Listening to episode 93 of the House of Comments podcast Emma tells us in that 50% of people stopped and searched by police are black (or was it West Indian,? I can't be bothered to listen again) which is more than the population distribution, as if that is something we should be worried about. It struck me as a either typical left wing thoughtless use of statistics or deliberately misleading. Having listened to her and read her blog for a while now I'll go with thoughtless use of statistics.
The first thing we should be asking ourselves is where is Stop and Search carried out? We'd like to think it is carried out in areas of high crime where it is likely to act as a detterrent.
Then we'd like to know whether or not is was working by, say, seeing crime come down. If it is then great, do we need more of it to deter more crime or can we get the same deterrence with less of it?
If it isn't deterring crime then why not? Do we need more of it? Do we need to change the way its done?
Only now should we be looking at the colour of those being stopped and searched compared to the local population. Unfortunately I haven't been able to track anything down which looks at top and Search by detailed location and I suspect that would be informative.
I say that because I remember reading some analysis in The Economist some time ago where they broke down the Stop and Search statistics for Leeds. What they found, IIRC, is that Stop and Search was carried out in mainly high crime areas, as expected, and the main area was Chapeltown. At the time that area had a fearsome reputation for high crime and it was also had a large British Caribbean community. Again relying on memory the outcome was that although the proportion of Stop and Search was in line with the population statistics for that area.
Now lets turn this round and look from the other end of the telescope. The statistics tell us that we are stopping too many blacks, so what do we do? If we reduce Stop and Search and crime goes up that's not much comfort to the local community who are most at risk and I'd bet there would be lots of complaints from blacks and whites. If we reduce the numbers of blacks being stopped and searched and crime goes up it doesn't look good for local race relations. Suppose we stop more whites and crime rates stay the same? Now we are wasting time and resources and still demonstrating that there is a problem with black crime in that area, again a problem.
I'd really like to find some analysis by local areas to understand what is really going on rather than rely on emotive use of statistics out of context.