Hate Crimes in Black and White

Which should the law treat more severely: Killing a guy because he cut you off in traffic or killing a guy because you don’t like his race?

Elsewhere on the web (link omitted because the source is the invitation-only blog of a personal friend), I read the following:

In the former case, you’re a danger to the person who wronged you. In the latter, you’re a danger to tens of millions of people, and that’s just in the US.

Hate crimes are different because the perp’s target list is vastly larger, with the built-in implication of recidivism.

There’s so much wrong with this I’m not sure where to begin. First of all, when a guy kills another guy for cutting him off in traffic, I’m inclined to think the likelihood of recidivism is pretty high. It’s not like nobody’s ever going to piss him off again. Second of all, I’d think that severity of punishment should be tied primarily to its effectiveness as a deterrent to others, not as a deterrent to recidivism. We can deal with recidivism partly by keeping an eye on past offenders, but when it comes to deterring unknown others, punishment is all we’ve got.

But I mention those issues only in passing on my way to what I think is the really interesting question, namely: Which is more harmful? Targeting a specific individual for death or targeting a randomly chosen representative of some race?

And while we’re at it: Which is more harmful? Targeting someone for being black, or for being white?

Some thoughts:

1. Generally speaking, people will pay more than a million times as much to avoid certain death as they will to avoid a one-in-a-million chance of death. By that measure, targeting a specific individual is more harmful than targeting a randomly chosen Albanian. And indeed, we allow this fact to guide public policy all the time. We are willing to spend, typically, maybe 5 or 10 million dollars for a guard rail that we expect to save a single randomly chosen life, but we’re willing to spend far more than that to save a particular miner trapped underground. I’ve argued (for example, in Chapter 15 of More Sex is Safer Sex that we ought to rethink those policies, but most people (unlike me) seem to be entirely comfortable with them. If you’re one of those people, then you ought to be willing to spend more to deter the death of one specific person than one who’s randomly chosen from a large population. In other words, you should view the “hate crime” motive as a mitigating factor, not an aggravating one.

2. To put this another way: When you target a specific victim, you do a lot of damage to one person. When you target a randomly chosen Albanian, you do a small bit of damage to each of many Albanians. It’s not immediately obvious in which case you’ve done more total damage. But all of the micro evidence points to the first case, and almost all existing public policy is formulated as if this micro evidence is definitive.

3. Suppose, however, that you believe for some reason that in this case, those arguments don’t apply. For example, you might believe that hate crimes against Albanians tend to unsettle Albanians at higher levels than are commensurate with the actual risks they face. In that case, you might be able to rescue the conclusion that hate crimes cause more harm than non-hate crimes. But you’ll also almost surely be forced to the conclusion that hate crimes against Chinese are worse than hate crimes against Albanians, simply because there are more Chinese than Albanians, each of them feeling outsized distress when crimes are committed against their group. As another application of the same principle, you’ll be forced to conclude that hate crimes against whites should be treated more severely than hate crimes against blacks.

4. The question on the table is: How does the size of the potential victim pool affect the severity of the crime? To focus attention on this question, we can consider Crime A with a victim pool of one (the guy who cuts you off in traffic), Crime B with a much larger victim pool (a randomly chosen black) and Crime C with an even larger victim pool (a randomly chosen white). (Obviously I’m envisioning a locality where whites outnumber blacks; if you live in a locality where the opposite is true, you can reverse Crimes B and C.) If your argument (like that of the blogger I’m responding to) is that “larger victim pools make things worse”, then you’ve got to conclude that Crime C is the worst. If your argument (like the one I’ve sketched, but not endorsed, in point 1) above) is that “smaller victim pools make things worse”, then you’ve got to conclude that Crime A is the worst. But you’d have to go through a lot of contortions to conclude that Crime B is the worst. I am sure a sufficiently clever debater could manage those contortions, but I’m skeptical that he could manage them in a terribly compelling way.

5. Conclusion: My blogger friend, if his argument is both serious and honest, should endorse the conclusion that hate crimes against whites are worse than hate crimes against blacks. Either that, or he should rethink his argument.

6. We really ought to have formulated this question a lot more carefully in the first place. “Which crime should be treated more severely?” is not a well-defined question. Does it mean, for example “Which crime should be prosecuted most vigorously?”, or “Which crime should be punished most harshly?”. Those are not the same question, and I (like my blogger friend) have been careless about conflating them. But I suspect that no matter how careful we are, it’s going to be very hard to avoid some conclusion along the lines of point 5).

7. (This was in my original draft and somehow got inadvertently cut before I posted; I’m restoring it now): You could also argue that the death of an aggressively bad driver is marginally less tragic than the death of a randomly chosen black. If you expand on this argument in the obvious way, it’s an argument for treating hate crimes more severely, but a slight variant (namely, that individual targeting can also deter desirable behaviors) goes the other way. In any event, I take this to be off topic, since the post I’m responding to is specifically about the size of potential victim pools as an argument for or against hate crime legislation.

Addendum: If I ever kill you in a rage, it’s not going to be for cutting me off in traffic; it’s going to be for ignoring the logical consequences of your own arguments. Just sayin’.

Share/Save

109 Responses to “Hate Crimes in Black and White”


  1. 1 1 Roger

    I see a couple of arguments for hate crime laws.

    If the KKK were engaged in an organized policy of burning down black churches, then maybe some sort of federal hate law could destroy the KKK and solve the problem.

    If bullying be portrayed as some sort of hate epidemic, then the LGBT lobby can use that as an excuse to get their propaganda into the schools.

  2. 2 2 Tristan

    Not to be pedantic, but if we’re talking about why the average person regards hate crimes as especially objectionable I think it’s worth pointing out that their arguments are not going to fall along the traditional cost-benefit framework we’re used to. While it might be interesting and internally consistent to make up defenses of the concept of “hate crime” along the lines of smaller/larger groups etc etc I think it misses the obvious argument most people have in mind.

