Politicians have a point, police numbers really do affect the levels of crime | Torsten Bell

Police officers are back in the headlines – and for once not because of terrible misconduct.

The government announced mission accomplished last week on its manifesto commitment to recruit 20,000 extra coppers. Labour hit back, saying this was just reversing the government’s own cuts since 2010. The reality is somewhere in between: there has now been an increase in police numbers since 2010, but we’re talking 3,500, not 20,000, and forces have grown more slowly than the populations they police.

The focus of politicians on police numbers, rather than wider questions of good policing and the underlying causes of crime, often annoys experts, who point out that tackling crime is far more complicated than putting police officers on the streets. They’re right, of course, but, to defend the politicians for a second, police numbers do matter.

Researchers struggle to measure the impact on crime of hiring more police because the relationship is a two-way street – looking for evidence of police bringing down crime is hard when more crime itself often leads to more police. But research from the US gets round this problem, comparing areas that got specific federal grants to hire extra police officers with areas that also applied (ie had similar desire for more officers given their crime situation) but did not receive the funding.

It shows that a 10% increase in police numbers reduces violent crime by 13% and property crime by 7%. This happens without higher arrest rates for such crimes, suggesting the impact is via deterrence (there is an increase in arrest rates for less serious crimes). Tackling crime, like many societal problems we face, is complicated, but don’t rubbish the politicians or public wanting to see bobbies on the beat.