Using DOJ stats to prove or disprove the criminality of a race is a crock of feces

Great Debates
21

http://boards.straightdope.com/sdmb/showthread.php?threadid=119382

This Pit thread is what inspired this GD thread. This was just the last one in the “blacks are criminals” genre of threads. The train of thought that leads people like BS and others down the “well everybody knows blacks are criminals and whites just ain’t” starts with these DOJ stats.

I just can’t understand how anyone with even just a little bit of critical thinking skills can look at these arrest and conviction numbers put out by our government and not question them. Hell, not only are they not questioned, but they are trumpeted as proof of the inferiority of minorities.

Another tangent used is the “poor commit more crimes and blacks are poorer than whites” thinking. Now, I do believe that this is true, but this does not begin to explain the disparities in arrest and convictions. Which is the reason why I chose drug stats to highlight my argument.

All the drug surveys show that minorites do more drugs than whites percentagewise. These differences can be explained in socio-economic terms. There is a 1 to 2% difference in drug use rates between blacks and whites. But, fuckity jam on potato crisps, how do you explain the disparity in arrest and conviction rates?

22

I believe that the poor commit more crimes proportionally speaking then do the middle or upper classes. I also believe that this is a primary reason why blacks are disproportionately represented in DOJ statistics as they are also disproportionately poor. In addition I also believe that the DOJ stats reflect the bias inherent in our society which exacerbates the disproportionate report rate of crimes among minorities. However, even if there were no bias inherent in society nor reporting I feel that minorities would still have disproportionately high rates of crime by reason of them being disproportionately poor.

Having said that, and assuming that there is a degree of bias in the DOJ figures, is it your contention that the DOJ statistics are completely unreliable and are therefore worthless in regards to categorizing crimes by race?

Grim

23

Shodan You want a cite that says "yes, although all of these people are classified as accidents/disappearances, they really are murders’???

Ok. how about this: Chandra Levy was classified as a missing person until her remains were found.

The same may be true of any other missing person.

There was a famous case in MIchigan where a woman was found dead, her husband claimed it was a fall from her horse. And so it was classified as an accident. Her death was later ruled to be a homicide.

the same may be true of any death ruled an accident.

These do happen, and not infrequently.

My point is that anytime that you’re making a statement about ‘the number of murderers’ you really need to clarify it as “number of known murderers” etc.

So, once again "you don’t have any way of determining how many actual crimes occur -in any category- so at best all you can demonstrate and look at is ‘how many crimes were reported, solved, persons convicted for’. So the extrapolation is not X race is responsible for y% of the crime, at best it’s "X race is convicted for y% of the reported and solved crimes.

24

Pretty much, yes. Let me reiterate:

The Feds own numbers say that 15% of drug users are blacks and 84% are white. The percentages or whites arrested for drug abuse crimes is 64% as opposed to 35% for blacks–again, the Feds own numbers. This makes them pretty much unreliable in my book.

We haven’t even started talking about conviction rates.

25

The disparity in rates of arrest for drug crimes is explainable by law enforcement tactics in dealing with drug crimes. Poorer inner city communities are the main places where open air drug markets exist, policing aimed at these drug markets will catch predominatly poor people in these neighborhoods, which goes without saying will have heavy black and hispanic populations. However, when you change tactics slightly like in Oakland’s Beat Feet program, which targets users (cars impounded and sold), you get a much more representative arrest rates. On a weekend in the past year Oakland Police conducted a sting, 80% of those arrested/cited were in fact white users coming into Oakland from surrounding communities.

I’m pretty sure I cited this before in a different blacks are scary and evil thread thread, but I’ll look it up if requested. Btw, this is not to say that blacks sell more drugs than whites either, it’s again the way enforecement works. For obvious reasons, drug sellers in the suburbs operate differently than those in the inner city, therefore aren’t as likely to draw police attention.

26

wring -

Your examples hardly rise to the status of “anecdotal”.

Come on - you claimed that there were a significant number of murders misclassified as accidental deaths - significant enough that the DoJ figures were meaningless for determining that blacks murder in numbers disproportionate to their representation in the population.

In order to prove this, you have to show

a) that enough of the accidental deaths in the US are actually misclassified murders to affect the figures, and

b) that a much higher than 50% rate of these mis-classified deaths are actually murders committed by whites.

To prove this, you cite two examples. Two.

Fine. I will then counter by mentioning OJ Simpson. This is a case in which everyone not blinded by prejudice or flat-out stupid knows is a case of two murders committed by a black man. They remain, however, officially unsolved.

