Skip to content

Keep the inputs VAT is computed from non-negative - #572

Open
MaxGhenis wants to merge 3 commits into
mainfrom
fix/non-negative-imputed-inputs
Open

Keep the inputs VAT is computed from non-negative#572
MaxGhenis wants to merge 3 commits into
mainfrom
fix/non-negative-imputed-inputs

Conversation

@MaxGhenis

@MaxGhenis MaxGhenis commented Oct 12, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Fixes #565.

Status: ready for review. Approved by Sol at this head (below), and CI passes. uk-data merges only in a batched release (d833), so this PR is not to be merged on its own.

What is wrong on 1.58.0

Measured on the released file (enhanced_frs_2024_25.h5, SHA-256 5ba399ea5ab9a6c7d4e9aaf6c623080beb34d9ad5f330f6244dbad72b8d960b6) with policyengine-uk 2.125.1 and policyengine-core 3.33.0:

  • 76 households have negative VAT in every year from 2024 to 2030. The lowest is −£7,091 in 2024 and −£7,501 in 2026. Together they are 32,000 weighted households and −£67m in 2024.
  • The cause is in this repo's inputs, not in policyengine-uk. Its formula is vat = (consumption × full_rate_vat_expenditure_rate × standard rate + consumption × reduced-rate share × reduced rate) / microdata_vat_coverage (vat.py), where consumption is the sum of twelve COICOP inputs. The formula needs no floor if its inputs are non-negative.
  • 44 households have negative total consumption. Each stores a negative housing_water_and_electricity_consumption that outweighs its other eleven categories. In all, 94 households store a negative value there; the lowest is −£5,498.
  • 32 households store a negative full_rate_vat_expenditure_rate (the lowest is −1.02).
  • The two groups do not overlap, and no negative-VAT household is outside them.
  • The same rate has an upper-tail fault: 2 households store a rate above 1, one of them 55.45. The rate is a share of expenditure, so it should be at most 1.

Why

Both imputations are quantile regression forests. They sample conditional quantiles of their training values, so their predictions stay within the range of those values. A negative training value can therefore reach the dataset.

  • Consumption (consumption.py, generate_lcfs_table):
    • LCFS 2023-24 records a negative weekly amount for housing, water and electricity (P604) in 9 of 4,202 households. That is 0.2% by weight, and the lowest is −£265 a week.
    • None of the other eleven categories is negative.
    • The table went to the model uncleaned, although the same function already clamps its derived gas and electricity at zero.
  • VAT rate (vat.py, generate_etb_table):
    • The training target is (totvat × (1 − reduced-rate share) / standard rate) / (expdis − totvat).
    • Of 4,199 ETB training rows, 4 have expdis − totvat ≤ 0, which gives a negative rate (down to −3.4).
    • 13 rows have a rate above 1. The highest is 55.45, in a household with under £1 a week of expenditure after VAT.

What this PR changes

  • The consumption training table clamps every imputed amount at zero. The weighted mean of the housing category rises 0.17%.
  • The VAT training table keeps the rate a share of expenditure:
    • a row with no expenditure left after VAT is dropped (4 rows);
    • a rate above 1 is capped at 1 (13 rows).
    • The mean training rate falls from 0.429 to 0.417, almost all of that from the one row at 55.
  • impute_consumption and impute_vat fail the build if an imputed amount is negative or a rate falls outside [0, 1].
  • Each model's metadata gains a key, so a cached model trained on the old tables is retrained.

Impact

Source: this PR's CI (run 40190348100, head 9fd9164e). It builds main and this branch the same way: TESTING settings (32 epochs), one OA clone, and the head's lockfile. It then measures both builds with policyengine-uk 2.122.2, with a full microsimulation for each year on each side. All figures are £bn.

2025-26 2026-27 2027-28 2028-29 2029-30 2030-31
VAT, base 324.14 332.85 340.77 348.97 357.52 366.31
VAT, this PR 320.46 329.08 336.90 345.01 353.46 362.15
Change −3.68 (−1.13%) −3.78 −3.86 −3.96 −4.05 −4.15
Income tax, change +0.66 (+0.22%) +0.69 +0.71 +0.73 +0.73 +0.73
CGT, change −0.08 (−0.33%) −0.09 −0.09 −0.09 −0.10 −0.10
Household net income, change +2.61 (+0.16%) +2.73 +2.86 +2.93 +3.05 +3.03
Cost of a 2pp standard-rate rise 31.96 → 31.60 32.82 → 32.45 33.60 → 33.22 34.41 → 34.02 35.26 → 34.85 36.12 → 35.71
  • Negative VAT is gone.
    • On the CI head build, no household has negative VAT in 2025 or 2030, and every VAT input is in range: test_built_vat_is_non_negative_for_every_household and test_built_vat_inputs_are_non_negative pass, out of 2,674 tests passed.
    • The aggregate tool's own sign:vat check fails on the base build and passes on the head. That is the comparison's one "fixed" check; none is broken.
  • The VAT fall is far larger than the negative VAT removed (−£67m on 1.58.0). Both models are retrained on the cleaned tables, so every household's imputed consumption and rate change, not only the 76 households'. The training-table mean rate falls 2.8%, and built VAT falls 1.13%.
  • Noise:
    • The CI calibration is unseeded (dropout_weights draws torch.rand_like), so the two builds also differ by calibration noise.
    • For a reference, take Widen CGT taxpayer count bound for the 1.59.1 release #631's CI. It changed only a test, so its base and head ran the same code. Its 2025 differences were CGT +0.41%, VAT −0.07%, income tax +0.01% and household net income +0.01%.
    • Against that one pair, the VAT move is signal, and the CGT move is within noise.
    • The income tax and household net income moves are 16 to 22 times that pair's. They come from changed weights, not from VAT itself. In policyengine-uk 2.122.2, household_tax adds vat_change, not vat, so baseline VAT does not enter household net income.
    • The tool reports CGT, not capital gains. This PR does not change the gains imputation.
  • Calibration fit against the base build: pass. Targets within 10% are national 64.5% → 64.6% and local 75.0% → 75.1%. obr/vat is in the same error band on both builds.
  • Income distribution and single ages against the base build: pass.

Invariants

  • Every imputed consumption and spending amount is non-negative, both in the training table (Hypothesis) and in the built dataset.
  • The full-rate VAT expenditure rate is in [0, 1], both in the training table (Hypothesis) and in the built dataset.
  • A training value that was already valid is unchanged (Hypothesis).
  • impute_consumption raises if its output has a negative amount. A test runs the real function with a mocked model, and it fails when the guard call is removed.
  • VAT is non-negative for every household of the built dataset, in the calibration year and in 2030, under policyengine_uk.Microsimulation. The test has no reduced-build slack, because the property belongs to the inputs, not to the calibration.