    That argument is that punishing certain crimes as being hate crimes sends a message about racial equality in general. A bunch of whites getting together to kill the black gentleman in their community because they don’t take kindly to him dating a white girl is punishable in a way that their less overt racism is clearly not. We can’t put them in jail for regarding him as their genetic inferior, but we can at least approximate that by giving him more jail time when he does commit a crime that explicitly identifies him as a racist. So you can fight racism by proxy by making an example out of those racists who happen to get on the wrong side of the law. Not saying whether I agree with this argument for hate crimes framed this way, but I think that’s obviously what most people have in mind…? Also Otto, thanks for lowering the bar for my comment!

  3. 3 3 RichardR

    “When you target a specific victim, you do a lot of damage to one person. When you target a randomly chosen Albanian, you do a small bit of damage to each of many Albanians. It’s not immediately obvious in which case you’ve done more total damage.” I disagree because when you randomly target one Albanian you do a lot of damage to the one Albanian you have targeted AND a little bit of damage to each single Albanian (because of the fear caused), total damage is therefore a lot greater.

    “you’ll also almost surely be forced to the conclusion that hate crimes against Chinese are worse than hate crimes against Albanians, simply because there are more Chinese than Albanians”. No, the amount of fear caused to individuals within the targeted group decreases as the size of group increases but increases because the amount of fear is multiplied by more persons. There is no a priori reason to assume that one effect is bigger than the other.

  4. 4 4 Navin Kumar

    @RichardR

    “I disagree because when you randomly target one Albanian you do a lot of damage to the one Albanian you have targeted AND a little bit of damage to each single Albanian (because of the fear caused), total damage is therefore a lot greater.”

    When you kill a guy for cutting you off, you’re imposing fear on other drivers. This is not a trivial matter, btw. I live in Delhi, and everyone knows that you should avoid confrontation with certain types of people on the road, because they have guns and will shoot you.

  5. 5 5 Andy

    On 3. I think you could reasonably argue that if Albanians are likely to feel more outsized distress (for instance because they have been forced into slavery in the past, and note that this could still be the case even if you think they have no right to feel that way after all this time) than the Chinese, then a hate crime against Albaninas is worse than against Chinese despite there being more Chinese people. Now replace Albanians with blacks and Chinese with whites and your last argument in 3. is no longer valid.

  6. 6 6 Harold

    Point 2. RichardR #4 got there before me with that point – the large amount of damage to the individual victim is the same in both cases, it is only the collateral damage that we need to to consider for any given murder. As Navin Kumar says #5 it may not be obvious in which case the collateral damage is greater.

    Raising the severity for particular types of crimes should have some reason. One reason may be that this particular class of crime is more common than similar types. So if blacks getting murdered by whites because of their color is quite common, but whites getting murdered by blacks (because of their color) is almost unheard of, then it may be a good idea to raise the penalties for the former. This is probably too specific, so if we raise the stakes for “hate crime” in general, we will get our desired effect for whites killing blacks, and will suffer little harm for “overpunishing” the blacks because there are not many of them. Thus there are very few attacks by gays on hetersexuals because of their sexuality. If we have a hat crime of attacks based on sexuality, we do not over-punish many gays

    So in Delhi, if murdering other drivers is more common than other similar murders, then extra penalties for road rage murders may be appropriate.

    So we don’t generally create special classes of crime that we would find abhorent, but which do not occur in practice. We create hate crime classes because they are relatively common.

    The question asked “Which is more harmful? Targeting a specific individual for death or targeting a randomly chosen representative of some race?” is therefore not really the right question.

    Putting it another way, murder usually requires extreme motivation. If this motivation is brought into the almost mundane for a lot of people, such as someone merely being a particular color, sexuality or poor driver, then some measure may need to be taken to deter these crimes over and above the normal deterent for murder.

  7. 7 7 Alan Gunn

    It’s not as if killing people either because of their race or because they cut you off in traffic is legal. The main reason for hate-crime laws is to give members of Congress the opportunity to look good to members of various racial or other groups at little cost, by seeming to share their concerns. Few members of Congress seem to see that need when it comes to getting votes from bad drivers.

    About the only practical effect these laws have is in a handful of well publicized cases in which an unpopular defendant gets off in state court and is then retried in Federal court for the same offense. This ought to violate the double-jeopardy clause, but the “dual sovereignty” notion, invented by the Supreme Court many years ago to allow prosecution of murderers who had gotten off because of racist white southern juries, says it’s OK. A nice example of “hard cases make bad law.”

  8. 8 8 suckmydictum

    A more extreme example would be Ted Bundy, who, I believe, only killed women, which is the largest victim pool anyone’s talking about. And he killed them because they were women. Hate crime or not? Is half the population too big of a victim group to count as a hate crime?

    The more I think about his, this more i think hate crime law really is unalloyed thought-crime law. If a group of open neo-nazis spend their weekends stewing about blacks and Jews and eventually go out and kill one, if this is classified as a hate crime, what the hate crime law seems to be punishing is either the previous stewing about blacks and Jews or the affiliation with nazi ideology, which are not punishable offenses by themselves. How this is substantively different from Ted Bundy stewing about killing women in his alone hours is not quite clear to me.

  9. 9 9 Xorph

    The collateral damage difference between the “all drivers” pool and the “all blacks/whites” pool is that even though you may do a mistake once in a while, you *can* avoid cutting people off, and if you do you’ll be fine (and the whole drivers pool will even benefit from your avoiding this).

    @Steven Landsburg
    This kind of posts is really my favourite of yours. Reading and following a logical argument and being suddenly trapped and forced to go against one’s own prejuges is absolutely exhilarating and enlightening.

  10. 10 10 suckmydictum

    @8

    “The main reason for hate-crime laws is to give members of Congress the opportunity to look good to members of various racial or other groups at little cost, by seeming to share their concerns.”

    If that’s true why wouldn’t crimes against women count as a hate crime? That’s a huge part of the electorate.

  11. 11 11 Keith

    On 3: Assuming you are going to kill only one person, there being more Chinese than Albanians means an individual Chinese person will suffer *less* distress than an Albanian. (This is because the probability of death for a Chinese person will be smaller than that for an Albanian.) And since the relationship between “probability of death” and “distress” is superlinear, this should mean total distress is smaller for the Chinese than for the Albanians.

