William Bennett: "If you want to reduce crime...abort every black baby in America."

The BBQ Pit
221

Let’s say we have a class of 30 kids. Twenty are white, ten are black. Five of the white kids are cut-ups. Five of the black kids are cut-ups. So what that means is that:

50% of the black kids are “criminal”, representing 50% of the problem element.
25% of the white kids are “criminal”, representing 50% of the problem element.
30% of the class is criminal.

As the teacher, you decide that you’re fed up with all these cut-ups. Because the black kids are the biggest source of the problem, you decide to kick them all out of your classroom. Get your black asses out of the classroom, black kids.

So now we’re left with twenty kids. Five of them are cut-ups. Now, only twenty-five percent of the class is the problem element. We can rest easy now.

BUT…

Let’s say in the classroom next door, I have a class of 30 kids. Twenty are white, ten are black. But only two of the black kids are cut-ups. Four of the white kids are cut-ups.

20% of the black kids are “criminal”, representing 30% of the problem element.
20% of the white kids are “criminal”, representing 70% of the problem element.
Overall, 20% of the class is criminal.

You decide to come to my class and kick out the black kids like you did in your classroom. Now we have twenty kids, 20% which is “criminal”. Status quo.
So yes, assuming that black criminals are disproportionately represented (as was the case in the first example) and that the criminals and non-criminals are independent entities*, then eliminating the entire black population will lower the proportion of criminals in the total population. But that diminishment will–heh–diminish as you approach proportional representation.

*This is a huge assumption to make. In the classroom example, you might be kicking out the classes’s biggest law enforcers: the tattle-tellers, the admirable role-models, the teacher’s pets (aren’t they the ones who always get to “take names”?), and the Protectors of the Underdogs (you know, the boy who beats up on the class bully every now and again). These individuals may keep the “problem element” in check. There’s no reason to believe that the “law enforcement element” must always be more evenly distributed than the “problem element.” Kick out the “do-gooders” along with the “bad doers” and who knows might happen to your class? Maybe the proportion of your bad apples will go from 20% to 30%?

Yeah, I’m over analyzing the question a LOT.

222

I guess that is Damocles’ sword. Do you think that I am a “race pimp” trying to smear William Bennett or that I am honestly relating to you how I perceive his comments? Did I not acknowledge his disclaimers and consider the context of his hypothetical before using my judgment? Did I in any way misrepresent his views by saying that he wanted to abort all black babies?

Did you in any way misrepresent my views by saying that I support the killing of Jews to reduce prices? Did you likewise consider the context of my hypothetical when you said the following:

What I actually said was:

I submit that you deliberately attributed to me a stereotype which I do not hold but that I used because it is offensive and I am aware that many people do hold that stereotype. That is unlike William Bennett in that it is he himself who holds the stereotype that cutting back on the number of Blacks born would lower the crime rate.

I did not change his quote or the meaning of his quote. You took mine out of context and changed the meaing.

You are hereby sentenced to being a Republican Conservative for the rest of your life.

223

Not that there’s anything wrong with that. :wink:

224

Do you accept the statistics that say american blacks murdered at higher rate than American whites?

If so, what do you think is explanation? Most obvious explanation is that American blacks tend to live near other American blacks.

It hard to imagine that American government is exxagerate murder rate for American black.

225

So much for that argument a while back that native speakers were incapable of making an ungrammatical utterance.

226

I hadn’t thought of it that way. Thanks for a new viewpoint :slight_smile: And sorry for the hijack.

227

No. In neither statistics not logic does the existence of a correlative relationship mean there is a causitive one. If 10% of the population of Group A likes ice cream and 50% of the population of Group B likes ice cream, Group A+B likes ice cream at a rate (assuming A and B contain the same number) of 30%. If you indiscriminantly elininate a sizable enough portion of Group B—ice cream lovers and non-ice cream lovers alike—the rate of ice cream liking among Group AB drops. This is true whether liking ice cream is genetic or due to outside influences.

I know of no one who has concluded such a thing. I know I haven’t. I think you are reading things into this, allowing yourself to be steered by your fears. But the facts remain. The correlation stands. And it is mute on the issue of causality.

228

Uh… smartguysmile doesn’t have a location, so what makes you think he’s a native speaker? I don’t think he is given his syntax patterns.
Back on topic - Bigot or not, Bennett should take this as a sign that he needs to work on the whole “think things out before uttering them” thing. I bet the station wishes the show wasn’t live.

229

Your example holds only if the percentages in each group are independent of one another. In other words, there very well could be a relationship between members of Group AB that isn’t borne out by the rates of Group A and Group B alone.

Possibly, but it isn’t entirely clear from the example Bennett gave. The logic of his statement would have been much clearer had it been in the form - If A, then B, given C.

230

No, you did not, and yes, you did, respectively.

Yes, I did.

Nope. It is not a stereotype, but an empirically established fact.

Sauce for the goose, sauce for the gander.

Oh, come on. You don’t think records are kept of the number of arrests and convictions for each person who enters the criminal justice system? The suggestion is absurd.

And I am not considering only incarceration rates in determining the crime rate, so your caveats about imprisonment vs. probation do not apply. Blacks are arrested at disproportionately high rates, and convicted at disproportionately high rates, as well as imprisoned at disproportionately high rates.