Reviews

  • GPT-6.1 Sol, round 1, at d14e0c7b: APPROVE WITH NITS. It found two P3 nits, both fixed in 6be52fe8:
    • the guard's call site had no test;
    • the comments overstated how exactly a QRF reproduces training values.
  • GPT-6.1 Sol, delta review, at 9fd9164e: APPROVE, with no findings.
    • It ran the 12 related test files plus the workflow-artifact, fit-report and distribution tests: 196 passed, 19 skipped.
    • Its own guard-removal mutation gave the one expected failure.

Related

  • Microcosm, which the Budget runs on, already has both fixes in its own pipeline:
    • microcosm#1063 floors negative LCFS consumption on the donor;
    • microcosm#1225 drops ETB rows with no expenditure after VAT and caps the rate.
  • 16 households store a negative num_vehicles on 1.58.0 (the lowest is −8). That comes from a different imputation, the WAS wealth imputation, and does not reach VAT: 16 households have a negative vehicle count on 1.58.0 (WAS missing-value code in the wealth imputation) #574.
  • VAT in total runs well above the OBR's receipts target with the 0.38 microdata_vat_coverage that this branch's locked policyengine-uk 2.122.2 uses. policyengine-uk#2187 (0.469, first released in 2.135.7) addresses that, not this PR.

axiom: n/a: data

🤖 Generated with Claude Code

@MaxGhenis @claude
Keep the inputs VAT is computed from non-negative
d14e0c7 
On the 1.58.0 enhanced FRS, 76 households have negative VAT (uk-data#565):
44 because an imputed housing, water and electricity amount is negative
and outweighs their other consumption, and 32 because their imputed
full-rate VAT expenditure rate is negative. Two more households have a
rate above 1, one of 55.

Both imputations are QRFs, which predict values they were trained on. Nine
LCFS households record a negative P604. Four ETB households report
expenditure at or below their VAT, and thirteen a ratio above 1.

- Clamp the consumption training table's amounts at zero.
- Keep the VAT training rate a share of expenditure: drop rows with no
  expenditure left after VAT, and cap the rest at 1.
- Fail the build if an imputed amount is negative or a rate leaves [0, 1].
- Mark both models' metadata so cached models from the old tables are
  retrained.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@MaxGhenis MaxGhenis mentioned this pull request Oct 12, 2026
Open
This was referenced Oct 12, 2026
Closed
Open
MaxGhenis and others added 2 commits October 25, 2026 07:44
@MaxGhenis @claude
Test the consumption guard's call and describe QRF sampling accurately
6be52fe 
Round 1 review nits (Sol, APPROVE WITH NITS at d14e0c7): impute_consumption
now has a test that fails if its non-negativity guard is disconnected, and the
comments say the QRF samples conditional quantiles within the training range
rather than reproducing training values.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@MaxGhenis @claude
Merge origin/main into fix/non-negative-imputed-inputs
9fd9164 
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 25, 2026

Copy link
Copy Markdown

Calibration fit against the base build: pass

Targets within 10% (those both builds have) Base Head Change
National (647) 64.5% 64.6% +0.2pp
Local (11,228) 75.0% 75.1% +0.0pp

A drop of more than 2.0% fails.

Notes

  • 12 national targets are in a better band on the head.

National targets in a worse error band on the head (10)

Target Base Head
dwp/uc/claimants_with_5+_children within 10% 25-50%
dwp/uc_payment_dist/COUPLE_NO_CHILDREN_annual_payment_4_800_to_6_000 within 10% 10-25%
dwp/uc_payment_dist/LONE_PARENT_annual_payment_20_400_to_21_600 within 10% 10-25%
dwp/uc_payment_dist/SINGLE_annual_payment_2_400_to_3_600 10-25% 25-50%
hmrc/dividend_income_count_income_band_50_000_to_70_000 10-25% 25-50%
hmrc/private_pension_income_count_income_band_150_000_to_200_000 within 10% 10-25%
hmrc/private_pension_income_income_band_300_000_to_500_000 25-50% 50-100%
hmrc/property_income_income_band_15_000_to_20_000 within 10% 10-25%
hmrc/self_employment_income_count_income_band_500_000_to_1_000_000 within 10% 10-25%
ons/north_east_age_10_19 within 10% 10-25%

Flags are for review: builds of the same code move 2 to 12 national targets to a worse band (#598). Bands are of |estimate / target - 1|; no target's estimate or error is published.

Workflow run

Income distribution and single ages against the base build: pass

HBAI FYE 2025 (DWP), period 2024: equivalised income in £ a week, rates and Gini in per cent. A value outside HBAI's 95% interval is marked (outside).

Measure Base Head HBAI 95% interval
median_bhc 678.42 (outside) 678.76 (outside) 719.48 703.57 to 735.58
median_ahc 584.40 (outside) 584.42 (outside) 623.14 605.21 to 637.07
mean_bhc 800.21 (outside) 802.04 (outside) 854.98 836.37 to 870.35
mean_ahc 691.60 (outside) 693.30 (outside) 749.82 732.33 to 764.50
gini_bhc 33.14 33.20 32.74 31.89 to 33.98
gini_ahc 38.44 (outside) 38.46 (outside) 36.78 35.90 to 38.29
relative_low_income_bhc_all 18.20 (outside) 17.95 (outside) 15.91 14.45 to 17.64
relative_low_income_bhc_children 23.87 23.22 20.52 17.36 to 23.96
relative_low_income_bhc_working_age 17.47 (outside) 17.23 (outside) 14.28 12.91 to 16.02
relative_low_income_bhc_pensioners 13.77 14.01 15.90 13.69 to 17.99
relative_low_income_ahc_all 23.47 (outside) 23.13 (outside) 19.64 17.99 to 21.27
relative_low_income_ahc_children 33.52 (outside) 32.58 (outside) 27.41 23.49 to 30.64
relative_low_income_ahc_working_age 23.16 (outside) 22.86 (outside) 18.58 17.01 to 20.34
relative_low_income_ahc_pensioners 12.06 12.36 13.87 11.62 to 15.90
Single ages against the ONS projection Base Head
Mean error, ages 0 to 79 11.0% 11.0%
Largest error, ages 0 to 79 43.8% 44.4%
Largest error, ages 15 to 19 36.9% 37.7%

Flags

None.

These measures report and flag but do not fail yet: their same-code spread is unmeasured (#598).

Aggregate impact of the data change: pass

Base dataset enhanced_frs_2024_25.h5 (the base's build, sha256 c50ac2b57ad3) against head dataset enhanced_frs_2024_25.h5 (this pull request's build, sha256 4f5f89b2c7f8), both measured with policyengine-uk 2.122.2, fiscal years 2025-26 to 2031-32. Each year is a full microsimulation run on each side.

Checks: 0 broken by this change, 3 failing on both sides, 1 fixed. Moves flagged for review: 188 (thresholds in report_spec.yaml; a flag does not fail the job).

