The surprising geography of the “mansion tax”


In brief:

  • Boris Johnson hinted that the 2020 budget will include a mansion tax. We’ve analysed Land Registry data to see which constituencies will be most impacted. 
  • Assuming only houses worth more than £1mn are considered “mansions”, the tax would impact the average Tory and Labour constituency roughly equally.
  • Lib Dem constituencies, concentrated in the South East, would be the most impacted. Not surprising, but noteworthy given they’ve historically been the biggest proponents of a mansion tax.
  • The constituencies which the Tories won from Labour in the 2019 general election would be relatively unaffected by a mansion tax.  It could be a play to defend recently-won Tory seats in the North.

In early February, Boris Johnson hinted he would like to introduce a “mansion tax”. That is, a tax imposed on people who own very high-value properties.  Following a backlash from Conservative MPs, Johnson has reportedly since cooled on the idea. But his government are yet to confirm whether the tax will or will not be part of the Budget on 11th March.

To understand why Johnson is/was interested in introducing a mansion tax, we’ve analysed Land Registry data to see which constituencies it would impact the most.  Our findings, summarised in the chart below, were as follows.

First, assuming only houses worth more than £1mn are considered “mansions”, the tax would impact the average Tory and Labour constituency roughly equally. The Tory base is much broader than country manors.

Second, Lib Dem constituencies, concentrated in the South East, would be the most impacted. Not surprising, but noteworthy given they’ve historically been the biggest proponents of a mansion tax.

Third, the constituencies which the Tories recently won from Labour would be relatively unaffected by a mansion tax.  Such a policy could be a play to defend recently-won Tory seats in the North.

Fourth, the choice of threshold will tell us much about the government.  A threshold of £1 million would suggest it wants to raise revenue and rebalance towards newly won Northern constituencies. Anything much higher – eg. £2 million – would have limited impact.


Methodology:

  • We download all house transactions in the UK since 1995 from the Land Registry.
  • Using ONS data, we then adjust the price paid in each transaction to 2019 prices – accounting for inflation. So a house costing £520,000 in 1995 would be £1,000,000 in 2019 prices.
  • We then map each house into its parliamentary constituency using the ONS’s postcode lookup dataset.
  • And then find which party won each constituency using data on the 2019 from the House of Commons library.
  • We implicitly assume the tax will be based on the price a house was most recently purchased for, adjusted for inflation.  If the government used estimated market value, they would capture more properties (but also people who are income-poor but have seen large house price rises).