It probably would 
According to Levitt, yes. I’m oversimplifying slightly though. According to Levitt, the statistical risk factors are poverty, whether or not you have a single parent, the education of that parent, the person’s education, and something Levitt calls “wantedness,” which he explains in detail.
Several abstracts of Levitt’s work and and an abstract of the chapter in question are available on line. Freakonomics is a good book, as he doesn’t delve too much into the math, but rather the theory and the methodology he uses. He makes about as nice an explanation of regression analysis theory as I’ve read, and regression analysis is really the main theme of the book. Using various examples, he discusses what is essentiall Wring’s objection which is the multiplicity of factors that exist in any statistical sample. For example, we can take a look at Wring’s issue of jail populations. By itself it’s not useful, because people in jail can have committed more than one crime, not everybody guilty of a given crime goes to jail, not everbody in jail is guilty, and a person may have a higher probability of getting jailed for a crime dependant on factor’s like race, appearance, ability to hire a lawyer, or geographic location, as well as a host of other factors.
A regression analysis is a statistical tool that properly used, enables a person to single out the effect of a single factor from a series of interdependant factors at work in a complex and chaotic system.
Well, yes. He’s done work on this kind of thing. Not really up to Levitt’s quality. And Levitt basically bitchslaps Bennett’s superpredator theory.
Well, here’s the thing. I’m not talking down to you, but try to understand what I’m saying rather than jumping to a conclusion based on my first sentence, ok?
Bennett’s statement is accurate.
If you abort every black baby, the crime rate will go down. It’s really an unassailable fact, for the same reason that if you abort every Greek orthodox baby, the crime rate will also go down. If you abort every white baby the crime rate will go down.
If you reduce the number of people in the next generation of any group, down the road you have a smaller proportion of people who can potentially be criminals.
Also:
Levitt’s work shows us is that if you are born of a young, single, uneducated, poor woman, who does not want you, you are in the highest risk group to become a criminal.
If abortions are readily available, a young single undereducated, poor woman, who does not want and cannot readily care for her child is likely to take advantage of this availability and have an abortion. The result is, that since Roe versus Wade, relatively a much smaller proportion of the children in this risk group are being born. The crime rate drops proportionately as the highest risk pool for criminal behavior shrinks.
It just so happens that because of prejudice, discrimination, and the legacy of slavery, black people have the highest proportion of their population in this risk group.
We know that Bennett is familiar with this work of Levitt’s and we can reasonably surmise that Bennett is making a shorthand, and condensed reference to it when he makes his reduction to absurdity argument.
Intellectually, it makes perfect sense, and computes fine.
Where it doesn’t make sense is from the standpoint of sensitivity, politeness, and political correctness.
Now, personally, I would rather be accurate then sensitive. This however is quite a faux pas. Race is a hot topic and it makes sense to be sensitive, especially if you’re a public figure who’s views are widely disseminated.
Having read books by both Levitt and Bennett, I think I know where Bennett was coming from and what he was trying to say, and I beleive his intention is quite the opposite of racial insensitivity. His point is to show how outrageous such viewpoints are. So, I see nothing to be upset about.
But, this is a sensitive topic.
For some, simply the idea that crime rates for black people are significantly higher than the rest of the population is in itself an offensive, false, and bigoted viewpoint that needs to be attacked and refuted.
I can understand this. Such information, without context is fuel for racists. You only have to visit forums like Stormfront to see that such thinking is alive and well. People use this information to justify their bigotry.
While understanding this, I still disagree. That the crime rates are higher is a fact, and needs to be addressed. When we understand why the crime rates are higher, and what drives them we are confronted with statistical proof about the ongoing effects of generations of prejudice and discrimination as well as the legacy of the slavery. The higher crime rates for black people are not an indictment of black people. They are an indictment of society which condemns people to poverty, and a lack of educational resources even today, based on race. It is proof that we live in a prejudiced and discriminatory society.
The fact that these difference in crime rates vanish when we correct for these discriminatory factors is statistical proof that the bigots are wrong.
So, I beleive that hiding these things, or denying their existance is counterproductive.