And again, by your standard you cannot establish any link between any of the other factors blamed for crime rates - not poverty, not single parenthood, nothing. You can, however, determine that racism is not responsible for differing attributed crime rates between blacks and whites - by holding other factors constant and examining race. When you do that, the correlation between race and crime disappears.

This would not be the case if racism were the cause of the differing crime rates. Racists do not care if the members of the hated class are the children of single mothers, are rich or poor, whatever. If racism were the cause of the differences in crime rates, middle and upper class blacks who were the product of two-parent homes would be arrested and convicted at the same rates as any other blacks. Mark Fuhrman[sup]*[/sup] was not alleged to have framed poor, illegitimately-born blacks only, or exclusively.

But they aren’t. If you correct for other, non-race-related factors, crime rates do not differ between blacks and whites. Thus, either white racists have magic insight into the backgrounds of their victims, or race correlates with higher crime rates, but does not cause it.

Which was Bennett’s point all along.
Regards,
Shodan

[sup]*[/sup]Ten years ago today, a black man walked for a double murder. You wanted to talk about “unsolved” murders that are really committed by white people instead of blacks? Go help OJ on his life-long hunt for the real killers. :wink:

231

I’m not sure I follow. Could you explain this a little more, possibly with an example. Thanks.

Possibly. But it seems pretty clear to me. To misconstrue his words or his meeting takes effort and ignores logic.

232

Crime rate for population A is X%

Crime rate for sub-group I of population A is Y%
Crime rate for sub-group of population J is Z%

If I and J are independent entities, then it follows that adding the crime of both groups together (Y and Z) should equal X (assuming the same number of people in sub-groups I and J).

However, I and J aren’t necessarily independent entities constituting population A. In other words, dividing population A into sub-groups I and J is arbitrary (not entirely, of course; but there are other ways of sub-dividing population A into sub-groups - say G and H, that might allow for the assumption of independence to hold stronger. Rather than subvidive according to race, subdivide according to income or marital status, for example).

Another way of stating this is that - your previous example holds if the liking of ice cream of Group A is independent of the liking of ice cream of Group B. It wouldn’t necessarily hold if the liking of ice cream from both groups is dervied from the liking of ice cream for the entire population (AB). There could be some underlying relationship between A and B affecting each group’s liking of ice cream.

Well, it wasn’t entirely clear to me (as well as to others). He could have helped matters by clarifying what he meant.

233

I think what you’re saying is simply that there might be other ways to divide the group that would get more to the heart of the matter, income, for instance. If so, I think you are correct. But just because a stronger correlation may exist—one that might even be closer to causitive—doesn’t invalidate any other correlation that might exist.

Could he have made his point better? Absolutely. Was it a racist comment, in the judgmental use of the term? No.

234

He advocated it? Cite? And please don’t forget to add what he said after his hypothetical.

The shrillness from dopers against an academic hypothesis actually surprises me. I guess I overestimated the intelligence level of this board.

235

I just cannot get worked up over an off-the-cuff remark that, as a very superficial analysis of the crime stasitics is correct, but almost certainly is not correct with a much more nuanced and analytical study. I expecially cannot get upset when the example was meant not as a means to support a policy, but as an attempt to refute the use of statistics for policies when there are other factors (such as morality and ethics) which would mitigate those policies.

And trust me, I am no conservative.

236

Shodan - once again you misunderstand. Of course, corrections keeps those records.

BUt the data that is reported and everyone points at does not contain that information.

When you (or anyone) looks at prison incarceration data and points at the relative percentages of people, they point at x % white, y% black, then compare the stats to overal population, then make the statement that blacks are responsible for “more crime”.

And (follow with me) that is not fucking supported by the data presented.
Two bodies in prison may represent 1 crime, 2 crimes, zero crimes, 1000 crimes, and you simply do not know from the incarceration data.

what you can legitimately state is that blacks are over represented in prisons.
Period.

You cannot go further than that and make any reasonable claim that they’re responsible for “more crime” (or ‘proportionately more crime’) because the data is not presented "how many crimes to these people represent. Further, incarceration data ignores all of those folks on probation. and the unsolved crime. and the unreported crime - all of which have unknown percentages of race.

get it?

Blacks are over represented in prisons.= supported by the data available.

Blacks are responsible (proportionately) for “More crime” than whites = unknowable therefore unsupportable by the data available and publicized (yes, one could potentiall gather up all those individual records and add 'em up, but no one is doing that, and I don’t foresee that as a probability in the future).

237

My comment had nothing specifically to do with Bennett’s comment. It was merely a response to fushj00mang’s ridiculous attempt to incorporate abortion politics into his “analysis.”

238

Sorry to get in this one late and have this discussion go on so long, but Bennett was pretty much in the clear on this one. His whole point was that you could reach morally abhorrent conclusions if you used faulty logic based on gross, aggregate statistics. In other words, his point was CONTRA racism. Especially given that he’s against abortion period, the hubbub here is mostly knee-jerk and then trying to defend the knee-jerk.

239

Also please see the internationally recognized definition posted by crowmanyclouds (welcome, by the way!):

(emphasis mine)

240

My pleasure. Take care. :slight_smile: