Originally Posted by Aylmer
I agree that critical thinking is essential. But in order to be critical, you should have well-founded reasons to be skeptical and not just a convenient "gut feeling". You can't be critical of data, but not critical of your own assumptions or anecdotes. It isn't a one-way street.
So if you think that crime statistics may be tainted by underreporting, show me some hard facts about underreporting and how often it happens. It's that simple.
I can give you two real-life examples of underreporting crime. Among other things, I am in the process of fixing up (and overseeing, when it's stuff I can't legally do myself) one of my properties in a nearly all black neighborhood that used to be rough a few years ago, but is improving. Windows got broken on it while I was away in Quebec (clearly just vandalism by teens having fun -- I found the rocks in the house) and I didn't bother reporting. Then one of my NEW windows got again broken recently, after I replaced them -- still did not get reported.
Those two instances of crime against property (or whatever it's technically called) were not recorded, and therefore do not officially exist; they are not going to impact the official
crime rate of this black neighborhood.
If I'd bothered, I'm sure it would have changed some data somewhere. (And not done me any good.)
Now I know all the neighbors in that area, and they know me; and I've also got the lights on over there... and I'm hoping it won't be a problem now.
Multiply this by 100 or 1,000 or 10,000 and you can quickly have a discrepancy between reported crime and actual crime. Underreporting DOES happen. Now, please don't ask me to provide lots of solid data on underreporting -- you'd remind me of fellow SSPer Crawford when he was insisting that undiscovered voter fraud is undiscovered, therefore does not officially exist and therefore is not a problem at all because the data shows there's no voter fraud.
Underreporting exists (even your links acknowledge it as an important factor), that was my point. Does it explain everything? Maybe, maybe not. As pointed out already, it's obvious that selected immigrants (selected for their wealth and/or education) will be less crime-prone than people who are poorer and less educated; wealth and education are much better indicators than ethnic origin.