Mr. Levitt had already made clear that race was not a component of his theory. Are you trying to say that Bennett misunderstood him, or was somehow oversimplifying the theory?
Here it is from Levitt himself, on his blog:
Maybe Mr. Levitt didn’t get the talking points in time. But I imagine the recent publicity won’t hurt his book sales.
Scylla I’ve stated very explicitly what my point is. over and over again.
that incarceration demographic data represents only the deomographics of those who are incarcerated, and cannot be reliably extrapolated to mean anything about the demographics of who commits crime.
I’m not talking economics.
Shodan - you’ve been an absolute ass to me here. I had expected better of you.
I’m not sure I fully understand the point of contention here, but I do know two things:
-
Nobody should ever expect Shodan to be anything but a completely useless prick.
-
Scylla is correct, inasmuch as he is saying that there are marked differences in crime rates between African Americans and Caucasians, and that these differences tend to be best explained by economic disparities. He’s way off in defending Bennett’s statement as in any way “correct.” It just simply is not. I’m not sure how anyone who claims to understand the first point could claim the second.
Also, wring is entirely correct that you cannot tell dick about population rates based on a selected sample, especially one selected for incarceration status. Anyone arguing otherwise is unspeakably stupid. If we aborted all the stupid fetuses, we wouldn’t have to deal with this stupidity.
My research group is heavily involved in just this type of study, although my own interests are more along the lines of constructs from clinical psychology (the disruptive behavior disorders, for example). One interesting analysis that is being conducted examines two cohorts of boys followed up for many years. The two groups are racially similar (both containing primarily African American and Caucasian children) but show different rates of criminal and violent behavior. These differences are clearly explained by economic factors withiin the community, which went through some notable changes during the period which the boys were followed.
Likewise, Jane Costello, at Duke, published a nice paper demonstrating the marked reduction in conduct problems that occurred within a particular American Indian community. She had longitudinal data before, during and after the community established a casino and spread wealth among members of the community.
In my own work on Oppositional Defiant Disorder and Conduct Disorder, there are differences in the rates of these behaviors between racial groups, but once again, they disappear after you control for economic and other factors.
Oh, horseshit. You keep repeating your own point and refusing to address anyone else’s.
Let’s try it one more time -
Is the following statement true, false, or unproven?
Why, or why not?
Regards,
Shodan
While I agree with your arguments in this thread, this segment of your post puzzles me. The only way I can make sense of it is if you are considering the short term while the existing population of the proposed group remains. That would reflect on the fact that crime predominates amongst the youth and younger adults.
If that is not what you meant then you might explain.
If we assume for arguments sake that whites and blacks each make up 50% of the population of 100 people and that amongst that population, there is usually 4 black criminals and 1 white criminal then the overall crime rate could be described as 5%. Now if you abort every white baby and the white population disappears, you have a total of 50 people, all black and 4 criminals. Now you have a higher crime rate of 8% for the total population. If you abort all the black kids into oblivion, then your crime rate goes down to 2%
Somehow I think that is what Bennett had in mind. To be clear, all we are doing here is playing with simple math.
-formerly grienspace
wrong. on the first. I’ve made the exact same point. You keep tossing other interpretations of my point (“you think whites cause more crime” sort of shit). The only thing I care about in this thread is to get folks to stop using incarceration data as a premise to extropolate to “racial characterisitics of who committs crime”.
Your other stuff wrt economic issues was not anything I’d investigated, claimed, cared about (except as an incidental noticing FME that folks w/public defenders did worse in sentencing etc. than others w/paid attorneys). I asked over and over for the data you wanted me to look at, you finally gave some up, I started to look at it and you again acted jerk like in my direction.
Since the economic factor was never really of interest to me, and not at all necessary for the point I was making, I decline to get involved in it at all. The only issue I cared about was trying to get folks to cease attempting to make the link between demographics of incarcerated folks and demographics of those committing crimes.
See - selecting non random sampling of a group means you should not expect the same results for the larger group.