Workflow run and full report

Notes

  • In the years a figure from fewer than 10 records is suppressed (spending.dfe_education_spending, spending.disabled_students_allowance, spending.sda), every total containing it, and the decile incomes, are rebuilt from their other components, so no published total includes it. Those totals are marked †. A change is withheld where the two sides leave out different figures and can't be lined up.

Fixed by this change

  • VAT is never negative (sign:vat): 2025-26: 65 units below zero, a weighted £-82,810,728 in total; 2026-27: 65 units below zero, a weighted £-85,037,298 in total; 2027-28: 65 units below zero, a weighted £-87,058,971 in total

Moves flagged for review

Measure Year Base Head Change
VAT 2025-26 £324.14bn £320.46bn -£3.68bn
VAT 2031-32 £375.24bn £370.99bn -£4.26bn
VAT 2028-29 £348.97bn £345.01bn -£3.96bn
VAT 2027-28 £340.77bn £336.90bn -£3.86bn
VAT 2030-31 £366.31bn £362.15bn -£4.15bn
VAT 2029-30 £357.52bn £353.46bn -£4.05bn
VAT 2026-27 £332.85bn £329.08bn -£3.78bn
Government balance, each receipt once 2025-26 £432.62bn £429.03bn -£3.60bn
gov_balance 2025-26 £435.75bn £432.16bn -£3.60bn
Government balance, each receipt once 2026-27 £459.94bn £456.24bn -£3.70bn
gov_balance 2026-27 £463.07bn £459.37bn -£3.70bn
Government balance, each receipt once 2027-28 £490.45bn £486.61bn -£3.84bn
gov_balance 2027-28 £493.58bn £489.74bn -£3.84bn
Government balance, each receipt once 2028-29 £518.26bn £514.31bn -£3.95bn
gov_balance 2028-29 £521.39bn £517.44bn -£3.95bn
Government balance, each receipt once 2029-30 £547.98bn £543.89bn -£4.09bn
gov_balance 2029-30 £551.11bn £547.02bn -£4.09bn
Government balance, each receipt once 2030-31 £581.31bn £577.23bn -£4.08bn
gov_balance 2030-31 £584.44bn £580.36bn -£4.08bn
Government balance, each receipt once 2031-32 £608.57bn £604.34bn -£4.22bn
gov_balance 2031-32 £611.70bn £607.47bn -£4.22bn
Extended childcare entitlement 2030-31 837k 804k -32.8k
Disability Living Allowance 2031-32 1,404k 1,365k -38.3k
Disability Living Allowance 2028-29 1,385k 1,348k -37.8k
Disability Living Allowance 2029-30 1,391k 1,353k -38.0k
Disability Living Allowance 2030-31 1,398k 1,360k -38.1k
Disability Living Allowance 2025-26 1,370k 1,332k -37.4k
Disability Living Allowance 2027-28 1,380k 1,342k -37.7k
Disability Living Allowance 2026-27 1,375k 1,337k -37.5k
Personal Independence Payment 2031-32 £35.43bn £35.85bn +£420.9m
Personal Independence Payment 2030-31 £34.58bn £34.99bn +£410.9m
Personal Independence Payment 2029-30 £33.75bn £34.15bn +£401.0m
Stamp duty land tax (own purchases, and corporate until the 2031-32 £25.98bn £25.58bn -£400.6m
Personal Independence Payment 2028-29 £32.95bn £33.34bn +£391.4m
Stamp duty land tax (own purchases, and corporate until the 2030-31 £24.78bn £24.40bn -£383.3m
Personal Independence Payment 2027-28 £32.17bn £32.55bn +£382.2m
Personal Independence Payment 2026-27 £31.33bn £31.70bn +£372.2m
Stamp duty land tax (own purchases, and corporate until the 2029-30 £23.66bn £23.29bn -£366.3m
Carer's Allowance 2031-32 924k 942k +17.9k
Personal Independence Payment 2025-26 £30.07bn £30.42bn +£357.2m

148 more flags are in the report artifact.

Receipts by tax head

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Income tax +£661.2m +£690.8m +£706.5m +£728.6m +£726.7m +£733.4m +£756.7m
National Insurance (employee and self-employed) +£69.3m +£70.0m +£72.7m +£73.4m +£77.8m +£78.3m +£80.4m
National Insurance (employer) +£273.0m +£283.6m +£291.6m +£299.0m +£307.2m +£316.1m +£328.5m
Capital gains tax -£83.4m -£86.0m -£88.9m -£92.1m -£95.2m -£98.6m -£102.1m
Student loan repayments +£10.4m +£10.8m +£11.1m +£11.3m +£11.6m +£12.0m +£22.6m
VAT -£3.68bn -£3.78bn -£3.86bn -£3.96bn -£4.05bn -£4.15bn -£4.26bn
Fuel duties +£113 -£634 +£229 +£453 -£548 +£239 -£253
TV licence fee -£4.8m -£6.0m -£5.3m -£5.2m -£5.4m -£5.5m -£5.2m
Council tax +£18.8m +£19.9m +£21.1m +£22.5m +£23.9m +£25.3m +£25.5m
High value council tax surcharge 0 0 0 -£4.4m -£4.5m -£4.6m -£4.7m
Northern Ireland domestic rates -£1.6m -£1.7m -£1.7m -£1.8m -£1.9m -£2.0m -£2.0m
Business rates +£676 +£4,426 -£3,264 -£5,284 -£3,441 +£1,803 -£2,060
Stamp duty land tax (own purchases, and corporate until the -£301.9m -£318.2m -£334.3m -£350.1m -£366.3m -£383.3m -£400.6m
Land transaction tax (Wales) -£0.2m -£0.2m -£0.2m -£0.2m -£0.3m -£0.3m -£0.3m
Land and buildings transaction tax (Scotland) -£0.4m -£0.4m -£0.4m -£0.4m -£0.5m -£0.5m -£0.5m

