Does your police commissioner's party predict crime?

The raw figures

Every force with an elected commissioner, grouped by the party that won the most recent election for that post, over the 12 months to June 2026. A force has no resident population of its own, so the rate is the crime and the population of the local authority areas it polices, added up.

Labour and Labour Co-op are counted together: candidates stand on a joint ticket. Forces whose districts have no usable figures are absent, because a force cannot be given a rate built on months its officers never filed. Three forces have no elected commissioner at all and cannot appear: the City of London, British Transport Police and the Police Service of Northern Ireland.

What happens when you account for deprivation

The site already fits recorded crime against the Index of Multiple Deprivation across every English local authority area. That fit knows nothing about politics. Running each party's areas through it gives the rate their deprivation alone would predict, and the difference between that and what they actually record.

The raw gap is 33%. Measured against what deprivation predicts for each side, 10 percentage points of it survive. Deprivation alone accounts for roughly 70% of the difference between Labour and Conservative areas, before anyone has looked at commuting, town centres or how differently forces record what is reported to them.

England only: the Index of Multiple Deprivation 2025 covers England, and Wales publishes a different index on a different basis that the two cannot share. Welsh forces are in the raw table above and out of this one. A correlation is not a cause in either direction, and the index is not fully independent of crime: recorded crime is one of its seven domains, worth 9.3% of the score, so the model is to that extent predicting crime partly from crime. That works against the finding on this page rather than for it, because it makes deprivation look like a better predictor than it is.

Why the raw table is mostly a map

Commissioner elections are fought on force boundaries, and those boundaries divide England and Wales into metropolitan forces and shire ones. Labour wins in the metropolitan forces: West Midlands, Merseyside, South Yorkshire, Northumbria, Cleveland. Conservatives win in the rural and suburban ones. So a table of crime by party is close to a table of crime by how urban a place is, which is a thing this site can already show you directly.

Urban areas record more crime per resident for reasons that have nothing to do with who was elected. A rate divides the crime of everyone who passes through a place by the number of people who sleep there, so a city centre with a station, a shopping street and a nightlife strip carries the offences of hundreds of thousands of visitors against tens of thousands of residents. Deprivation is concentrated in the same places, and the relationship between deprivation and recorded crime is one of the strongest in the data.

There is also a timing problem that no amount of arithmetic fixes. These figures cover 12 months. Commissioners serve terms of three or four years, and several of the people in the table below were elected part-way through the window or after it closed entirely. A figure cannot be evidence about a person who was not in post while it was being recorded.

Every force, its commissioner and its rate

The underlying rows, so the grouping above can be checked rather than taken. Where an area records more than its deprivation predicts the difference is positive, in the same colour this site uses for a rising rate everywhere else.

The same question for the country

Recorded crime across England and Wales, month by month, set against the party in government. The archive opens in December 2010, so it holds one change of governing party: everything to June 2024 sits under a Conservative-led government, everything after it under Labour.

As published, recorded crime has run 4.6% lower a month under Labour than under the Conservative-led governments before it. Most of that is not crime. The number of forces filing a return moves between 41 and 45 across the archive, and the Labour months average 1.3 fewer forces than the Conservative ones. Compare only the 34 forces that filed in every single month, so the same places are measured against themselves, and the difference falls to 1.4%.

What actually moves the national line

Whole calendar years, the same 34 forces throughout, so nothing here is a change in who was supplying data.

On this panel recorded crime fell 19% from 2011 to 2014, then rose 21% again by 2020. Every one of those movements happened inside a single government, and the rise is mostly not a rise in crime. After a 2014 inspectorate report found forces were failing to record a large share of what was reported to them, forces were pushed to record more of it, and the recorded totals rose accordingly. A line that moves by a fifth because of a change in paperwork cannot be read as a scorecard for whoever was in office while it moved.

Two years to treat with particular care: 2020 and 2021 contain the pandemic, when anti-social behaviour reports rose sharply as breaches of the restrictions were logged while other offence types fell. How the figures work sets out what recorded crime does and does not count.

Questions

Do areas with Labour commissioners have more crime?

They record more: 108.1 crimes per 1,000 residents against 81.3 in areas with a Conservative commissioner. That is a real difference in the figures and a poor guide to anything else. Labour holds the metropolitan forces, urban areas record more crime per resident whoever runs them, and once each side is measured against the deprivation it actually has, most of the gap is gone.

What about Plaid Cymru and Reform UK?

Plaid Cymru holds one commissioner post, Dyfed-Powys Police, won in May 2024. One force is an anecdote, not a party record: it is a fact about Dyfed-Powys Police, and it is in the table because leaving it out would be its own distortion. Reform UK holds one commissioner post, Norfolk Constabulary, won in July 2026. These figures cover the 12 months to June 2026, so the measurement period ended before that commissioner took office and nothing counted here happened on their watch. One force is an anecdote, not a party record: it is a fact about Norfolk Constabulary, and it is in the table because leaving it out would be its own distortion.

What about councils, Labour-run versus Reform-run?

Councils in England and Wales have no policing powers, so council control is the wrong variable: a county council cannot direct a chief constable and does not set the force budget. Policing is overseen by the elected commissioner, or by the combined authority mayor where that role has absorbed the commissioner's functions, which is what this page uses. How the figures work sets out the rest.

Has crime gone up or down under this government?

On a constant panel of forces, recorded crime is running about 1.4% lower a month than the average of the thirteen and a half years before July 2024. That is a small difference measured over a short period, against a series that has moved by a fifth for reasons of recording practice alone, and it covers one change of government. It is not enough to credit or blame anyone with, and anybody presenting it as though it were is choosing their window.

Could a commissioner change these figures at all?

In the direction people expect, only slowly and only at the margin. The clearest thing a commissioner can move is how crime is recorded rather than how much of it happens: a force that tightens its recording practice books more of what is reported to it and its rate rises, which looks like a place getting worse and is a force getting more accurate. That is one of several reasons this site treats recorded crime as a measure of what police wrote down, not of what happened.