What is this logical fallacy called?

[Reg Prescott]

2 hours ago, John Cuthber said:

But the (assumed) view of the police is that blacks are violent. and that's why they arrest so many of them.

That factor is inherent in the process of arrest.
So, it's a self fulfilling prophecy.
Looking at arrest rate as a proxy for criminality should be reasonable, but, if (as we suspect) the police are biassed then it's no longer valid.
If you use that measure of criminality to drive the arrest rate then there's a feedback loop.

Essentially they are arrested for being black, so they are arrested for being black.

Well, if the police were to advance an argument of the form you just described, i.e.

Premise 1: Blacks are arrested for being black
Conclusion: Blacks are arrested for being black

they would indeed be guilty of begging the question. But the OP has given us no information that such an argument has been advanced.

 

Or as another question-begging example:

Premise 1: Blacks engage in violence to a higher proportion than other races
Premise 2:
[...]
Conclusion: Blacks are inordinately violent compared to other races

we see the conclusion is simply a restatement of Premise 1. Nothing has been proven or derived or inferred: it was there to begin with. Any person making such an "inference" (i.e. no inference at all) can be charged with committing the fallacy of begging the question, regardless of whatever his or her personal opinion of blacks is. Personal bias is irrelevant.

 

But once again, no such argument has been advanced. So a charge of begging the question can, I think, be safely ruled out.

The crux of the issue lies with the (non question-begging) argument (BIV) that has actually been offered by the police:

Premise: Black arrests for violent crime are disproportionately high
Conclusion: Blacks are inordinately violent

Now, the argument as it stands is clearly not deductively valid -- the truth of the premise does not guarantee the truth of the conclusion. Thus, anyone making such an inference, and claiming deductive validity, would be mistaken. If there's any logical fallacy to speak of, it would simply be invalid deduction.

 

The argument could be rendered valid, though, by adding extra premises. E.g.

Premise 1: Black arrests for violent crime are disproportionately high
Premise 2: People who are arrested for violent crime are violent
Premise 3: Arrests are made indiscriminately
Conclusion: Blacks are inordinately violent

Deductive validity is now secure ... I think. Assuming the truth of premise 1 (i.e. the police are not fabricating statistics), anyone who finds the conclusion unpalatable would have to challenge the second or third premise.

 

Alternatively, rather than invoking deduction, the original argument (BIV) might be recast as an inference to the best explanation (IBE). In other words, it might be argued that of all the candidate explanations for the undisputed fact "Black arrests for violent crime are disproportionately high", the best explanation is that "Blacks are inordinately violent", thus we can infer -- depending on how IBE is construed -- to the truth, or the likely truth, or similar, of the proposition "Blacks are inordinately violent".

Those unhappy with the conclusion, assuming they grant IBE as a valid form of inference, would have to contend "Blacks are inordinately violent" is NOT the best explanation for "Black arrests for violent crime are disproportionately high".

Once again, though, as with inappropriate deduction, I don't think there's any standard name for this. It's simply an invalid inference.
 

Edited November 18, 2018 by Reg Prescott

[John Cuthber]

That's all very well, but it loses sight of the feedback.
It is certainly plausible that 

(1) The police, for some arbitrary reason- possibly sheer chance, arrest a disproportionately large number of black people.
(2) As a consequence of that, the arrest rates are higher for black people.

(3)The police interpret that as implying that black people commit more crime (which should be a valid implication).

(4)On that basis, they target black people for "suspicion" (Again, this should be a valid way to act; if you see some group as being more likely to be involved in committing crime, it makes sense for the police to target that group) and,

(5) as a consequence of spending more time looking at black people, they arrest disproportionately more black people. (and, yet again, that should be a sensible outcome)

(6) And that drives the arrest rate for blacks up still further.

 

Nobody, as far as I can tell, has made a logical error except in failing to check the statistical (and other) validity of the first step.

 

In reality, it's possible that part of the reason for the initial high arrest rate is racism, but it's not necessarily  the cause.

The real problem is that an arrest rate should be an indicator of criminality, but it may not be.

 

[Reg Prescott]

@ John (post directly above)

All of that may or may not be true, but the topic is logical fallacies, and since we lack access to the mental states of those involved, all we can do is appraise the arguments as presented to us.

Supposing the officers involved were to make explicit their thought processes, we could then examine the cogency of their reasoning. As things stand, we cannot. All you've said is hypothetical.

Of course we could take a stance "If the police are reasoning in such-and-such a way then ... whatever". Not sure if that's what the OP has in mind.

 

Edit P.S. -- What you've offered us in your (1) - (6) above is a causal account for the unusually high arrest rates. That's not what concerns us here (if I understand the OP). What we're interested in is relationships between statements, i.e. what can be inferred from what. Not what causes what.

Oh, and most of all, does the form of fallacious reasoning involved, if indeed there is one, have a name?

 

Edited November 18, 2018 by Reg Prescott

[Reg Prescott]

On 11/17/2018 at 7:55 AM, dstebbins said:

So what exactly would this sort of logical fallacy be called? Where you argue your actions are justified based on evidence that you yourself have a lot of influence, if not total control, over.

Reading through the OP again, and focusing on the last sentence (quoted above), it seems to me inappropriate in a case like this to be speaking of logical fallacies at all.

A logical fallacy would normally (I think) be a case where, due to some error of reasoning, an unjustified conclusion is yielded. The person arguing has not provided us with a good reason to believe her conclusion. The epistemic justification for the conclusion is insufficient.

But that's not the situation here as you've described it. The question is not "Why should we believe your conclusion?", which is a matter of epistemic justification, but rather "Why do/did you act as you do/did?". What's being demanded is explanatory justification for one's actions, not epistemic justification for one's conclusion.

 

To continue, the accused might plead in her defense, "The reason for my action is X". X is her alleged justification; her explanation (or perhaps excuse) for acting as she did.

But the accused, according to the OP, "has a lot of influence, if not total control, over X".

Then what do we say (if we're in a particularly suspicious frame of mind)? Manipulation? Machiavellian manoeuvring? Skulduggery?

Still doesn't seem to me a logical fallacy, i.e., a flaw in one's modes of inference.

 

Edited November 18, 2018 by Reg Prescott

[swansont]

On 11/16/2018 at 9:42 PM, dstebbins said:

Neither of those seem to fit 100% what I'm looking for.

Circular reasoning is close.

Arrest and prosecution statistics tell us group X is more prone to commit crimes, therefore we must not give them the benefit of the doubt when faced with the question of arresting or prosecuting a crime — which leads to group x being arrested/prosecuted more often — which bolsters the statistic

It's not necessarily solely that, since the original statistical deviation may be due to some bias (and in your example, likely is). But that particular justification is circular reasoning. That's going to be an issue when you have more complex scenarios — more than one issue is in play. It's not necessarily going to distill down to textbook examples of a single fallacy.

 

 

[iNow]

Tautology comes to mind, as well