Spending by benefit and service

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
State Pension +£319.8m +£325.9m +£368.2m +£343.6m +£353.8m +£364.2m +£375.0m
Pension Credit +£8.5m +£10.6m +£7.2m +£8.6m +£7.4m +£6.4m +£0.3m
Winter Fuel Payment +£3.8m +£5.5m +£3.2m +£2.7m +£2.7m +£3.4m +£3.6m
Pension Age Winter Heating Payment (Scotland) -£95,102 -£4,363 -£0.1m -£0.2m -£0.2m -£0.2m -£0.1m
Universal Credit -£6.4m +£20.7m +£36.3m +£79.3m +£101.4m +£35.2m +£28.4m
Housing Benefit +£17.0m +£17.4m +£16.8m +£17.7m +£18.1m +£18.4m +£19.1m
Council Tax Reduction -£15.4m -£14.6m -£16.4m -£17.7m -£18.1m -£15.1m -£14.5m
Employment and Support Allowance (income-related) -£4.1m -£4.2m -£4.3m -£4.4m -£4.5m -£4.6m -£4.8m
Employment and Support Allowance (contribution-based) +£2.0m +£2.1m +£2.1m +£2.2m +£2.2m +£2.3m +£2.3m
Jobseeker's Allowance (income-based) +£1.9m 0 0 0 0 0 0
Jobseeker's Allowance (contribution-based) -£1.6m -£1.7m -£1.7m -£1.8m -£1.8m -£1.8m -£1.9m
Income Support +£6.6m 0 0 0 0 0 0
Personal Independence Payment +£357.2m +£372.2m +£382.2m +£391.4m +£401.0m +£410.9m +£420.9m
Disability Living Allowance -£216.4m -£225.5m -£231.5m -£237.1m -£242.9m -£248.9m -£254.9m
Attendance Allowance +£75.5m +£78.7m +£80.8m +£82.7m +£84.8m +£86.9m +£89.0m
Carer's Allowance +£75.6m +£78.8m +£80.9m +£82.9m +£84.9m +£87.0m +£89.1m
Carer Support Payment (Scotland) -£9.4m -£9.7m -£10.0m -£10.2m -£10.5m -£10.8m -£11.0m
Carer's Allowance Supplement (Scotland) 0 -£1.3m -£1.3m -£1.3m -£1.3m -£1.3m -£1.3m
Severe Disablement Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Industrial Injuries Disablement Benefit -£4.4m -£4.5m -£4.6m -£4.8m -£4.9m -£5.0m -£5.1m
Armed Forces Compensation Scheme +£46.2m +£47.5m +£48.6m +£49.8m +£51.0m +£52.2m +£53.5m
Child Benefit -£27.8m -£28.2m -£30.2m -£30.8m -£31.6m -£32.2m -£31.6m
Scottish Child Payment +£8.4m +£8.5m +£8.7m +£9.0m +£9.2m +£9.5m +£10.2m
Healthy Start vouchers -£1.4m -£1.4m -£1.4m -£1.4m -£1.4m -£1.4m -£1.4m
Statutory Maternity Pay -£2.7m -£2.8m -£2.8m -£2.9m -£3.0m -£3.0m -£3.1m
Statutory Sick Pay -£25.3m -£25.9m -£26.6m -£27.2m -£27.9m -£28.5m -£29.2m
Sure Start Maternity Grant +£76,146 +£76,436 +£76,719 +£77,026 +£77,365 +£77,713 +£78,047
Bereavement Support Payment -£17.4m -£17.8m -£18.3m -£18.7m -£19.2m -£19.6m -£20.1m
Tax-Free Childcare +£0.4m -£5.0m -£4.8m -£5.0m -£5.1m -£12.2m -£12.3m
Extended childcare entitlement -£48.4m -£59.9m -£52.0m -£59.1m -£60.1m -£115.2m -£69.5m
Universal childcare entitlement -£16.7m -£12.5m -£12.8m -£12.0m -£12.4m -£12.6m -£13.0m
Targeted childcare entitlement -£5.9m -£6.1m -£6.3m -£6.4m -£6.6m -£6.1m +£5.9m
Parents' Learning Allowance +£0.5m +£0.5m +£0.5m +£0.5m +£0.6m +£0.6m +£0.6m
Adult Dependants' Grant +£84,059 +£84,378 -£0.4m -£0.4m -£0.4m -£0.4m -£0.4m
Disabled Students' Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
NHS spending +£39.2m +£39.3m +£39.5m +£39.6m +£39.8m +£40.0m +£40.2m
State education spending suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Transport subsidy spending +£34 +£41 +£9 -£25 -£47 -£82 +£208

Caseloads

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
State Pension +12.1k +11.3k +14.6k +12.5k +12.5k +12.6k +12.6k
Pension Credit +4.6k +11.5k +5.3k +4.6k +5.3k +6.2k +4.9k
Winter Fuel Payment +17.5k +24.0k +15.4k +12.1k +11.9k +16.0k +17.4k
Pension Age Winter Heating Payment (Scotland) +2.1k +2.2k +1.6k +1.4k +1.4k +1.4k +1.4k
Universal Credit -8.2k -6.7k -4.5k -6.5k -6.4k -3.6k -6.1k
Housing Benefit +1.9k +1.8k +1.8k +1.8k +1.8k +1.8k +2.0k
Council Tax Reduction -958 +1.1k -3.3k -5.7k -1.8k +311 +8.0k
Employment and Support Allowance (income-related) +791 +794 +797 +802 +805 +809 +813
Employment and Support Allowance (contribution-based) +184 +184 +185 +186 +186 +187 +188
Jobseeker's Allowance (income-based) +460 0 0 0 0 0 0
Jobseeker's Allowance (contribution-based) -286 -287 -289 -290 -291 -292 -293
Income Support +297 0 0 0 0 0 0
Personal Independence Payment +49.3k +49.5k +49.7k +49.9k +50.1k +50.4k +50.6k
Disability Living Allowance -37.4k -37.5k -37.7k -37.8k -38.0k -38.1k -38.3k
Attendance Allowance +12.4k +12.4k +12.5k +12.5k +12.6k +12.6k +12.7k
Carer's Allowance +17.5k +17.5k +17.6k +17.7k +17.7k +17.8k +17.9k
Carer Support Payment (Scotland) -2.2k -2.2k -2.2k -2.2k -2.2k -2.2k -2.2k
Carer's Allowance Supplement (Scotland) 0 -2.2k -2.2k -2.2k -2.2k -2.2k -2.2k
Severe Disablement Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Industrial Injuries Disablement Benefit +216 +217 +218 +218 +219 +220 +221
Armed Forces Compensation Scheme +5.4k +5.4k +5.5k +5.5k +5.5k +5.5k +5.5k
Child Benefit -22.7k -22.5k -22.8k -22.9k -23.0k -23.0k -22.4k
Scottish Child Payment +5.9k +5.8k +5.8k +5.9k +5.9k +5.9k +6.2k
Healthy Start vouchers -7.0k -7.0k -7.0k -7.1k -7.1k -7.1k -7.1k
Statutory Maternity Pay -1.7k -1.7k -1.7k -1.7k -1.7k -1.8k -1.8k
Statutory Sick Pay -4.9k -5.0k -5.0k -5.0k -5.0k -5.0k -5.1k
Sure Start Maternity Grant +152 +153 +153 +154 +155 +155 +156
Bereavement Support Payment -2.7k -2.7k -2.7k -2.8k -2.8k -2.8k -2.8k
Tax-Free Childcare -208 -4.7k -5.7k -6.0k -6.5k -12.4k -11.6k
Extended childcare entitlement -11.2k -12.6k -10.8k -11.7k -11.7k -32.8k -12.7k
Universal childcare entitlement -4.8k -3.5k -3.5k -3.2k -3.3k -3.3k -3.3k
Targeted childcare entitlement -1.2k -1.2k -1.2k -1.2k -1.2k -1.1k +1.1k
Parents' Learning Allowance +268 +269 +270 +271 +272 +274 +275
Adult Dependants' Grant +24 +24 -103 -104 -104 -104 -109
Disabled Students' Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed

