Transfer taxes and household mobility: Distortion on the housing or labour market?

Abstract

We estimate the effect of the UK Stamp Duty Land Tax on household mobility using micro data. The UK (until 2014) provided an ideal setting to explore the impact of housing transfer taxes on mobility decisions. This is partly because the stamp duty liability is quite substantial, at least for more expensive housing (the top rate is currently 7 percent of the purchase price), and partly because the stamp duty liability, until December 3 of 2014, jumped sharply at various cut-off values, providing various ?discontinuities? that can be exploited empirically. Our analysis focuses on a discontinuity where the stamp duty jumps particularly strongly from 1 percent to 3 percent of the purchase price. This discontinuity allows us to isolate the impact of the stamp duty from other determinants of mobility. In our core analysis we use data from the British Household Panel Survey (BHPS) and compare homeowners with self-assessed house values on either side of the cut-off, while controlling for flexible but smooth functions of house values. We find that the stamp duty has a significant negative effect on household mobility and that this effect is confined to short-distance moves and to moves that are housing- rather than job-related. Our core estimates indicate that the 2 percentage-point increase in the stamp duty reduces the annual rate of mobility by between 2 and 3 percentage points. This is a very substantive effect given that in the UK about 5 percent of all owner-occupier households move each year. To further assess the validity and quantitative significance of the response of households to the stamp duty, we turn our attention to a different dataset from the Land Registry that consists of actual transaction prices. Analysing the distribution of transaction prices of all housing sales in England and Wales, we find additional evidence of a strong behavioural response. We document bunching of observed transaction prices at the cut-offs where the tax rate increases and, consistent with the results of our core analysis, we find that a 2 percentage point increase in the tax rate decreases the volume of sales by roughly 30 percent. The contribution of our study to the existing literature is twofold. Firstly, we identify the long-term (equilibrium) effects of the stamp duty on actual household mobility. Secondly, we are able to distinguish between different types of moves. In particular, our analysis distinguishes between short- and long-distance moves and between housing- and job-related moves. This paper is to our knowledge the first quasi-experimental study that directly evaluates the effect of a real estate transfer tax on actual household mobility.