If you are going to take anecdotes as evidence, my anecdotes prove that at least 50% of the “unsolved” murders in the US are committed by blacks - just like the solved ones. Since neither of your cites mentioned race, we do not know if Chandra Levy’s murderer is black, white, or Martian. Nor in the other case you cite. So as far as the four specific ‘unsolved murders’ mentioned on this thread, at least 50% - possibly 100% - were committed by black men.

You can do better than this. If you want to be taken seriously, you have to.

Simply to dismiss the figures because “everyone knows they’re racist” is a cop-out. They are the best figures available, and they show a clear trend. Dismissing them out of hand because they do not back up your allegations is arbitrary.

If you simply want a caveat added to every cite of the DoJ figures that says “these cover only the ones we know about”, fine. If you want to base an opinion on the unknown murders in the US, feel free to do so, but you aren’t basing it on anything anyone knows.

Biggirl’s OP seems to be saying that higher arrest and conviction rates of blacks are a function of racism. That is a possible assumption.

Prove it.

Regards,
Shodan

27

Shodon - nope, I never claimed the number. What I did say (repeatedly) is that when people talk about how blacks commit more crime than other races and use the DOJ stats to back it up, they’re basing that statement on an incorrect analysis of the data base.

All crime is not reported.

all crime is not solved.

prison #s = those people convicted for known crimes.

The subset of ‘drug crimes’ (for the set all crime) was brought up by BigGirl. You brought up the subset (for all crime) ‘murder’, trying to get a similar picture.

I demonstrated that we still do not know ‘all crime’. You asked for ‘proof’ that there are missing persons/accidents that are actually ‘murders’. I pointed out some examples. Do I know the actual number? no, and neither does anyone else.

I have long attempted to argue that **no one will ever, ever ** be able to substanitate any claim relative to actual proportions of people of any demographic who commit crimes (because nowhere near all crime is reported, nowhere near all crime is solved), and that any assumptions based on subsets of the data (all crime that is reported and solved in this or that category), will suffer from the same problem - an incomplete analysis based on insufficient and non random data in the data base.

While it’s true that one doesn’t need to analyze a full set of data in order to make some reasonable assesments of trends and so on, one must however have some assurance that the subset of data that one examines is as free from bias as possible.

and it just ain’t possible with crime rate data.

28

I don’t know whether the “prove it” part was directed at me or wring, but, as I’ve already stated, the reason I chose drug abuse rates is because we have independent numbers to show how drug abuse breaks down by race.

If my contention is that the DOJ numbers are skewered because of racism, I don’t see how using the skewered numbers proves that they are not skewered.

There are more black people arrested and convicted of murders because there are more poor blacks than poor whites and poverty has been shown to increase the crime rate, however this factor alone does not explain the disparity in arrest and conviction rate. You can’t just turn around and say "well, the DOJ stats show that blacks get arrested and convicted of murder at a much higher rate than whites, so therefore the DOJ numbers are not skewered because of racism.

I am using the drug rates and extrapolating. Explain how I can’t or shouldn’t extrapolate or how the drug abuse numbers are not a function of racism.

29

It seems to me that BigGirls original objection (and a valid one IMHO) was her perception that blacks were being characterized as somehow more apt to commit crime, simply by virtue of being black. I hope that I’ve demonstrated that I don’t think that this is the case at all and that my stance is that crime is more related to socio-economic factors NOT some inherent “evilness” factor in blacks. Since the OP though BigGirl’s and wring’s posts seem (correct me if I’m wrong here) to have switched to the stance that “DOJ figures are completely unreliable having been biased to a great degree by racism”. This second argument is very different from the first and is the one I take exception to.

I don’t know about this wring. Yes, it’s true, we have incomplete data on crimes perpetrated and it will always be this way. There will never be 100% accuracy in reporting all crimes that have occurred. However I disagree that the DOJ values are valueless simply by virtue that the subset of data gathered is not completely bias free. The question is essentially is:

“To what degree is the DOJ data skewed by bias in the gathering of said data (by virtue of officers pulling over blacks disproportionately, pursuing crimes committed by blacks more aggressively, etc.) as well as the gathering and reporting of said data?”

Once this question is answered we can determine how accurate DOJ figures are in regards to perpetration rates among minorities. To simply decide out of hand that the degree of bias (an acknowledged but unknown value at this point) in the DOJ figures is great enough to make the entirety of the DOJ race stats meaningless, is IMHO, premature. To me it suggests a willingness to deny that there is some validity in the data even though the conclusion may be distasteful to ones preferences.

Some questions:

  1. Do the poor commit crimes at a disproportionate rate then do the middle and upper classes?

  2. Are blacks disporportionately in the poor classes?

  3. If 1 and 2 are answered in the affirmative is it reasonable that DOJ figures show a disproportionate amount of crimes committed by blacks?