Government balance, each component counted once

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Government receipts, each component once -£3.04bn -£3.11bn -£3.19bn -£3.28bn -£3.38bn -£3.48bn -£3.56bn
Government spending +£559.8m † +£586.8m † +£649.7m † +£668.7m † +£705.2m † +£597.9m † +£663.6m †
Government balance, each receipt once -£3.60bn † -£3.70bn † -£3.84bn † -£3.95bn † -£4.09bn † -£4.08bn † -£4.22bn †

Model totals

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
gov_tax -£3.04bn -£3.11bn -£3.19bn -£3.28bn -£3.38bn -£3.48bn -£3.56bn
household_tax +£367.3m +£378.9m +£380.5m +£381.6m +£366.0m +£354.2m +£369.6m
pre_budget_change_household_tax +£356.9m +£368.2m +£369.5m +£374.7m +£358.8m +£346.8m +£351.7m
gov_spending +£559.8m † +£586.8m † +£649.7m † +£668.7m † +£705.2m † +£597.9m † +£663.6m †
household_benefits +£559.8m † +£586.8m † +£649.7m † +£668.7m † +£705.2m † +£597.9m † +£663.6m †
gov_balance -£3.60bn † -£3.70bn † -£3.84bn † -£3.95bn † -£4.09bn † -£4.08bn † -£4.22bn †

Components the model counts twice

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
corporate_sdlt, counted again inside expected_sdlt +£43 +£408 -£307 -£507 -£329 +£169 -£208

Household income totals

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Household market income +£1.62bn +£1.69bn +£1.73bn +£1.77bn +£1.81bn +£1.86bn +£1.93bn
Household net income +£2.61bn † +£2.73bn † +£2.86bn † +£2.93bn † +£3.05bn † +£3.03bn † +£3.18bn †
Household net income (HBAI definition) suppressed suppressed suppressed suppressed suppressed suppressed suppressed

Population

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
People -10.0k -10.0k -10.0k -10.1k -10.1k -10.2k -10.2k
Benefit units -24.0k -24.1k -24.2k -24.3k -24.4k -24.5k -24.6k
Households -6.9k -6.9k -6.9k -6.9k -7.0k -7.0k -7.0k

Poverty rates

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Absolute poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed

People in poverty

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Absolute poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed

Inequality

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Gini coefficient, equivalised HBAI income before housing costs suppressed suppressed suppressed suppressed suppressed suppressed suppressed

Mean household net income in each side's own deciles

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Decile 1: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 2: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 3: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 4: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 5: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 6: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 7: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 8: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 9: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 10: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed

Standard reform costings (change in gov_balance)

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Raise basic rate by 1pp +£10.4m +£10.0m +£10.1m +£10.1m +£9.7m +£9.7m +£11.0m
Raise higher rate by 1pp +£21.2m +£22.9m +£24.3m +£25.4m +£26.7m +£28.0m +£29.3m
Raise personal allowance by ~800GBP/year -£8.6m -£1.8m -£1.3m -£3.6m -£90,258 -£1.1m -£0.7m
Raise child benefit by 25GBP/week per additional child -£19.1m -£16.4m -£16.5m -£18.3m -£17.4m -£17.1m -£16.6m
Reduce Universal Credit taper rate to 20% -£27.9m -£24.9m -£30.3m -£28.7m -£46.5m -£50.1m -£70.3m
Raise Class 1 main employee NICs rate to 10% +£13.7m +£14.1m +£14.8m +£12.8m +£13.2m +£12.7m +£15.3m
Raise VAT standard rate by 2pp -£367.0m -£376.8m -£385.8m -£395.1m -£404.8m -£414.7m -£424.8m
Raise additional rate by 3pp +£6.8m +£6.5m +£6.2m +£6.6m +£4.5m +£3.1m +£2.8m
Raise the Universal Credit standard allowance by 10% -£5.9m -£5.6m -£8.8m -£13.2m -£15.8m -£16.4m -£17.1m
Raise the State Pension by 10% -£34.9m -£28.9m -£38.4m -£40.0m -£41.2m -£42.4m -£46.1m

Standard reforms: change in each reform's target

Measure 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Raise basic rate by 1pp: change in income_tax +£8.9m +£9.0m +£8.8m +£8.3m +£8.6m +£8.7m +£8.8m
Raise higher rate by 1pp: change in income_tax +£21.3m +£23.1m +£24.5m +£25.5m +£26.8m +£28.1m +£29.2m
Raise personal allowance by ~800GBP/year: change in income_tax -£8.4m -£2.0m -£1.4m -£3.8m -£0.5m -£1.8m -£0.9m
Raise child benefit by 25GBP/week per additional child: change in child_benefit +£1.3m +£1.4m +£1.0m +£0.9m +£0.9m +£0.8m +£0.9m
Reduce Universal Credit taper rate to 20%: change in universal_credit +£28.0m +£24.9m +£30.3m +£28.8m +£46.7m +£50.3m +£70.4m
Raise Class 1 main employee NICs rate to 10%: change in national_insurance +£11.9m +£11.8m +£12.3m +£12.4m +£13.5m +£13.3m +£13.6m
Raise VAT standard rate by 2pp: change in vat -£367.0m -£376.8m -£385.8m -£395.1m -£404.8m -£414.7m -£424.8m
Raise additional rate by 3pp: change in income_tax +£6.8m +£6.5m +£6.2m +£6.6m +£4.5m +£3.1m +£2.8m
Raise the Universal Credit standard allowance by 10%: change in universal_credit +£6.2m +£5.9m +£9.1m +£13.6m +£16.2m +£16.8m +£17.5m
Raise the State Pension by 10%: change in state_pension +£32.0m +£32.6m +£36.8m +£34.4m +£35.4m +£36.4m +£37.5m

Household net income by decile

The datasets hold different households, so households are not matched; compare the mean income in each side's own deciles.

Checks failing on both base and head (3)
  • Each model total's components are exactly the partition's heads and declared double counts (closure): gov_spending: the partition puts 'healthy_start_vouchers' in this total, but the model does not add it; gov_spending: the partition puts 'scottish_child_payment' in this total, but the model does not add it; gov_spending: the partition puts 'statutory_paternity_pay' in this total, but the model does not add it
  • The partition's heads and declared double counts add up to gov_spending (identity:gov_spending): 2025-26: 358 households differ, by a weighted £539,873,913 in total (households holding a figure from fewer than 10 records aren't checked this year); 2026-27: 360 households differ, by a weighted £561,039,261 in total (households holding a figure from fewer than 10 records aren't checked this year); 2027-28: 360 households differ, by a weighted £574,714,761 in total (households holding a figure from fewer than 10 records aren't checked this year)
  • The partition's heads and declared double counts add up to household_benefits (identity:household_benefits): 2025-26: 1,819 households differ, by a weighted £92,416,699 in total (households holding a figure from fewer than 10 records aren't checked this year); 2026-27: 1,791 households differ, by a weighted £94,351,503 in total (households holding a figure from fewer than 10 records aren't checked this year); 2027-28: 1,739 households differ, by a weighted £94,705,918 in total (households holding a figure from fewer than 10 records aren't checked this year)
Counted-once check

Counted-once check not run: not requested.