There’s a number of dorms on MSU campus. Should I take the demographics and GPR’s of the folks living at Wilson Hall (predominately folks on the sports teams) and attempt to claim that they’re representative of the whole student body? No and no one would attempt to argue that.
But take the subset of “incarcerated” folks and everyone and their brother seems to believe that it somehow magically rendered itself into a random sampling of folks who commit crimes.
Are you going to answer the question?
We all understand your point about incarceration. That is why I have repeatedly stated that the notion that blacks are disproportionately involved in crime is not based solely on incarceration data. All measures of the crime rate - not just the incarceration figures - all measures point to the same conclusion. Blacks are disproportionately involved in crime at a level beyond their representation in the general populace.
Thus, Bennett’s hypothetical is correct. If you eliminate the part of the population that is disproportionately responsible for crime, then the crime rate decreases disproportionately.
Sheesh! You talk about me acting like a jerk - you keep making the same argument, over and over, and refusing to address the fact that the thread has moved beyond it - and it didn’t have that much to do with the OP to begin with.
Regards,
Shodan
Black fetuses are responsible for exactly 0% of the prevalence of crime. Aborting them will have no effect on the rate of crime.
Secondly, if you understand that economic factors tend to explain the disproportionate rate of criminal behavior within racial groups, why do you continue to argue for a racially motivated strategy for altering future crime rates?
But this line of defense for Bennett goes against what both Levitt and Bennett have said.
Why did Levitt say the race-based crime data is “generally deemed unreliable” if it is an unquestioned fact?
How can anyone here defend Bennett’s remarks based on what Levitt had written, when Bennett clearly stated he disagreed with Levitt’s theory?
The first link is the debate between Levitt and Steve Sailor on Slate. If you read the whole debate, you’ll see that it was Sailor who used race-based examples. Bennett disagreed with Levitt’s theory, just as Sailor did.
Maybe Bennett confused Levitt with Sailor. Maybe Bennett still had the “let’s smear black people because they made the government look bad by expecting help after Katrina” talking point on his mind. Maybe it was pandering to a racist audience. Maybe it had more to do with furthering a pro-life agenda by inciting liberals to speak against abortion.
Maybe it was just a thoughtless off-the-cuff remark. Who can say for sure? And who cares? No one is going to lynch him, the government isn’t going to force him off the air and thwart his right to speak. Why does he need defending?
What I take issue with is how the “Black people are naturally more criminal and violent” meme has been slipped into the discussion by the Bennett defenders (with a wink and a nod to “accuracy over sensitivity”).
It’s disingenuous, and passive-aggressive racism, and a red herring that hides the very real and dire problem of institutional racism in the criminal justice system.
Off-topic: Zeeny, what will it take to convince you to go from “guest” to “member”?
Thank you for saying that. I’m worried about two things - this board is addictive, and I fear getting kicked out because my (internet) mouth is too big. My husband isn’t happy about the possibilities of either one, and would never let me forget it if I signed up leading to a bad consequence. 
But I’m working on it. 
Which question? the one about economic data? see, please my post immediately before yours.
not from the data I’ve seen. Not from data that’s been posted here. In fact (see post below yours) there is the statement "None of our analysis is race-based because the crime data by race is generally not deemed reliable. "
the data you’ve pointed at (incarceration, and some vague reference to arrest rates) is not reported as “x category is responsible for y number of crimes”. It just ain’t. no matter how often you attempt to make it say that, it is not reported that way. And even if it was, you also face the inherant problems of unreported and unsolved crimes, and the fact that the sample is anything but random. Any number of factors skew the factors, all of which has been painsteakingly pointed out to you.
nope.
The thread has not seemed to have ‘moved beyond it’, since in this post you continue to assert that one racial category is responsible for “more crime” than another. Did it have anything to do w/the OP. Why yes, it does indeed. The “it” of the OP, of course has to do w/Bennet’s statement which contains once again the implicite assumption that some how or another we can positively assert that one race is responsible for more crime than another. and here we are again.
So, once again, my very narrow interest in this thread is to try show people like you, who continue yet in this same post to assert that one race is responsible for more crime than another, based on data available that y you’re wrong to do so.
