We have just finished an 18-month study of value for money in British policing, which the Police Foundation is kindly helping us launch in the House of Lords on Monday 20 March. Our main finding is that crime, as reported by the public, has not risen since 2010, nor has public confidence in the police declined. As the period has seen a sharp reduction in public expenditure on policing, that suggests that value for money has been improving. The creation of Police and Crime Commissioners (PCCs) in 2012 may have contributed modestly to this result

The study was commissioned by CIPFA (the Chartered Institute of Public Finance and Accountancy) and conducted by the Gwilym Gibbon Centre for Public Policy at Nuffield College, University of Oxford. We interviewed a wide range of PCCs, senior officers, and heads of finance many thanks to all who found the time to speak to us. We also conducted some statistical analysis. And, although the funding side of the equation was not part of our original brief from CIPFA, the recent collapse of a Home Office funding formula led us to write about the strengths and weaknesses of Britain’s three distinct ways of funding the police. Wales is slightly different from England, and Scotland differs radically from both. (Our study did not cover Northern Ireland).

Our interviewees told us many stories of efficient work in hard times. Many of them stressed the need for police authorities to become more proactive. Almost all of them stressed the advantages of joint working with other agencies, given that so much police time is taken up by the needs of vulnerable people, and the police are the only social service available 24 hours a day and seven days a week. Experiments in joint working in Greater Manchester, Cheshire, and Shrewsbury among others should be watched carefully.

We looked at the creation of PCCs and the first two elections. We found that:

• The introduction of PCCs was botched because of the late change in the date of the first elections and the use of an unacceptable electoral system.
• The legitimacy of the role increased somewhat after the 2016 round of elections.
• Although most PCCs now come from political parties, neither election was strongly partisan. Party and independent candidates made very similar promises.
• Average HMIC (Her Majesty’s Inspectorate of Constabulary) ratings of forces improved slightly after the introduction of PCCs, but the measured improvement was too small to be statistically significant.

Police funding is full of perversities, one of which Sir Robert Peel would recognise. Part of it comes from local taxes council tax and business rates. However the poorest areas, with the most social deprivation, are those least able to rely on the local tax base. There needs to be an equalisation formula. But the old formula and the one the Home Office abortively tried to introduce were radically flawed. The old one has too many indicators, which are too prone to lobbying. The one proposed in 2015 would have fixed that, but it was based on a commercial database which the Home Office refused to release to forces. When the PCC’s office in Devon and Cornwall bought one, it found that the Home Office had not used it correctly. This helps to confirm the usefulness of PCCs. We hope that the working party now working on the next attempt pays a visit to Australia, whose Commonwealth Grants Commission is a model of how to do equalisation.

Measuring police force effectiveness is difficult. Everybody recognises that crime as recorded by forces is an unreliable statistic, because forces will always have discretion on what to report. One consequence is that if you appoint more officers, more crime will be reported. Most statistical studies of effectiveness fail at this very basic level. The only reliable relevant statistic is the survey-based report of how much crime people have experienced, and how satisfied they are with the police. We therefore used the Crime Survey of England and Wales (CSEW) as our basic source for data. The data are available at force level, although it took many months for the Office for National Statistics to make them fully available to us.

One of our results shows how slippery this is. Although HMIC ratings of forces went up after PCCs were introduced, the proportion of the public thinking that local crime had risen also went up, not down as predicted. We think that the very introduction of PCC elections provided a new forum to talk about crime rates. If people talk about things, they think they are happening.

There is a lot of fascinating detail in our findings (well, we think it is fascinating). Our five working papers are available for free download at http://ggcpp.nuff.ox.ac.uk/index.php/working-papers/ and hard copies of our final report will be available at the launch on 20 March.

Professor Iain McLean, Gwilym Gibbon Centre for Public Policy, Nuffield College, University of Oxford