Provenance
Side Commit Runtime Environment hash Partition hash Time
base policyengi uk 2.122.2 (files 0cb87d75a516), core 3.32.13, py 3.13.16 x86_64 a9ab0df50691 e8ea8a019873 202.44s
head policyengi uk 2.122.2 (files 0cb87d75a516), core 3.32.13, py 3.13.16 x86_64 a9ab0df50691 e8ea8a019873 174.86s

Base dataset: enhanced_frs_2024_25.h5, sha256 c50ac2b57ad332a0a657e2ed78ebd33aa2a19420f56beea61f744ddc251bd2a5, the base's build.

Head dataset: enhanced_frs_2024_25.h5, sha256 4f5f89b2c7f87aeb12329f1b4337809e1e35e04a96728c8450359610a546e83e, this pull request's build.

Receipts by tax head: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Income tax £298.05bn £317.52bn £333.46bn £347.85bn £363.79bn £381.16bn £401.03bn
National Insurance (employee and self-employed) £51.15bn £53.36bn £55.11bn £56.69bn £58.60bn £60.41bn £62.74bn
National Insurance (employer) £151.13bn £157.57bn £162.46bn £166.98bn £173.69bn £179.13bn £186.18bn
Employer NI change borne by capital £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Employer NI change borne by consumers £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Capital gains tax £25.06bn £25.91bn £26.86bn £27.82bn £28.79bn £29.85bn £30.94bn
Student loan repayments £7.71bn £8.00bn £8.22bn £8.43bn £8.65bn £8.90bn £8.29bn
VAT £320.46bn £329.08bn £336.90bn £345.01bn £353.46bn £362.15bn £370.99bn
Change in VAT from its baseline £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
VAT on private school fees £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Fuel duties £24.04bn £24.14bn £26.10bn £26.16bn £25.89bn £25.24bn £25.82bn
Alcohol duties £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Tobacco duties £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Vehicle excise duties £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
TV licence fee £4.50bn £4.66bn £4.68bn £4.70bn £4.72bn £4.74bn £4.77bn
Council tax £48.71bn £51.55bn £54.81bn £58.35bn £61.87bn £65.55bn £65.83bn
High value council tax surcharge £0.00bn £0.00bn £0.00bn £0.69bn £0.71bn £0.73bn £0.74bn
Northern Ireland domestic rates £1.12bn £1.18bn £1.23bn £1.29bn £1.35bn £1.42bn £1.42bn
Business rates £31.73bn £31.73bn £31.73bn £31.73bn £31.73bn £31.73bn £31.73bn
Stamp duty land tax (own purchases, and corporate until the £19.49bn £20.36bn £21.33bn £22.29bn £23.29bn £24.40bn £25.58bn
Land transaction tax (Wales) £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn
Land and buildings transaction tax (Scotland) £0.06bn £0.06bn £0.06bn £0.06bn £0.07bn £0.07bn £0.07bn
Unspecified tax change borne by capital £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Unspecified tax change borne by consumers £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Wealth tax £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Wealth tax excluding the main residence £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Land value tax £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Carbon tax £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Spending by benefit and service: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
State Pension £133.47bn £138.54bn £140.83bn £143.18bn £147.40bn £151.77bn £156.23bn
Pension Credit £6.54bn £7.15bn £7.03bn £7.08bn £7.19bn £7.30bn £7.31bn
Winter Fuel Payment £1.69bn £1.66bn £1.62bn £1.59bn £1.58bn £1.58bn £1.57bn
Pension Age Winter Heating Payment (Scotland) £0.09bn £0.09bn £0.09bn £0.10bn £0.10bn £0.10bn £0.10bn
Universal Credit £78.88bn £83.10bn £84.92bn £86.47bn £87.39bn £84.60bn £85.09bn
Housing Benefit £8.34bn £8.67bn £8.86bn £9.07bn £9.35bn £9.59bn £9.55bn
Council Tax Reduction £2.17bn £2.23bn £2.30bn £2.36bn £2.42bn £2.50bn £2.46bn
Employment and Support Allowance (income-related) £3.20bn £3.28bn £3.36bn £3.43bn £3.52bn £3.60bn £3.69bn
Employment and Support Allowance (contribution-based) £4.74bn £4.87bn £4.99bn £5.11bn £5.23bn £5.36bn £5.49bn
Jobseeker's Allowance (income-based) £0.05bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Jobseeker's Allowance (contribution-based) £0.29bn £0.29bn £0.30bn £0.31bn £0.32bn £0.32bn £0.33bn
Income Support £0.17bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Working Tax Credit £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Child Tax Credit £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Incapacity Benefit £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Personal Independence Payment £30.42bn £31.70bn £32.55bn £33.34bn £34.15bn £34.99bn £35.85bn
Disability Living Allowance £8.12bn £8.46bn £8.69bn £8.90bn £9.11bn £9.34bn £9.57bn
Attendance Allowance £7.78bn £8.11bn £8.33bn £8.53bn £8.74bn £8.95bn £9.17bn
Carer's Allowance £3.98bn £4.15bn £4.26bn £4.36bn £4.47bn £4.58bn £4.69bn
Carer Support Payment (Scotland) £0.31bn £0.32bn £0.33bn £0.34bn £0.34bn £0.35bn £0.36bn
Carer's Allowance Supplement (Scotland) £0.00bn £0.04bn £0.04bn £0.04bn £0.04bn £0.04bn £0.04bn
Severe Disablement Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Industrial Injuries Disablement Benefit £0.62bn £0.64bn £0.65bn £0.67bn £0.69bn £0.70bn £0.72bn
Armed Forces Compensation Scheme £0.59bn £0.61bn £0.62bn £0.64bn £0.65bn £0.67bn £0.69bn
Child Benefit £16.35bn £16.99bn £17.41bn £17.82bn £18.23bn £18.67bn £19.07bn
Scottish Child Payment £0.48bn £0.50bn £0.51bn £0.53bn £0.54bn £0.55bn £0.56bn
Two Child Limit Payment (Scotland, counterfactual) £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Healthy Start vouchers £0.06bn £0.06bn £0.06bn £0.06bn £0.06bn £0.06bn £0.06bn
Maternity Allowance £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Statutory Maternity Pay £3.84bn £3.94bn £4.03bn £4.13bn £4.23bn £4.34bn £4.44bn
Statutory Paternity Pay £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Statutory Sick Pay £0.97bn £1.00bn £1.02bn £1.04bn £1.07bn £1.10bn £1.12bn
Sure Start Maternity Grant £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn £0.01bn
Bereavement Support Payment £1.13bn £1.16bn £1.19bn £1.22bn £1.25bn £1.28bn £1.31bn
Tax-Free Childcare £0.45bn £0.44bn £0.44bn £0.45bn £0.46bn £0.47bn £0.47bn
Extended childcare entitlement £3.76bn £4.17bn £4.29bn £4.40bn £4.52bn £4.68bn £4.82bn
Universal childcare entitlement £1.50bn £1.59bn £1.64bn £1.68bn £1.72bn £1.74bn £1.79bn
Targeted childcare entitlement £0.37bn £0.38bn £0.40bn £0.40bn £0.41bn £0.41bn £0.42bn
Care to Learn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Childcare Grant £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Parents' Learning Allowance £0.19bn £0.19bn £0.19bn £0.19bn £0.19bn £0.20bn £0.20bn
Adult Dependants' Grant £0.06bn £0.06bn £0.05bn £0.05bn £0.05bn £0.05bn £0.05bn
Travel Grant £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Disabled Students' Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
16 to 19 Bursary Fund £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
NHS spending £208.41bn £209.21bn £209.98bn £210.82bn £211.75bn £212.70bn £213.62bn
State education spending suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Transport subsidy spending £25.15bn £25.25bn £25.34bn £25.44bn £25.55bn £25.67bn £30.94bn
Cost-of-living support payment £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Energy Bills Rebate £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Energy Price Guarantee subsidy £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Basic income £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Disability basic income £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Other public spending budget change £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Caseloads: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
State Pension 11,748k 11,642k 11,397k 11,307k 11,357k 11,408k 11,457k
Pension Credit 1,369k 1,392k 1,346k 1,329k 1,328k 1,326k 1,307k
Winter Fuel Payment 7,150k 7,027k 6,805k 6,671k 6,648k 6,617k 6,606k
Pension Age Winter Heating Payment (Scotland) 710k 703k 691k 684k 687k 690k 693k
Universal Credit 6,374k 6,404k 6,430k 6,432k 6,407k 6,396k 6,373k
Housing Benefit 1,211k 1,208k 1,202k 1,198k 1,204k 1,204k 1,201k
Council Tax Reduction 3,612k 3,640k 3,682k 3,705k 3,728k 3,748k 3,723k
Employment and Support Allowance (income-related) 310k 311k 312k 313k 315k 316k 317k
Employment and Support Allowance (contribution-based) 668k 670k 673k 675k 678k 681k 684k
Jobseeker's Allowance (income-based) 9k 0k 0k 0k 0k 0k 0k
Jobseeker's Allowance (contribution-based) 60k 60k 60k 61k 61k 61k 61k