  4. If 3 is also an affirmative what is it about racism in our society (at this point an admittedly unquantified value as applied to crime stats) that makes the DOJ crime stats valueless? Especially when we would already assume (as per 1 and 2) that there would be a disproportionate number of minority crimes by virtue of being disporportionately poor.

Grim

30

I’ll try and answer what I think you’re saying Grim.

Even if #'1 & 2 are true, there’s additionally the unintentional bias that results from bad law.

We’ve noted certain laws that while on the surface are not biased, have the effect of targeting minorities or those living in poorer neighborhoods. This results in more arrests/convictions for minorities, minorities filling the prisons, people looking at the prison numbers and seeing the relative racial characteristics, which then get built into policies, leading to profiling, leading to a larger skew of the results, and the corresponding assumption on the part of many that certain ‘types’ of people are more likely to commit crimes (it’s easier to look at a large group of people and guess their racial background than socioeconomic, plus the DOJ figures give out data on race, not S/E).

I’m not saying that the DOJ stats are ‘valueless’, but that they shouldn’t be used as a predictor/basis for substantiating the need for racial profiling, or conclusions about racial characteristics of those who would most likely commit crimes.

Consider this:

The crime of drug possession is most often the result of a search done in an ancillary fashion to some other event (traffic stop etc.). IOW, in order to find the stuff, the cops already have to have a reason to search you.

At any given point in any given day, there’s likely to be more than 50% of the students at the local college campus in possession of some illegal substance (booze for the underage, drug paraphenalia, drugs, etc.). however, when the local narcotics squads are looking to bust folks, they don’t go to the campus, they’ll hang out at the inner city 7/11 type stores and look for people on foot, ask if they’re carrying, know some one who is, etc. ( One of my clients got asked while he was waiting for the bus. ) Gee, which place is more likely to have minorities? Who then, do you think then gets arrested/convicted?

I’ve worked in the system for 25 years. There were some ‘givens’:

If you were a minority and were charged w/a crime in certain areas of my state (roughly above a line mid mitten, plus a select number of counties in the lower part, for different reasons), the liklihood of you being sent to prison was pretty damn high. A white from those same areas, not so much. In a local rural county, police called to a domestic dispute, in the home where the whites lived, they left w/o arresting anyone, in the home where the minority woman lived, she left in handcuffs.

I talked to a local prosecutor recently, about a case where a minority guy was above that county line, and his response was “what was he thinkin’?”. That’s how pervasive it is.

So, when I hear folks talk about how minorities are more likely to be criminals, 'cause gosh darn, just look at the prisons, my first thought is “self fufilling prophecy works yet again”.

Yes, I believe that socio economic factors have an enormous effect, but not just the old adage about poor folk stealing more often (I"m not convinced that’s necessarily true - grew up in middle class America and I was one of the few of my contemporaries that didn’t shoplift at least a couple of times). But it’s not the only factor.

31

**
Again, I’ll repeat (everyone must be tired of seeing me type this)This is why I’ve been using the drug stats. We have established that blacks do more drugs percentagewise. We even have the figures-- 1 to 2% percent more. I’ve also said, though have not given any evidence of this, that this descrepancy can be explained with socio-economic factors.

I’ve also established that, although whites constitute 85% of the drug using population, the are arrested for only 68% for drug abuse crimes. Blacks, on the other hand, are 15% of the drug using public yet represent 35% of the arrests for drug abuse.

They are more than valueless when used to describe the difference in the criminality between minorities and whites, they are harmful. They feed on the perception of the criminality of minorities-- especially blacks-- perpetuate it.

Now I have a question for you. Do you think that socio-ecomomics explains why a 2% difference in usage translates to a 30% difference in arrest rates?

32

If anyone has access to a university database and can find an online copy of Everybody does it : Crime by the public (Gabor T; 1994; Toronto CA : University of Toronto Press), please put up your hand. I don’t have the actual book here, only my course notes, and I don’t want to start quoting stuff without being able to refer people to the original sources if I start quoting data here.

I would, however, like to point out that the so called “index crimes” (which are essentially crimes of violence and serious property crimes) are not going to show the same demographics as white collar and corporate crimes and that when we are talking about overall crime rates we need to make sure that we are including ALL classes of crime (this is a great deal easier said than done).

33

To give people an illustration of what I’m talking about, it’s a pretty major deal here to deprive our Taxation Office of revenue, but many of those court doing so will come to an arrangement with the Taxation Office which includes paying massive penalty taxes and therefore never show up in any kind of criminal justice statistics. Similarly, many people who defraud the Commonwealth by claiming social security payments to which they are not entitled are dealt with by administrative arrangements and thus don’t appear in our crime figures.

