It is reported this week that the Metropolitan Police is reviving a proposal, last aired (and rejected by the Labour government) in 2007, to allow the introduction of ethnic quotas in recruitment.

This time round, Assistant Commissioner Simon Byrne aims to recruit a white police officer only ‘if you can recruit a black or minority ethnic person at the same time’. In other words, 50 per cent of new recruits would have to be BME. Senior police are linking this proposal to what is claimed to be increasing racial and religious tension in the capital.

Inspiration may partly have been drawn from the United States, where from the early 1970s many American cities faced legal challenges to their police recruitment strategies. The courts found in most cases that their use of entrance examinations had ‘disparate impact’, producing more successful white candidates than African-Americans or, later, Hispanics. The courts in many cases subsequently insisted that police departments had to hire according to ethnic quotas until their police forces looked more representative of their cities.

The classic study of the effects of these quotas, by Justin McCrary of the University of Michigan, found that affirmative action led over 25 years to an average 14 percentage point gain in the fraction of African-Americans among newly hired officers.

So should we change the law to allow police authorities to engage in positive discrimination? Many readers will dislike this idea on principle. But there are also practical problems which are worth emphasising.

For one thing, the US experience has not been without real difficulties. Although minority recruitment certainly rose, low levels of quits by existing police officers meant that positive discrimination usually failed to reach target levels quickly and thus quotas, originally thought to be temporary, remained in place for many years. There were sometimes morale and organisational problems which led to temporary falls in productivity and clearance rates. Rapid changes in city populations meant that quotas needed to be adjusted over time, for instance as Hispanics and Asians rose in numbers relative to African-Americans in many cities, and this proved contentious. And over time the political tide has turned against this type of intervention. In police recruitment and in the comparable area of university education where similar quotas have been imposed, increasing court challenges by aggrieved white candidates have narrowed their scope, and positive discrimination may even be ruled unconstitutional in the near future.

In London there are particular problems in imposing ethnic quotas of this kind. Much has been made of the apparent mismatch between London’s population (40% BME in 2011) and its police force which, while having made considerable strides in more diverse recruitment over the last decade, still has only just over 10% BME officers (though 17% amongst new recruits). This is said to produce problems in policing minority communities who feel alienated from the majority and in the case of black youths seem to suffer excessive numbers of stop-and-search interventions. There is something in this, but will crude quotas of the type Simon Byrne is proposing address the complexity of London’s diversity? Few if any American cities have quite the range of communities to which London is home.

‘Black and Minority Ethnic’ is a broad statistical category which has no real meaning: it is just an administratively convenient way of aggregating data. Just to take some obvious categories, London has large and distinct areas where people of Indian, Black Caribbean, Black African, Bangladeshi, Turkish and Jewish (not usually discussed in this context) origins are in a clear majority. There are also very large numbers of Pakistani, Chinese, Arab and Polish (again not counted in BME) spread more evenly across the capital, plus many smaller numbers of people from more than a hundred different countries, speaking many different languages. These communities have little in common with each other: they have far more in common, in fact, with the white British group whose influence the Met wishes to diminish.

Unless there is to be an attempt to set sub-quotas (a very difficult task indeed), therefore, the problems of alienated communities may remain even if positive discrimination takes off; a black policeman in a Chinese community has no obviously greater insight into that community than a white officer. And while persuading some groups to join the Met may be relatively easy, others are likely to prove intractable. Research has shown that London’s Indian-origin young people tend to reject policing as a career choice because they aspire to professional status. So do young people from a Chinese background. A degree-level recruitment scheme might attract some of them, but this would be less attractive to other groups who do much less well in formal education, such as Black Caribbean males. Pakistani Muslims report very strongly negative feelings towards the police on political/religious grounds. Bangladeshi women have an extremely low participation rate in the labour market and many do not speak English. Over half of Black Africans in London are first generation immigrants: 20% have arrived in the last five years, many specifically to work in healthcare, and police work is unlikely to interest them.

Recruiting somebody just because they fit into a statistical aggregate, therefore, may achieve relatively little in making different ethnic communities feel part of the mainstream. They are likely to provoke further criticisms of the Met as community ‘leaders’ jockey for position and consultancy opportunities.

Social engineering is always much more problematic than critics of our admittedly – unavoidably – imperfect society tend to believe. We could probably do a lot more to encourage minority groups to apply for policing jobs. Recently-retired Chief Superintendent Dal Babu has suggested that more police posts could be advertised for people (whether white or BME) with language skills, and that recruits could be sought from those who have done community work amongst minority communities. This latter suggestion seems sensible: all recruits, white or BME, would gain from working with people from other communities before they join. The Met could also make use of commercial recruiters tailoring their approaches to minority groups which are particularly under-represented.

However I remain unconvinced that quotas are necessary or desirable even in this particularly contentious area. Once started they will be difficult to remove, they have the potential to increase hostility between different communities, and are likely to spread to other areas, such as education and civil service recruitment, where the case for their use is far weaker.