Income Support 26k 0k 0k 0k 0k 0k 0k
Working Tax Credit 0k 0k 0k 0k 0k 0k 0k
Child Tax Credit 0k 0k 0k 0k 0k 0k 0k
Incapacity Benefit 0k 0k 0k 0k 0k 0k 0k
Personal Independence Payment 4,133k 4,149k 4,164k 4,181k 4,199k 4,218k 4,236k
Disability Living Allowance 1,332k 1,337k 1,342k 1,348k 1,353k 1,360k 1,365k
Attendance Allowance 1,538k 1,544k 1,550k 1,556k 1,563k 1,570k 1,577k
Carer's Allowance 919k 922k 926k 929k 934k 938k 942k
Carer Support Payment (Scotland) 71k 71k 71k 72k 72k 72k 73k
Carer's Allowance Supplement (Scotland) 0k 71k 71k 72k 72k 72k 73k
Severe Disablement Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Industrial Injuries Disablement Benefit 158k 159k 159k 160k 161k 161k 162k
Armed Forces Compensation Scheme 66k 67k 67k 67k 67k 68k 68k
Child Benefit 7,955k 7,969k 7,983k 8,009k 8,036k 8,067k 8,081k
Scottish Child Payment 340k 342k 343k 345k 343k 344k 343k
Two Child Limit Payment (Scotland, counterfactual) 0k 0k 0k 0k 0k 0k 0k
Healthy Start vouchers 231k 232k 233k 234k 235k 236k 237k
Maternity Allowance 0k 0k 0k 0k 0k 0k 0k
Statutory Maternity Pay 304k 305k 306k 307k 308k 310k 311k
Statutory Paternity Pay 0k 0k 0k 0k 0k 0k 0k
Statutory Sick Pay 152k 152k 153k 153k 154k 155k 155k
Sure Start Maternity Grant 22k 22k 23k 23k 23k 23k 23k
Bereavement Support Payment 183k 184k 184k 185k 186k 187k 188k
Tax-Free Childcare 965k 951k 951k 957k 934k 944k 933k
Extended childcare entitlement 792k 783k 790k 793k 795k 804k 822k
Universal childcare entitlement 438k 448k 451k 452k 454k 452k 456k
Targeted childcare entitlement 77k 77k 77k 78k 78k 76k 75k
Care to Learn 0k 0k 0k 0k 0k 0k 0k
Childcare Grant 0k 0k 0k 0k 0k 0k 0k
Parents' Learning Allowance 95k 95k 96k 96k 96k 97k 97k
Adult Dependants' Grant 17k 18k 15k 15k 15k 15k 14k
Travel Grant 0k 0k 0k 0k 0k 0k 0k
Disabled Students' Allowance suppressed suppressed suppressed suppressed suppressed suppressed suppressed
16 to 19 Bursary Fund 0k 0k 0k 0k 0k 0k 0k
Cost-of-living support payment 0k 0k 0k 0k 0k 0k 0k
Energy Bills Rebate 0k 0k 0k 0k 0k 0k 0k
Energy Price Guarantee subsidy 0k 0k 0k 0k 0k 0k 0k
Basic income 0k 0k 0k 0k 0k 0k 0k
Disability basic income 0k 0k 0k 0k 0k 0k 0k
Government balance, each component counted once: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Government receipts, each component once £983.23bn £1,025.12bn £1,062.97bn £1,098.07bn £1,136.63bn £1,175.49bn £1,216.15bn
Government spending £554.20bn † £568.88bn † £576.36bn † £583.75bn † £592.74bn † £598.27bn † £611.80bn †
Government balance, each receipt once £429.03bn † £456.24bn † £486.61bn † £514.31bn † £543.89bn † £577.23bn † £604.34bn †
Benefits in household_benefits but not in gov_spending £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn £0.00bn
Model totals: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
gov_tax £986.36bn £1,028.25bn £1,066.10bn £1,101.20bn £1,139.76bn £1,178.62bn £1,219.28bn
household_tax £514.77bn £541.60bn £566.74bn £589.21bn £612.60bn £637.34bn £662.11bn
pre_budget_change_household_tax £507.06bn £533.60bn £558.52bn £580.08bn £603.24bn £627.72bn £653.08bn
gov_spending £554.20bn † £568.88bn † £576.36bn † £583.75bn † £592.74bn † £598.27bn † £611.80bn †
household_benefits £554.20bn † £568.88bn † £576.36bn † £583.75bn † £592.74bn † £598.27bn † £611.80bn †
gov_balance £432.16bn † £459.37bn † £489.74bn † £517.44bn † £547.02bn † £580.36bn † £607.47bn †
Components the model counts twice: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
corporate_sdlt, counted again inside expected_sdlt £3.13bn £3.13bn £3.13bn £3.13bn £3.13bn £3.13bn £3.13bn
Household income totals: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Household market income £1,661.20bn £1,723.43bn £1,776.20bn £1,826.88bn £1,892.35bn £1,950.89bn £2,022.88bn
Household net income £1,660.17bn † £1,708.72bn † £1,742.65bn † £1,777.18bn † £1,812.87bn † £1,850.84bn † £1,909.07bn †
Household net income (HBAI definition) suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Population: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
People 70,730k 70,999k 71,262k 71,547k 71,862k 72,185k 72,496k
Benefit units 38,918k 39,066k 39,210k 39,367k 39,540k 39,718k 39,889k
Households 32,058k 32,180k 32,299k 32,428k 32,571k 32,717k 32,858k
Poverty rates: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Absolute poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
People in poverty: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Absolute poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Absolute poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, before housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: everyone suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: children suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: working-age adults suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Relative poverty, after housing costs: pensioners suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Inequality: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Gini coefficient, equivalised HBAI income before housing costs suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Mean household net income in each side's own deciles: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Decile 1: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 2: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 3: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 4: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 5: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 6: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 7: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 8: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 9: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Decile 10: mean household_net_income per household suppressed suppressed suppressed suppressed suppressed suppressed suppressed
Standard reform costings (change in gov_balance): head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Raise basic rate by 1pp £6.04bn £6.31bn £6.52bn £6.71bn £6.91bn £7.14bn £7.39bn
Raise higher rate by 1pp £3.76bn £4.06bn £4.29bn £4.52bn £4.78bn £5.06bn £5.28bn
Raise personal allowance by ~800GBP/year £-3.59bn £-3.69bn £-3.80bn £-3.86bn £-3.91bn £-3.98bn £-1.67bn
Raise child benefit by 25GBP/week per additional child £-2.05bn £-1.82bn £-1.69bn £-1.59bn £-1.48bn £-1.37bn £-1.26bn
Reduce Universal Credit taper rate to 20% £-16.79bn £-17.42bn £-17.66bn £-17.86bn £-17.91bn £-18.13bn £-18.16bn
Raise Class 1 main employee NICs rate to 10% £10.63bn £11.04bn £11.34bn £11.60bn £11.89bn £12.19bn £12.63bn
Raise VAT standard rate by 2pp £31.60bn £32.45bn £33.22bn £34.02bn £34.85bn £35.71bn £36.58bn
Raise additional rate by 3pp £3.27bn £3.52bn £3.73bn £3.94bn £4.18bn £4.43bn £4.79bn
Raise the Universal Credit standard allowance by 10% £-2.97bn £-3.12bn £-3.17bn £-3.22bn £-3.24bn £-3.28bn £-3.33bn
Raise the State Pension by 10% £-9.77bn £-10.02bn £-10.10bn £-10.20bn £-10.43bn £-10.69bn £-10.97bn
Standard reforms: change in each reform's target: head levels
Measure (head) 2025-26 2026-27 2027-28 2028-29 2029-30 2030-31 2031-32
Raise basic rate by 1pp: change in income_tax £6.15bn £6.43bn £6.64bn £6.84bn £7.05bn £7.27bn £7.53bn
Raise higher rate by 1pp: change in income_tax £3.77bn £4.06bn £4.30bn £4.53bn £4.79bn £5.07bn £5.29bn
Raise personal allowance by ~800GBP/year: change in income_tax £-3.69bn £-3.81bn £-3.91bn £-3.98bn £-4.03bn £-4.11bn £-1.73bn
Raise child benefit by 25GBP/week per additional child: change in child_benefit £2.50bn £2.29bn £2.17bn £2.05bn £1.94bn £1.82bn £1.70bn
Reduce Universal Credit taper rate to 20%: change in universal_credit £16.83bn £17.46bn £17.71bn £17.90bn £17.96bn £18.17bn £18.20bn
Raise Class 1 main employee NICs rate to 10%: change in national_insurance £10.81bn £11.23bn £11.54bn £11.82bn £12.11bn £12.42bn £12.87bn
Raise VAT standard rate by 2pp: change in vat £31.60bn £32.45bn £33.22bn £34.02bn £34.85bn £35.71bn £36.58bn
Raise additional rate by 3pp: change in income_tax £3.27bn £3.52bn £3.73bn £3.94bn £4.18bn £4.43bn £4.79bn
Raise the Universal Credit standard allowance by 10%: change in universal_credit £2.97bn £3.12bn £3.17bn £3.22bn £3.25bn £3.29bn £3.33bn
Raise the State Pension by 10%: change in state_pension £13.35bn £13.85bn £14.08bn £14.32bn £14.74bn £15.18bn £15.62bn