  12. 12 12 Xorph

    @Keith
    This would be true only if distress was rational ; as a very basic instinct i rather think it is not, and also it is an individual feeling, meaning that in a paranoid way, it totally leaves group probabilities out. It’s an emotion : you have distress or you don’t. It doesn’t mitigate in probability like the killing does.

  13. 13 13 Ken B

    SL:” Second of all, I’d think that severity of punishment should be tied primarily to its effectiveness as a deterrent to others, not as a deterrent to recidivism. ”

    Hmmm. Potential recidivists are part of the sample space of future offenders, and the one with a proven propensity to commit the infraction in question. Seems entirely reasonable to think deterring recidivists should be weighted proportionately more than deterring some random person.

    The argument I am making here seems oddly like the one Steve is making in the post.

    Steve I believe advocates at least one policy which does have a disproportionate effect on potential recidivists: the death penalty.

  14. 14 14 Jonathan Campbell

    “When you target a specific victim, you do a lot of damage to one person. When you target a randomly chosen Albanian, you do a small bit of damage to each of many Albanians.”

    You are arbitrarily drawing a line at the wrong part of the continuum between the crime’s conception and execution. You could just as arbitrarily eithar
    a) change the last clause to “you do a lot of damage to a single Albanian”
    b) change the second part of the 1st sentence to “you do a small bit of damage to all people who you have feuds with” [or replace "feuds" with whatever it was that caused the problem with that particular person]

  15. 15 15 TjD

    It seems that you analyze on a micro level ( one man killing another man ) and then try to reason on the macro level ( laws.. )

    Maybe you should compare 15000 road rage victims vs 15000 Albanians getting killed each year because they are Albanian?

    To me, it seems obvious that people working together result in a more productive country, John Forbes Nash would say it better.. One man killing another man because of road rage is bad, one man killing another man because of race will incite rage and revenge, hence it is worse.

  16. 16 16 Rowan

    I’ll take a stab at a broader argument:

    Systematic oppression causes a huge range of ills in the oppressed group — for instance, poorer physical health, less education, and restricted access to economic advancement (or even security). The costs of those ills are not born by the oppressed group(s) alone, but by society as a whole.

    Hate crimes are the most acute expression of systematic oppression, and the most extreme way of perpetuating it. As such, they should be punished more harshly.

    (I also agree with the commenters above who said that you’re adding up the “damages” incorrectly in Point 2.)

  17. 17 17 James Kahn

    If people in the threatened group are aggrieved by, say, doubling the perceived probability of getting killed, then someone threatening to kill a random black will do more damage than one threatening to kill a random white, because of the relative safety in numbers of whites. Add to that, the perceived risk may be related to the number of potential attackers. If Albanians have a proclivity for attacking Chinese, that’s a lot ess threatening to the Chinese than vice versa.

    Having said all that, basing laws on vague collective distress scenarios like these is wrong.

  18. 18 18 Jonathan Kariv

    First thought is that the size of the perpetrator pool should probably matter to. If we have alot of people that hate group A and only a few that hate group B then you deter more murders by punishing group A more severely.

    Second thought is that sure the guy who kills someone for cutting him off in traffic is likely to kill again but I think your blogger friend is probably right about recidivism in that the expected number of future kills is probably higher for the person killing over race.

    Third thought. What about someone who wants to kill someone and doesn’t care who. This is surely further on the bigger population scale but it seems alot like someone who wants to kill a specified but undisclosed (to anyone but the potential murderer) person.

  19. 19 19 JonathanKariv

    *edit “punishing people who murder members of group A more severly”

  20. 20 20 Dan

    I don’t understand the argument. The certainty of death is supposed to be from the perspective of the the victim. Someone may very well pay more than a million times as much to avoid certain death as they will to avoid a one-in-a-million chance of death. But this assumes that the victim knows that the random (non-hate) crime will happen with certainty. Only the perpetrator knows this. If the victim knew they’d just call the cops!

    It’s the same thing from the policy maker’s perspective. They don’t know with certainty, otherwise they’s arrest the perpetrator.

    The rescue of a particular miner is therefore not a good example in this case.

    Both the victim and the policy makers face two uncertain events: probability of a random crime and the probability of a hate crime occurring.

  21. 21 21 Dan

    Also the size of the victim pool would have to be weighed against the probability of a hate crime from the other groups.

    Take an extreme example. Everyone’s white. The white victim pool is therefore very large. But in this case there would be no hate crime against whites since everyone’s white.

  22. 22 22 Alan Gunn

    @10

    Yes. That explains the two Violence against Women acts (one held unconstitutional, the other on the way, with the supposed infirmities fixed). Murdering women is a state crime everywhere, and under these acts a Federal crime as well. Murdering men is just a state crime, usually.

  23. 23 23 Harold

    Putting it another way, the quote is in error:
    “In the former case, you’re a danger to the person who wronged you. In the latter, you’re a danger to tens of millions of people” In the forme case you are a danger to anyone who cuts up while driving. That could be millions of people.

  24. 24 24 Ken B

    @22, 10
    Yeah, I gotta wonder where suck has been these past 30 years. Just compare how the law and politicians treats female rape compared to male rape. Politics especially, which is the topic under discussion after all. You can make bend over for the soap in prison jokes and still get elected is my bet.

  25. 25 25 Ken B

    My main reason for opposing these laws is that we should at least pretend to try treat people as individuals, not merely as instatiations of a type. It’s pretty sad when we cannot even pretend.

  26. 26 26 xan

    It seems to me that punishing hate crimes more severely is an attempt to shift out of an equilibrium where such crimes are subsidized by social approval and into an equilibrium where all people are targeted for murder equally.

    It is not a question of whether you do more harm targeting a specific individual versus a group. It is a question of the *aggregate* risk faced by each person. If everyone targets the same group, there is a disparity between their risk and everyone else’s. Perhaps, from behind the veil of ignorance, the average person would prefer to live in a world where the risk is distributed equally.

