Stamp Duty Land Tax

In October 2017 the OBR published semi-elasticity estimates for residential SDLT, derived from HMRC's analysis of the December 2014 reform.⁵ Each gives the percentage change in transactions per 1 percentage point change in the effective tax rate; roughly twice as large as previously assumed:

A semi-elasticity of −6.0 means a 1pp rate cut increases transactions by 6.0%. We use steady-state rather than Year 1 estimates; the latter are larger (up to −7.0 for cheaper properties) because retiming inflates the short-run response.

Every transaction suppressed by SDLT is a mutually beneficial exchange that does not occur. The welfare cost per suppressed transaction averages half the SDLT that would have been paid; the standard Harberger triangle.⁶ Because deadweight loss scales with the square of the effective rate, high-rate bands are disproportionately damaging per pound of revenue raised.

Our model overstates raw revenue because it applies the standard schedule to all buyers, ignoring first-time buyer relief. Revenue and DWL figures below are scaled proportionally so that total revenue matches HMRC's published £6.7bn standard-rate benchmark.

We apply the OBR's steady-state semi-elasticities to Land Registry transaction volumes for England. Both scenarios leave surcharges unchanged; the revenue effects relate only to the standard buyer schedule.

Annual transaction counts are from HM Land Registry Price Paid data for calendar year 2025, English sales only (864,313 transactions; Welsh postcodes excluded).

Predicted responses range from zero (properties below the nil-rate band) to over 50% at the top; figures above £2m extrapolate well beyond the original data and should be treated as indicative.

A nil-rate band to £250,000 and flat 5% above eliminates SDLT for buyers under £250k and delivers large reductions above £925k where the 10% and 12% bands are removed. This targets the most distortionary rates while retaining a broad revenue base.

One narrow way to put a number on the labour-market gain counts only long-distance job moves. We set it out because it is the calculation an official costing reaches for, and then explain why it understates. Assume 21% of the additional transactions are moves of 50 km or more by economically active households, and that such movers gain about 9% in earnings from a better job match.⁷ Applied to our transaction model, that gives 16,032 long-distance movers a year under full abolition.

Both parameters sit at the generous end of the survey evidence. Most moves are short and driven by household needs: the English Housing Survey puts moves of 50 miles or more at 7% of younger owner-occupier movers and 21% of older ones, and the wage-growth premium to migration in British panel data is largest for those who move for work.⁷ We apply the share uniformly, though it may be higher for expensive properties where job relocation is more common.

The standard approach applies the 9% gain to median full-time earnings of £39,039, producing ~0.002% of GDP per year.⁷ But the largest SDLT reductions fall on expensive properties whose buyers earn well above the median; using band-specific earnings produces a materially larger estimate.⁸

Each additional transaction generates spending on estate agents, solicitors, surveyors, removal firms, and mortgage brokers. Agent fees (1.2% of sale price) scale with property value; other costs (conveyancing, survey, removals, mortgage, searches) are modelled as a flat £6,000 per transaction, mostly VAT-able at 20%.

Agent fees dominate at higher price bands; ancillary costs dominate at lower bands.

Limitations

• Extrapolation. The semi-elasticities were estimated from the 2014 reform, which changed effective rates by 0–3pp. Full abolition implies reductions exceeding 10pp at the top of the market (effective rates approach the 12% top marginal rate); the predicted 50%+ transaction increase for the most expensive properties extrapolates well beyond the original data.
• Linearity. Semi-elasticities are local linear approximations. For large rate changes the true response is unknown; it may diminish (fewer locked-in transactions to unlock) or increase (credit constraints relaxed at lower rates). We apply the response as a proportional change, 1 + the semi-elasticity × the cut in points. A constant semi-elasticity strictly compounds, and the difference is large at the top: under abolition the exponential form gives £5m+ sales a 91% rise rather than 66%, and under the 5% cap it trims the revenue loss by about 2%. Our figures are the more cautious ones.
• Band means. Each band's response is computed at its mean price, and the effective rate is not linear in price. Recomputing sale by sale over all 864,313 transactions changes the abolition response from 8.8% to 8.5% and the cap's revenue loss by under 1%, because the bands are narrow and the schedule is piecewise linear within them.
• First-time buyer relief. FTBs below £500k benefit from a £300k nil-rate band. Revenue figures are scaled to HMRC's published £6.7bn benchmark accordingly.
• Price capitalisation. The OBR estimates price semi-elasticities of −1.5 to −2.0: a 1pp rate cut raises prices by 1.5–2.0%. Part of the buyer's saving is captured by sellers; a transfer, not an efficiency loss, but buyers do not retain the full benefit.
• Three-band resolution. The OBR provides only three semi-elasticity bands. The same value (−6.0) applies to a £1.1m house and a £10m property, which almost certainly have different buyer behaviour.
• Credit constraints. SDLT reduces deposit funds for marginal buyers near LTV limits. Abolition could unlock currently unaffordable purchases; a channel not captured by the semi-elasticities.
• Selection. The 9% earnings gain is an average for long-distance movers. Marginal movers; those just deterred by SDLT; may gain less from relocation, overstating the productivity channel.

Wider economic effects

A full cost-benefit analysis would also account for distributional consequences of replacement revenue, house price capitalisation,⁵ and dynamic fiscal offsets from higher transaction volumes (additional VAT, capital gains tax, and professional fees). Muellbauer's background paper to the 2025 Keynes Lecture sets out the wider channels: commuting, business location, investment pulled away from productive sectors when land prices rise, and a package of planning, land value capture and social housing whose parts reinforce each other.¹²