34

wring:

I can surely see how the DOJ stats could be misapplied or incorrectly used to support a racial profiling policy. My main objection here was what I perceived to be a willingness to consider the DOJ stats worthless. Providing a flawed picture, yes. Worthless, no. I do not support a profiling policy mainly for civil rights reasons (i.e. a person shouldn’t be subjected to a “stop-n-search” without cause simply because they fit a certain profile).

I agree that it’s not the only factor however I believe that it IS the major one.

Biggirl:

I would agree with this if the statement was amended to say "… far greater than can be explained only by socio-economic factors… ". Socio-economic factors IMHO are still the major cause here. Profiling, imbalanced laws or unconcious racial bias are secondary factors which exacerbate differences which are primarily the result of socio-economic factors.

To me the stats are more an argument about the difference between the criminality of the poor and middle/upper classes. They don’t signal to me that minorities are inherently more dangerous, simply that they are in a socio-economic class which has more motivation to pursue, and more exposure to (in general) criminal activity. However, I will grant that this may be a more nuanced view then most people get when they read the stats. Isn’t it always the case though with statistics that they only become meaningful after doing some research and rationally analyzing the causes?

I believe wring pretty well covered this earlier. Drug abuse convictions are commonly obtained as a secondary conviction to a primary criminal activity. Some of it is ancillary to crime work being done relative to lower classed based crime (where minorities tend to be disproportionately represented), however, at the same time some of it is also (IMHO) because of unintended racial bias.

In other words, I still hold position #2 from my earlier post. DOJ stats would still show disproportionate minority crime rates (even though the differences would be less dramatic) if racial bias were less prevalent (and by extension if that were reflected in police procedure and local/state law).

Grim

35

If I didn’t make myself clear my position is:

Grim

36

Sorry, wring, you still haven’t made your case.

If the statistics on murder are unreliable because so many murders are undetected, then they do not show that blacks are overrepresented among murderers if most of the undetected murders were committed by whites.

If this is not the case, and roughly 50% or less of undetected murders are committed by whites, then blacks are still overrepresented among murderers - just like the stats show.

If you are claiming that we can never know what percentage of undetected murders are committed by whites, you are making an unfalsifiable statement - what I referred to earlier as a “finagle factor”, roughly defined as that number which, when added to an equation, makes the answer come out to what you wanted in the first place.

The trouble is, we can equally apply such a factor to Biggirl’s OP, and say that the figures that lead us to conclude that lots more white drug users go un-arrested than black drug users are unreliable, because we do not know how many people lie on drug surveys.

In other words, why can we conclude that the system is biased and racist based on data which are just as flawed as they are when we conclude that blacks are disproportionately involved in crime?

Especially if, for crimes where we have little or no reason to suspect huge numbers of biassing, unreported crimes, the numbers for black arrest and conviction are just as high as they are for crimes with a higher potential for racial abuse?

You can’t just wave your hands over the data and dismiss them because you assume they are racially biased.

They are the best figures available. If you can’t prove anything from them - well, then you can’t prove anything from them.

Including racial bias.

Regards,
Shodan

37

Shodan you misunderstand.

I do not contend that the data is biased. I contend that

  1. The data only shows those crimes which are reported, and solved.
  2. the data also shows some demographics about the people who’ve been convicted.
  3. to attempt to extrapolate from this set of data, any reasonable inferance about the racial characterisitics of the greater group ‘people who commit crimes’ is unreasonable, and shouldn’t be done.

I also contend that there are laws that by their wording, will automatically subject those living in certain socio-economic areas more seriously, and that since those areas have a higher percentage of minorities than whites, that such laws will necessarily impact minorities more than whites.

The OP suggests (and I agree) that using DOJ stats to prove /disprove the criminality of a race is subject to inevitable and unredeemable flaws.

And I think the case has been shown.

You wish to argue about specific numbers of murders, as shown by the DOJ site, that’s not my particular concern. You brought up the sub classification of murder, I merely pointed out that even in that classification, the data regarding perpetrators is incomplete. (by the way, you also should note, according to this some 17,000 agencies report their figures to the FBI, resulting in the data base we’ve been talking about, so not all of the data goes into the numbers, either)

The data that we can be certain of, is # of people incarcerated (they keep pretty damn good track of that), but because of the problems I’ve already outlined, attempting to derive conclusions about criminal behavior based on racial classification from the numbers of incarcerated is inherently and irretrievably flawed.

38

I remember reading an article in The Economist which said that a higher proportion of blacks and hispanics were arrested for drug offences because they tended to deal on the streets whereas whites, who generally have more money, tend to deal behind closed doors where it is harder to catch them. Could this go some way to explaining this racial disparity?

39

Behold, *Gomez! * Killer of threads!