    In which case, spread those murderous tendencies around!

    (This is not *actually* an endorsement of current laws, just one small facet of a much bigger argument).

  27. 27 27 Ken B

    There is another aspect of these kinds of laws that is troubling, and does crop up from time to time, and that is a defence based on religious or historical enmity. If our road rager tried to argue he should be let off easy everyone would laugh. But in other cases you can hear this kind of defence. I have seen it in rape cases. And there really is some logic to it, “I was raised this way by my culture/group/religion in our attitudes towrds them”, if you accept in the first place that group membership should be a mitigating or aggravating factor.

    I find this really noxious myself.

  28. 28 28 suckmydictum

    @22

    The violence against women acts, as far as I know, do not fall under the umbrella of hate crime law. We could split some hairs and say the VAWA are just mutations of pre-existimg hate crime law supposedly meant to protect vulnerable minorities, but they seem different to me. I’m just not sure why they are different or why they should be different.

  29. 29 29 Ken B

    28: “they seem different to me. I’m just not sure why they are different”

    Perhaps because they refute your claim?

  30. 30 30 Bob Murphy

    “If I ever kill you in a rage, it’s not going to be for cutting me off in traffic; it’s going to be for ignoring the logical consequences of your own arguments.”

    If I am ever in a life-or-death struggle with Steve, I plan on stunning him by saying mathematics is an axiomatic system.

  31. 31 31 Ben

    I think Xorph’s response is right. There is a categorical difference between targeting blacks and targeting people who cut you off — it is socially acceptable for the latter to make you angry (because cutting you off is wrong, and people can avoid cutting you off), but not the former. As a society, we view overreaction (crime of passion, murder 2) as a less serious crime than cold premeditation (murder 1). Unless you’re prepared to argue that crimes of passion are just as serious as premeditated crimes, it seems perfectly reasonable to draw a distinction in the particular example you give in your first paragraph (though perhaps not for the reasons your friend gives).

    Now, if the driver went home and carefully planned his revenge against the other driver who cut him off, then I don’t see a reasonable distinction.

  32. 32 32 khodge

    I have a problem with #4, size of the victim pool. There was comment that Steve referred to in one of his rape victim responses suggesting the harm done to the observers; I think his scenarios consistently ignore the effect on the non-victim/non-aggressor group: Yes, the “crimes against whites” has a larger potential victim pool but all humanity participates in the general degradation of value when the actual crime occurs without any consequence. If it is less harmful to degrade the value of the Albanian because of the size of the pool, where does the “material” dividing line fall? Albanians/Italians? Italians/Mexicans? Mexicans/Chinese?

  33. 33 33 suckmydictum

    @29

    I’m not sure what claim you’re referring to. The only claim I think I made in this thread was that hate crime law is akin to thought-crime law (i.e. people are being punished for prejudices which existed before whatever crime they commit). I actually think you’re right to connect this to reasons of historical enmity.

    My confusion is as follows: if you ask someone who advocates for hate crime laws which groups ought to be protected under a hate crime law, women are not at the top of that list (usually you hear about racial or religious minorities or gays). If these people insist on having hate crime laws, why isn’t violence against women (on account of their being women) a hate crime? And if Alan Gunn is right that these laws are a way to curry political favor, why wouldn’t the VAWAs have been classified as hate crime laws?

    I dislike the laws for some of the same reasons you do, but I also dislike them because I can’t see why they can’t be extended to protect just about any social group a demagogue likes.

  34. 34 34 Ken B

    @29: Comment 10.
    (The only comment even remotely relevant to the discussion btw.)

  35. 35 35 Ken B

    @33
    Well, the magic isn’t in the phrase “hate crime”, it’s in passing laws to favor one group and hence to curry favor either with that group or (more often) with les biens pensants who think said group needs or deserves favor.

    In general it’s harder to suck up to large groups on issues like this because the implicit assumption is that the micreant will be from some *other* group. The bigger the group the less true that is.

  36. 36 36 suckmydictum

    @34

    I concede I rambled away from the original discussion. Apologies.

    @35

    I’m not sure it’s done quite as cynically as you make it sound. people who draft and vote for laws like this probably do think they’re doing something important and like feeling sympathetic to a group they believe need protection. It’s just unfortunate for them that this is vulnerable to the sort of argument SL made.

  37. 37 37 Ken

    For those claiming that the size of the victim pool should be discounted because in a larger victim pool you have a reduced chance of being victimized, your claims are wrong. Consider the fact that dozens of black women are raped by white men every year, but tens of thousands of white women are raped by black men every year (the ratio is on the order of 2000 white woman raped by a black man for every black woman raped by a white man). Thus even though there are roughly six times as many white women as black women, the probability of white woman being raped by a black man is still orders of magnitude greater than the probability of a black woman being raped by a white man, negating your assumption that a larger victim pool reduces the probability of a crime being committed against it.

    Additionally, even if your assumption was correct (which it is not), the total damage done by a single crime to a particular population is independent of population size. By your own assumption, population X has x people in it and population Y has y people in it. Say a crime costs C, then the individual fear anyone in population X will feel is C/x and the individual fear anyone in population Y will feel is C/y. But when totaling the fear in population X, you see the total fear is C. Similarly, the total fear in population Y is C. By your own assumptions, the punishment for any particular crime should be the same regardless of why that crime was committed for any malicious purpose. This means that a murder committed due to someone cutting another off in traffic is just as bad as the murder committed by someone due to racial hatred.

  38. 38 38 Ken B

    @36: No apology needed. I’m just glad you’re not arguing the box car moves up.

  39. 39 39 Rory

    I think that there are interesting ways to object to the size of population arguments, but the most compelling argument against your stance is the idea that increased fear is independent of group.

    If the crime in question is perpetuated by a majority against a minority on the grounds of the minorities non-inclusion in that majority, the implied threat is much greater. Now, arguably, the number of people so racist as to commit hate-motivated murder is small enough that we no longer need disproportionately punish hate crimes. But if we posit a circumstance where racism, and even violent racism, is a common cultural phenomenon that the government wishes to greatly reduce it seems much more reasonable to impose additional penalties for hate-crime.