@MaxGhenis
MaxGhenis marked this pull request as ready for review October 25, 2026 16:13
@MaxGhenis

MaxGhenis commented Oct 25, 2026

Copy link
Copy Markdown
Contributor Author

Hand-off to the UK hub (session local_f07bc361-04ae-4cb2-ac13-981fa073bc98, 2026-10-25).

  • Head: 9fd9164e8b0932fafde6838c812c458b2318d701 (ready for review, MERGEABLE).
  • Review: GPT-6.1 Sol delta review, APPROVE at this exact head, no findings (~/reviews/uk-data-cgt-vat-2026-10-11/review/r2_572.md). Round 1 (APPROVE WITH NITS at d14e0c7b, review/r1_572.md) nits are both fixed in 6be52fe8.
  • CI: 6 of 6 pass at this head (run 40190348100): Test 2,674 passed, including the built-data tests that VAT and its inputs are non-negative in 2025 and 2030. The fit, distribution and aggregate comparisons pass. The aggregate tool's sign:vat check fails on the base and passes on the head.
  • Impact (in the body, from CI's same-harness base and head builds): VAT −£3.68bn (−1.13%) in 2025-26, rising to −£4.15bn in 2030-31. Income tax +0.2% and household net income +0.16% are weight responses. The CGT move is within same-code noise.
  • Pending: nothing on the PR. It lands only in a batch on Max's go (d833), as a batch-4 candidate.
  • Depends on: nothing.
  • Follow-ups owed: none for this PR. The related issue 16 households have a negative vehicle count on 1.58.0 (WAS missing-value code in the wealth imputation) #574 (negative num_vehicles, WAS missing-value code) stays open. The CGT target-level question after the 2024-25 forestalling year (Start capital gains on HMRC's 2024-25 size-of-gain table #571's line) went to a separate session (chip task_035a55d3).

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

76 households have negative VAT on 1.58.0

1 participant

@MaxGhenis