on vehicles purchases while non-debtors responded to the increase in property taxes with a reduction of their savings.
- Fifth, the direct negative consequences of the changes in the IMU residential property taxes are estimated around 0.11% of GDP in 2012 vis-a-vis an increase in tax revenues of 0.90% of GDP (or 1.80% of government revenues).
On the other hand, the direct impact of the property taxes on the car industry in 2012 was large, representing a fall of around 14% relative to the market size in 2011.
Heterogeneous effects of the tax on consumer spending
Our study offers an unprecedented evaluation of the heterogeneous effects of residential property taxes on consumer spending using a large and unanticipated policy change which took place in Italy at the end of 2011.
Our analysis reveals that the taxes paid on the main dwelling triggered a large and very significant decline in household expenditure whereas the taxes paid on other residential properties caused a small and statistically insignificant change.
The adjustment was mostly borne by home-owners with mortgage debt and was concentrated on net vehicles purchases whereas households without debt responded with a proportional reduction of their savings. We conclude that while the short-run direct cost (in the form of forgone consumer spending) of the property tax changes for the Italian economy was small relative to the amount of extra taxes raised, the negative consequences for the car industry in 2012 were significant.
As for policy implications, our analysis contributes to two important debates about the design of housing taxes and the aggregate consequences of fiscal consolidation plans. More specifically, our evidence suggests that setting a multi-year plan of higher property tax rates for non-owner-occupied dwellings as well as providing owner-occupier mortgagors with property tax deductions based on their level of outstanding debt (as currently done in some OECD countries) can generate sizeable government revenues over a relatively short period of time, while minimising the contractionary effects that levying a property tax may otherwise induce. Furthermore, our analysis provides both one instance in which a housing tax appears highly recessionary (when borne by households with debt) and an instance in which the same tax does not seem recessionary at all (when borne by households without debt), suggesting that the policy decision of what specific group(s) of society to target could (and perhaps should) become another relevant dimension along which to evaluate the effectiveness and desirability of fiscal consolidation plans.
Agarwal, S and W Qian (2015), “Consumption and debt response to unanticipated income shocks: Evidence from a natural experiment in Singapore”, The American Economic Review (forthcoming).
Johnson, D S, J A Parker, and N S Souleles (2006), “Household Expenditure and the Income Tax Rebates of 2001”, The American Economic Review, 96(5):1589–1610, December.
Parker, J A , N S Souleles, D S Johnson, and R McClelland (2013), “Consumer Spending and the Economic Stimulus Payments of 2008”, The American Economic Review, 103(6):2530–53, October.
Surico, P and R Trezzi (2015), “Consumer spending and property taxes”, Finance and Economics Discussion Series 2015-057. Washington: Board of Governors of the Federal Reserve System