    Also, I think Andy makes a great point that differing groups might reasonably have different levels of fear based on historical factors (as well as majority/minority status). Whites in this country are historically dominant, and over-represented in the legal and political structures of the country, which certainly seems to suggest less rational basis for fear of systematic persecution.

  40. 40 40 Todd

    Virtually all attempts at stamping out racism start by institutionalizing the premise that different races should be treated differently. This necessarily implies that racial groups are meaningfully different from one another—the very premise underlying all forms of racism. When we accept that there is a meaningful difference between two things, it is only natural (if not entirely inevitable) that a person will have a preference for one over the other. Hate crime legislation, affirmative action, the NAACP—whatever the intention behind them, all of these actively promote racism’s most basic and necessary false premise.

    Ken B has it right. Let’s treat people as people.

  41. 41 41 Daniel

    @Todd.

    When do you think this change over should have occurred? For example, should we suddenly have treated everyone the same directly following the end of slavery? The fact was that we had to formulate policy regarding race following slavery to prevent the already existing concepts of race from prohibiting former slaves from being maltreated by the existing system. It seems to me that race was a concept which profoundly influenced the direction and nature of our country, and even if we pretend in law that it doesn’t exist doesn’t mean that it will suddenly disappear from influencing how people act or create policy. For example federal law is formulated such that you can’t craft policies that indirectly but obviously effect minority groups. How would you suggest we prevent, for example, voter suppression laws without a concept of a protected class. I do think that we should teach our children to treat “people as people” but I don’t think we’ve gotten to the point where law can completely ignore race as a concept.

    More on topic, I do think like other’s that the crime does proportional effect on all groups regardless of the size of the group, so the increased penalty must be related to some other factor. Harold, I think had the best explanation for another reason other than direct harm why policies such as this may be desirable. I’d just add that we generally craft policies that may have a meaningful impact. So people may believe rightly or wrongly that they can have a greater impact on directly targeting individuals than on seemingly random acts of rage.

  42. 42 42 Daniel

    And Todd,

    All the examples you’ve given as examples explicitly do not rely on treating different races differently.

  43. 43 43 Steve Landsburg

    Ken (#37):

    By your own assumption, population X has x people in it and population Y has y people in it. Say a crime costs C, then the individual fear anyone in population X will feel is C/x and the individual fear anyone in population Y will feel is C/y. But when totaling the fear in population X, you see the total fear is C. Similarly, the total fear in population Y is C.

    In other words, you didn’t read the post.

  44. 44 44 Neil

    Suppose a hate group threatens harm to a given number of an identifiable group. When the group is small each member feels more threatened because the probability of being the victim is greater. However, a small group may have too little political power to get the crime classified as warranting special enforcement or penalty. On the other hand, a threat to a large group may cause insufficient concern to individual members to overcome free riding and cause them to organize and incur the cost of exercising their greater political power. This means that we should expect that an intermediate size of group, if any, should be most likely to succeed in getting “hate crime” legislated.

  45. 45 45 Wildkits

    Assuming there was criminal activity, would the IRS scandal be considered a Hate Crime?

  46. 46 46 Daniel

    @sl,

    Could ken’s point restated make sense. A completely rational individual should feel fear proportional to their probability of being selected, so the harm done to both populations, assuming rationality, is independent of pop size?

  47. 47 47 Ken B

    @Danile 45:
    That is precisely what Ken is assuming, that irrational fears are rationally calculated.

  48. 48 48 Ken B

    @37:
    Then the worst punishment should be reserved for crimes against non-existent groups shouldn’t it?

  49. 49 49 Steve Landsburg

    Daniel:

    A completely rational individual should feel fear proportional to their probability of being selected

    You’re looking to dictate the preferences of a completely rational individual?

  50. 50 50 Daniel

    @Steve,

    You’re right, I guess I was more thinking about the expected dis-utility, is that a word?, of the direct effects of the murder (that they won’t be living afterward). There’s obviously lot’s of complications to this, and I can see your point that they could have rational reasons besides expected dis-utility of the direct effects of the murder.

    Or is it wrong to say that their expected dis-utility from the direct effects of the murder would be defined in this way?

  51. 51 51 Daniel

    The Assumption is of course that on average and by the same distribution, that the individuals in the different groups value their own lives equally.

  52. 52 52 Ken B

    @51: If that is a reply to my 47, no. Ken needs to assume much more than that. He has to assume people seek out and use the data as well, and do it rationally.

  53. 53 53 Steve Landsburg

    Ken B:

    Ken needs to assume much more than that. He has to assume people seek out and use the data as well, and do it rationally.<

    No, he needs to assume far more than that, and what he needs to assume is in conflict with virtually all micro data. He has to assume that people maximize expected value, which neither theory nor data give us any reason to expect. He also has to completely ignore the original post, which is all about the non-linearity of WTP to avoid a given probability of death.

  54. 54 54 Daniel

    @52,

    Sorry I was actually qualifying my direct prior statement. Should have been more clear. I can see where in practice the harms are not proportional to their expected dis-utility, I was just seeing if we were on the same page about the direct expected harm (leaving fear out of it) of losing your life?

    I’m also not quite clear with the premise this line from Steve’s post.

    “Targeting a specific individual for death or targeting a randomly chosen representative of some race?”

    Maybe it’s just that it doesn’t apply very well to the example given, but I’m straining to see the distinction of “a specific individual” versus “a randomly chosen representative of some race”. In what way is the person that’s cutting them off in traffic not a randomly chosen individual among those that may cut the murderer off? They’re only a specific individual after they’ve cut the murderer off right?

  55. 55 55 Ken B

    @Steve 53:

    Yeah I realized just how much more he needed when I was at lunch.

    However in partial mitigation I could plead that a rational “use” of the data would require linear response to risk. I’m pretty sure a Dutch book is available otherwise.

  56. 56 56 Ken

    Steve,

    In other words, you didn’t read the post.

    I see. It couldn’t possibly be that I read your post and that you were wrong, right? But snark is easier that a substantive response.

