Economic theory
Behavioural responses
Tax and benefit reforms change incentives to work, earn and realise capital gains. PolicyEngine’s UK model can add labour supply responses, through substitution effects (changing work effort) and income effects (adjusting for changes in after-tax income), and applies a capital gains realisation response by default.
Substitution effect
When tax rates fall, each hour of work yields more after-tax income. This encourages people to substitute toward more labour and less leisure, increasing earnings.
Income effect
When after-tax income rises (e.g., from a new tax credit), people can afford to work less while maintaining their standard of living, slightly decreasing earnings.
Budget constraint: baseline vs. reform
Illustration with PolicyEngine’s central values: a substitution elasticity of 0.25 and an income elasticity of -0.05.
Labour supply elasticities
PolicyEngine’s UK model is static on labour supply unless a reform turns the response on by setting its two elasticities. When PolicyEngine turns it on, it uses the central values below. A reform then changes each person’s employment income by the substitution elasticity times the percentage change in their net-of-tax marginal wage (one minus their marginal tax rate), plus the income elasticity times the percentage change in their household’s net income. Each change is capped at 100%, and the response applies to employment income only.
CBO, "How the Supply of Labor Responds to Changes in Fiscal Policy" (2012); the substitution value is the CBO’s average over primary earners. Central values from PolicyEngine’s UK behavioural responses report.
policyengine-uk also includes the OBR’s labour supply model, whose elasticities vary by demographic group (substitution 0.094 to 0.439, income −0.185 to 0; OBR, February 2024). It currently models hours worked only and runs only when an analysis calls it directly; policyengine.py’s simulations use the two elasticities above.
Capital gains elasticity
A capital gains tax reform changes how much gain people choose to realise. PolicyEngine scales each person’s realised gains by ((1 − t₁) / (1 − t₀))ε, where t₀ and t₁ are their marginal tax rates on gains before and after the reform and ε is the elasticity with respect to the retention rate, the share of each pound of gain kept after tax. Gains qualifying for Business Asset Disposal Relief take the same elasticity, unless a separate one (1.4, the OBR’s) is switched on. Reforms that leave marginal rates on gains unchanged are unaffected; setting the elasticity to 0 gives a static costing.
Advani, Lonsdale and Summers (2024), "Reforming Capital Gains Tax", CenTax, pages 36-37; the central value is their medium-term (five-year) effect, and the sensitivities the range they consider.