Methodology

Population data first. Community voice alongside. Cities in context.

DDPP exists for the phase many evaluations skip: after policies have run long enough for outcomes to appear in official statistics — and after communities have formed a view of what changed on the ground.

Government statistics as the backbone

We prioritise authoritative, population-level sources — ONS, departmental statistics, NHS and public health profiles, DfT and local open data — so findings can be cited in briefings, committee papers and business cases. Where early policy design relied on samples, modelling or pilot evidence, we reconstruct what should have been observable at scale and test whether it materialised.

Stress-testing sustainable policy

Transport, housing, growth and skills programmes often launch with optimistic assumptions from pilots or modelling. Our stress-tests ask: did indicators move in the claimed direction; are effects robust to alternative specifications; do health and equity outcomes hold when viewed by income, ethnicity, disability, age and geography? We are explicit about uncertainty, data lag and counterfactual limits.

Community forums and polling

Administrative data rarely captures perceived safety, trust, time poverty or informal adaptation to a scheme. We complement statistics with structured engagement — deliberative forums where appropriate, and polling where breadth and comparability matter. Community evidence is woven into synthesis, not treated as anecdote.

UK cities and their regions

City policies ripple outward: commuting, housing pressure, retail spend, air quality and hospital catchments do not respect borough boundaries. We routinely analyse urban cores alongside their travel-to-work areas and neighbouring authorities, so commissioners see spillovers and dependencies — not just a map shaded inside one council line.

Principles commissioners can expect

Independence

Clear scope, published methods, separation from implementation teams where required.

Proportionality

Depth matched to decision stakes — from rapid indicator scans to full outcome reviews.

Decision support

Outputs structured for action: continue, adapt, scale, pause or redesign with equity checks.