  57. 57 57 Ken B

    Ken: “But snark is easier that a substantive response.”
    Good snark is substantive. Since you were unimpressed with Steve’s, let me give it a try!

    “In other words you didn’t understand the post.”

    Better?

    Ken, much of the post is *about* the inapplicability of your sum of fractions. Claiming then that the summ refutes Steve “even if” he was right suggests you missed the point.

  58. 58 58 Steve Landsburg

    Ken:

    Of course I could be wrong. Where do you think the error is?

  59. 59 59 Will A

    @Ken #37:

    Consider the fact that dozens of black women are raped by white men every year, but tens of thousands of white women are raped by black men every year (the ratio is on the order of 2000 white woman raped by a black man for every black woman raped by a white man)

    It is actually the opposite with a caveat. Tens of thousands of black women are raped by white men, but this is of no interest to local news stations.

    Dozens of white women are raped by black men, and this is very much of interest to local news stations.

  60. 60 60 Daniel

    @56,

    So snark + substantive arguments are okay right? I can’t count the number of times you’ve insulted someone’s intelligence before responding substantively to their arguments. I just want to make sure I play by the rules of Ken in future arguments.

  61. 61 61 Ken B

    Ken and Will A are disputing numbers.

    @Will A:
    Got facts to support that?
    @Ken:
    Got facts to support that?

    The only American figures I have ever seen are U.S. Department of Justice’s National Crime Victimization Study (NCVS) 2001-2003.They are a lot closer to Ken’s hyperbole than to Will A’s hyper-hyperbole.

  62. 62 62 Daniel

    @ 61,

    Can you send the direct link. I googled it and I got several links from different years but not the years you were referring to. Can you also provide the table from which you culled your assertion that “They are a lot closer to Ken’s hyperbole than to Will A’s hyper-hyperbole.”

  63. 63 63 Capt. J Parker

    Don Boudreaux wrote what for me was a thought provoking post on hate crime laws. His argument is that what hate crime laws really do is create thought-crimes. The additional punishment administered to one who commits a hate crime is in effect punishment for having a belief that does not conform to some societal(or governmental)norm. If you believe that freedom of thought (like speech) is a fundamental human right and this freedom must not be abridged despite how repugnant the thoughts (or speech) may be then you should oppose attempts to regulate it via thought crime laws on the basis that would violate fundamental human rights.

    Don’t let me garble Boudreaux’s argument. Here’s the link.

    http://cafehayek.com/2012/03/as-far-as-thought-goes-scienter-should-suffice.html

    His blog also pointed to this WaPo Op Ed by Richard Cohen saying the same thing – hate crime laws are really thought crime laws.

    http://www.washingtonpost.com/wp-dyn/content/article/2009/08/03/AR2009080302222.html

  64. 64 64 Ben

    “”"
    Which is more harmful? Targeting a specific individual for death or targeting a randomly chosen representative of some race?

    And while we’re at it: Which is more harmful? Targeting someone for being black, or for being white?
    “”"

    They are both equally harmful since in each case one person dies.

    Likewise giving out school places by lottery is equally unfair as giving them to, say only people born on a Thursday. But it sure SEEMS different.

    If you want to be fairer, you have to give school places to people who either deserve them more, or can get more benefit from them.

    Likewise, to make murder fairer (assuming that there are the same number of murders), it is less harmful to murder someone who deserves it more (whatever that means) or derives less benefit from living (whatever THAT means.

    Don’t murder. It’s bad.

  65. 65 65 Todd

    @Daniel 41, 42

    “All the examples you’ve given as examples explicitly do not rely on treating different races differently”

    Hate crime legislation punishes identical crimes differently based on the race of the victim (or sexual orientation or some other group distinction), affirmative action enforces hiring practices based on race and gender, and the National Association for the Advancement of Colored People exists for the sole purpose of treating one race preferentially relative to all others. If these are not cases of treating races differently, then I’m really not sure what would be.

    You seem to imply that legislation was necessary to protect former slaves; I contend that legislation has only served to institutionalize and exacerbate racism, then and now. Changes in attitude can’t be legislated, and the free market does an excellent job of punishing discrimination in spite of government.

    Righting past wrongs is no simple task, but I’m pretty sure we should not start by granting the premise upon which those past wrongs were justified.

  66. 66 66 iceman

    Some call it “murder vs. super-murder”. I agree in practical terms it simply involves too much mind-reading (since people tend not to overtly explain their motives except on Scooby Doo).
    I’ve always assumed the theory (to the extent it’s not just good politics) had to be that members of a targeted group experience disutility disproportionate to the risk; however, “randomness” seems key here so I agree with those who think the road rage example is a bit off since it involves behavior not just inherent attributes (e.g. no need to experience fear if you’re not out there cutting people off). I guess one way to get to point B is to argue once the group size gets large enough people get back to the ‘norm’ of discounting low-probability events at or below the actual risk. But I also agree then that much of the important context about relative group sizes is local (unless nationwide copy-catting is of concern), which does not seem to factor into the policy discussion at all.

  67. 67 67 Daniel

    @65,

    “Virtually all attempts at stamping out racism start by institutionalizing the premise that different races should be treated differently. This necessarily implies that racial groups are meaningfully different from one another—the very premise underlying all forms of racism. When we accept that there is a meaningful difference between two things, it is only natural (if not entirely inevitable) that a person will have a preference for one over the other. Hate crime legislation, affirmative action, the NAACP—whatever the intention behind them, all of these actively promote racism’s most basic and necessary false premise.”

    You implied in this statement not that these laws include race in their language, but that the laws imply some treatment of races differently.

    The hate crime legislation prosecutes people that commit a crime based on someone’s social group. This can be a black person who hates other black people because they’re black, a white person who hates other black people because they’re black, a black person who hates other people because they’re black, etc.. It could be because of religion, sexual orientation, etc. How does this treat one group differently than another? The rule applies equally to all groups. If you infer that because race gets heightened scrutiny and is defined as a way of grouping and thus legislating against discrimination, than yes, the law involves a concept of race.

