> left wing Euro politicians control access to the data and are locking it down. In the same sense as left-wing and right-wing politicians everywhere control a

> left wing Euro politicians control access to the data and are locking it down. In the same sense as left-wing and right-wing politicians everywhere control access to crime data and are locking it down. > if the data supported that claim, wouldn't they release it rather than hide it? Let me describe for you a hypothetical world. I don't know how closely it resembles the real one. In this hypothetical world, Muslim immigrants are 2x more likely than other people to commit violent attacks such as rape, make up a small fraction of the population, and are widely hated and feared; violent attacks on Muslim immigrants are 3x more common than violent attacks by Muslim immigrants; the fear many people feel of violence from Muslim immigrants is -- if, e.g., you calibrate it against other kinds of violence they fear -- far out of proportion to their actual risk of being attacked by Muslim immigrants. So, in particular, if the public knew and understood and internalized that 2x figure, they should actually be reassured by it because the risk they face is very small; but, just as in the real world, the public in our hypothetical world don't always respond with perfect rationality. Now imagine you are a politician (of the left or the right, I don't care), you're aware of these numbers (or have guessed at them), and you are trying to decide whether published crime statistics should break out Muslim immigrants as a separate category of offender. I suggest that you are likely to think as follows: "If we do this, every right-wing anti-Muslim anti-immigrant group in the country is going to seize on that 2x figure and make hay with it. It will widely be taken as confirming fears that Muslim immigrants are a terrible threat to public safety. Violent attacks on Muslims will go through the roof. Lots more people will be hurt and killed, and any sort of integration will become more difficult. Which will probably lead to more crime by Muslim immigrants in the future, too. Even if by some magical process releasing those numbers made rape by Muslim immigrants completely go away, it would still lead to a lot more harm than good." I'm not saying you'd be right to think that. Maybe actually "better more data than less data" is a universal principle that trumps everything else. Maybe releasing the numbers wouldn't really have such bad effects. Maybe it actually would somehow lead to a big decrease in crimes committed by immigrants. I don't know. But the point is that I just sketched a perfectly plausible process by which politicians might decide it's better not to publish potentially inflammatory numbers, even if those numbers indicate that anti-immigrant politicians and journalists exaggerate the actual magnitude of the threat. > If I'm wrong and Muslim immigrants to Europe actually commit crimes at the base rate I doubt that. I bet Muslim immigrants to Europe are, e.g., systematically somewhat poorer than the population as a whole, which would probably mean more violent crime and (e.g.) less embezzlement even if there were no cultural differences at all. I bet they drink less alcohol, which would probably mean less of some kinds of violent crime and e.g. far less DUI. Etc. I'd be astonished if there weren't all kinds of differences in the rates at which Muslim immigrants and (say) Christian multigeneration natives commit all sorts of crimes. > it looks like there is a real problem, and it looks like it's being hidden from us. Suppose someone has a theory that financial misconduct like fraud and embezzlement and insider trading are much more widespread among the rich than in the general population. If they point out -- correctly, so far as I know -- that official crime statistics don't break out the richest 1% of the population when reporting rates of these crimes, are they justified in saying that "there is a real problem and it's being hidden from us"?