In short:
there is no compilation of data that goes:
Person: Race: Number of crimes responsible for:
them adds 'em all up. and even if there was, you would have to somehow factor in those cases where a single crime was committed by several people…
Not to mention the problems w/(again) the fact that this would still only have the data on the solved crimes (an overall very small portion of all reported crime, which is also a small portion of all crime, and certainly not anything remotely like a random sampling). None of which even touches the other concepts pointed out like targeting certain population and areas for increased scrutiny by the police, etc etc etc.
This one -
Is the following statement true, false, or unproven?
Regards,
Shodan
I’m not sure what started you on the kick, as I don’t think anybody was arguing incarceration data until you brought up the subject,
But, since you do bring it up, you are mistaken on both counts.
When you discuss what data can be extracted from a given sample, you are discussing economics, or at least the tools and purview of economics.
Saying that you cannot extrapolate “anything” about the demographics of who commits crime from prison demographics is simply a falsehood.
There is a causative correllation between being imprisoned and having been convicted of a crime. There is a causative correllation between being convicted of a crime and having committed one.
There are of course other causative correllations concerning imprisonment, but you seem to be under the impression that because there are other factors at work they somehow poison the well and no data can be extracted. That’s simply not true. Economics concerns itself with extracting such data. I earlier wrote about how this is done with regression analysis, but I hate to repeat myself. The reliability of this data depends on what you try to pull from it and how carefully and rigorously you do it.
On the other hand, it is of course true that the raw data of prison demographics is not necessarily demonstrative of who committs crime.
wring, I’m a little unclear as to your point, but there are many data sets that involve random community samples, that employ self-report as well as official records, that look at arrests, charges and convictions. They do look at prevalence of crime by racial categories, but fortunately, they go far beyond that as well.
If I’ve misunderstood your point, however, I apologize.
Well, Levitt, the author of Freakonomics agrees with me:
From:
http://www.freakonomics.com/2005/09/bill-bennett-and-freakonomics.html
"If we lived in a world in which the government chose who gets to reproduce, then Bennett would be correct in saying that “you could abort every black baby in this country, and
your crime rate would go down.”
Yes.
For simplicity’s sake we’re assuming that these abortions do not materially affect the total size of the population to any significant or measurable degree. This may or may not be true, but it makes for a simple example.
That’s a good question and a bit of a head scratcher. I really can’t answer as I don’t know what’s going on in Levitt’s head. It appears to me he’s being inconsistent.
He says here:
http://www.freakonomics.com/2005/09/bill-bennett-and-freakonomics.html
"It is true that, on average, crime involvement in the U.S. is higher among blacks than whites. "
Which seems contradictory with the statement you’ve quoted. If the data’s unreliable, how can he say that. Earlier I cited a work of Levitt’s dealing specifically with race, crime, and arrest rates in which deals with these statistics and extrapolates data from them, and he and his research partners, particularly Roland Fryer at Harvard have done some incredible work with racial crime statistics.
So, like I said, good question. I wish I had an answer for you.
I guess if people were always consistent the world would be a much better place. I think Bennett does accept aspects of Levitt’s work, and probably the conclusion that abortion rates affect crime. What he doesn’t seem to accept is that making moral decisions from economic data is a good idea. This last may be a bit of a strawman as I haven’t seen Levitt make that argument.
Then again, Bennet’s superpredator theory was basically destroyed and its corpse cut up and thrown to the dogs by Levitt, so maybe there’s a little bit of pissiness in Bennet wanting to disagree with Levitt on principle while he makes use of his conclusions.
In other words, I have no idea on this one either. Of course, this doesn’t mean that they’re not good questions.
Beats me.
“Accuracy over sensitivity” is my line, so it looks like you are referring to me. I have never said “Black people are naturally more criminal and violent,” as you quote me. In fact, I have argued the opposite.
Who has said this? Who has made this argument that you quote?
It’s a strawman. Nobody has. I expect you to retract this accusation or show this quote in context.
Nobody has made this argument here and it’s a shame you feel the need to make false accusations, especially in the midst of an otherwise substantive post.