    The same concept can be applied to affirmative action. The law simply states that a federal employer can not discriminate based on race, gender, etc. There are various measures that have been created to see if various federally funded sources have been following through with this law, but no where is one race or group treated differently than any other.

    Finally the NAACP is an organization and not a law, so I have no idea what you’re talking about here.

    Your premise that no legislation was required to protect former slaves from discrimination (a lot of times from their own State governments) immediately following the end of slavery is frankly ridiculous. I can list the many reasons why, if you’d like.

  68. 68 68 Harold

    \61. Such a breakdown seems difficult to find. The bureau of justice stats include the following.
    http://www.bjs.gov/content/pub/pdf/fvsv9410.pdf

    For 199-2004, the rate of rape and sexual assaults was 3.1 per 1000 for white women and 4.1 per 1000 for black women – a slightly higher risk for black women. The proportion of perpetrators was 60% white and 26% black. Given blacks formed about 11% of population the proportion of perpetrators is about 3 times higher for blacks.

    If I have figured this correctly, then white women are more likely to be assaulted by a black man than vice-versa because even if all the black victims were raped by black men, there would be more perpetrators than victims, so some of the black men must have attacked white women. The white atackers could have attacked only white women. I may be wrong about this.

    This is purely an effect of black men being highly over-represented in the attacker pool and black women being only slightly over-represented in the victim pool.

    The good news is that the numbers of assaults has dropped significantly.

    There is a case for hate crime law. For me it relies on their actually being more of that type of crime than would occur without hate. It is not about thought crime because if you only hate, you do not committ a crime. You only committ a crime when you act on that hate.

  69. 69 69 Harold

    @40 “This necessarily implies that racial groups are meaningfully different from one another—the very premise underlying all forms of racism.” There is a fallacy here – meaningful can have different meanings.

    If someone appears different, but is otherwise the same, is that meaningful? It has some meaning, because I can identify people with, say, blue eyes, or people shorter than 5’6″, or people that have skin darker than a certain shade. However, in the context the word is used in relation to race, these differences are not meaningful. By having a race law, we are saying that people have an identifiable difference, but not necessarily a meaningful one. In fact, it is the opposite. We are affirming that some identifiable difference is not meaningful.

  70. 70 70 Dan

    @68
    There is not enough information there to draw the conclusion you have.

    But, you are right that if you are interested in racial motivations for crime, at a minimum you should de-mean the average crime rate to capture socio-economic and other reasons for crime. e.g. look at diff-in-diff: (white on white – white on black) – (black on black – black on white)

    In other words, you want something like: conditional on the rape offender being black, what is the probability that the victim is white compared to the probability that the victim is black.

  71. 71 71 Jim W K

    Steve Landsburg >>. To put this another way: When you target a specific victim, you do a lot of damage to one person. When you target a randomly chosen Albanian, you do a small bit of damage to each of many Albanians. It’s not immediately obvious in which case you’ve done more total damage. But all of the micro evidence points to the first case, >>

    It seems to me that it is immediately obvious in which case you’ve done more total damage; and further it seems clear that the micro evidence points to the second case, not the first.

    If I call the killing of one person n damage, then the damage to the victim = n whether it is one person or an Albanian I’ve killed. But if I’ve done a small bit of damage to each of many Albanians by killing an Albanian then my total damage in the first case is n, and my total damage in the second case is n + a small bit of damage to each of many Albanians. It seems clear to me that there is more total damage to the second case, not the first as you suggest.

  72. 72 72 Ken B

    @71:
    Then the damage is greater the larger the victim group. A killer who cares not for your demographics but would kill all alike should, on your calculation, do more damage still shouldn’t he?

  73. 73 73 Todd

    @Daniel, 67

    “Your premise that no legislation was required to protect former slaves from discrimination (a lot of times from their own State governments) immediately following the end of slavery is frankly ridiculous. I can list the many reasons why, if you’d like.”

    It seems that at least one of us has a skewed idea of what racial legislation looked like after the Civil War. I don’t think there is anything ridiculous in claiming that Jim Crow laws and literacy tests designed to prevent blacks from voting served to institutionalize and exacerbate racism. I suspect you may have in mind something like the Civil Rights Act, but that didn’t come along until about a century after the Civil War. Perhaps I’m unaware of some early legislation which you see as having outweighed the negative effects of Jim Crow and legalized suppression of black votes.

  74. 74 74 Daniel

    @71,

    Yeah this is the part I haven’t really been understanding about Steve’s post. How is the death of one person random and the death of another person specific? It seems to me it’s only a random variable until the one person is chosen at which point the event is no longer random. Therefore it doesn’t seem that the conclusions that Steve comes to in this post are particularly relevant unless he can somehow convince us that the one case is “really” specific and the other “random”.

  75. 75 75 Harold

    I have been missing the point, and focusing on hate-crime legislation. An important point, but not the one SL was making.

    The example of the cutting up driver is not good, because as SL says, there will surely be many people who annoy, so it is not really targetting an individual. Better perhaps to cite matricide, since recidivism is not possible.

    We spend $10M on a guardrail at a canyon, we have saved one unidentified life at a cost of $10M/life. We rescue a miner at at the cost of $50M or $50M/life. Steve extends the miners vs guardrail to cover the murder situation. The matricide poses an absolute risk to the victim, whereas the bigot poses a small risk to millions.

    The problem with this analogy is that the miner is only identified after the life threatening situation has arisen and the descision to spend the $50M is taken. Even with matricide, we have the possibility of every person murdering their mother, so policy-wise we are still not dealing with an identified victim. The comparison only works if the individual is identified before the spending or policy descision is made. Otherwise we are comparing spending on safety in mines with spending on safety on cliff-edges.

    A closer comparison would be a case like Salman Rushdie, on whom a fatwa was issued making him a target for thousands of potential murderers. Now we have a named individual just like we have a named miner. In this situation, lots of money was spent on protection – probably much more than is spent on average to deter/prevent murder. To bring it in line with current discussion, the alternative of passsing Rusdie murdering legislation increasing the punishment for anyone who murders him would be the policy comparison.

  76. 76 76 Ken B

    @75
    What a good thought Harold! We’ve been talking about a single malefactor and a large group of his potential victims, but you are right: it can work the other direction. My initial gut reaction to an enhanced penalty in this case is different, but perhaps it shouldn’t be. Of course there are other incentives at work here. Thought provoking!

    I’d say millions not thousands btw.

  77. 77 77 Harold

    #70 Dan. I am not sure we don’t have enough. If we simplify and say the risk of assault is the same for black women as white women. In fact the incidence is slightly higher for black women, but ingnore this for now. So for every 1000 black women, 4 get assaulted and the same for white women. There are 8 times as many white women, so for every black women assaulted, there are 8 white women.

    On the attackers side, for every 60 white attackers there are 26 black attackers. Normalising, for every black attacker there are 2.3 white attackers.

    Normalising the attackers to the victims, we have 9 victims (1 black 8 white) so we need 9 attackers 2.8 black and 6.2 white.

    The greatest number of black women that can be attacked by a white man is 1. The smallest number of white women that can be attacked by black men is 1.8.

    Thus more white women must be attacked by black men than black women attacked by white men.

    This is a rather cumbersome way to work it out – any statistically minded folk out there please correct any mistakes. It is not a conclusion I particularly welcome.

    This has no relevance for the issue under discussion. For that we need to know if black men rape more white women than black men rape black women, and more particulary, that they are raping them *because* they are white. (I will avoid use of italics).

  78. 78 78 Harold

    Apologies- the ratio white/black is 5.5, not 8.

    So we have 6.5 victims, 1 black and 5.5 white victims, and 6 attackers, 1.8 black and 4.2 white. At least 0.8 white victims must be attacked by black men. The greatest number of black women that could be attacked by white men is 1, but in that case 1.8 white women must be victims of black men.

  79. 79 79 Harold

    Heck – I am half an attacker missing. You get the point.

  80. 80 80 Harold

    Wonder if that is where the extra half boy went?

  81. 81 81 Harold

    #76 – well, I suppose millions are a certain number of thousands.

  82. 82 82 Daniel

    @Todd,

    My point was that a law such as the voting rights act (which includes race in it’s language) was certainly required to protect people from the Jim Crow Law’s etc created by the State governments. The whole point was that the Jim Crow Law’s did not explicitly identify race, but had requirements that created de facto conditions for discrimination against another race. Only by identifying this as the true intention of the law (which can’t be accomplished without specifically mentioning race), could the federal government overcome this hurdle. It was required after the end of slavery but wasn’t implemented until the 60′s.

    There were attempts immediately following the Jim Crow laws to legislate against it but it was largely unsuccessful. You can read about it on the Jim Crow wiki linked below.

    http://en.wikipedia.org/wiki/Jim_Crow#Early_attempts_to_break_Jim_Crow

  83. 83 83 Ken B

    @81: I suppose Ken and Will A can say the same thing about dozens.

  84. 84 84 Will A

    @ Ken B #83

    I would say the difference between Ken and I is that I was trying to have him post the actual numbers.

    I have no idea what the numbers are. Dozens in a population of 300,000,000 usually doesn’t sound right to me.

  85. 85 85 Daniel

    @ Harold 77,

    I’m not sure your math is right. I’m trying to do my own analysis now, but just so you know, your ratio is still off. 5.5 to 1, would be the ratio if you’re not including the Hispanic distinction. Since the tables you linked to do include Hispanic as a category, and the footnote states explicitly that White and Black are specifically not Hispanic or Latino, your calculation is incorrect. The ratio of White to Black females is 3.83 to 1, not 5.55 to 1 as you had suggested. Re-do your calculations with this in mind and I’m going to continue my analysis while your at that.

  86. 86 86 Daniel

    @ Harold,

    Okay, so what I did was take the census total for women for each race category and multiplied by the rates reported for 2005-2010 (not sure why you would use the earlier years, but it actually makes your case stronger to use 2005-2010 since the perceived % of the attacker that is black is increasing.

    This is the number of rapes from 2005-2010, I got for each female race category:

    White: 226167.0466
    Black: 59026.03
    Hispanic: 35916.58
    American Indian/Alaska Native: 9324.288
    Native Hawaiin: 11133.8652

    By the percentage of perceived attackers (which is an extremely crude estimate, since I’m pretty sure there’s lot’s of research that shows black men are more often perceived as the committer of a crime even when a white man has committed the crime), we get this as the total of rapes by attacker race category.

    White: 195376.7872
    Black: 92906.44427
    Other: 21177.20421
    Mixed Group: 4098.813718
    Unknown: 28008.5604

    Can you use these numbers to know if blacks commit more rapes against white’s or white’s committed more rapes against blacks. I certainly think not because of the various complications with other racial groups included in the study. What I do think we can say for certain is that both Will A’s and Ken’s assertions were basically coming from nowhere.

    There’s also the problem that these are based on self-reports so that the victim’s perception of rape may be dependent on racial category and thus it is impossible to discern if one race category or another is under-reporting. But let’s just take a look at what happens when we make some very strong assumptions about the numbers listed above.

    Let’s say that the extreme is true and that all of the white men are raping white women. This would leave 30,790 white women that were raped outside of their racial category. Divide this by the total number of White women raped outside of their racial category, and you come to the conclusion that at least 13.6% of rapes committed against white women are outside their racial category. If you do the same assessment for black women, you see that at least 36.4% of black women are raped outside of their racial category. So taking these numbers as “the truth”, we’d be hard pressed to conclude that white women are more likely to be raped outside of their racial category than black women are. Of course this doesn’t adjust for the fact that there are more white men, than black men, so it says nothing about whether white men or black men are more likely to rape outside of their race category, but at least it’s a start.

    Overall, I just want to use this post as a cautionary tale. Don’t look at numbers, especially on a topics as sensitive as rape and race, and do some basic calculations and think that you’ve come up with your answer. There are too many complications to say anything definitively without a more thorough analysis. What we can say for certain is that Will A and Ken were both getting their facts from basically no where, and Harold although an admirable attempt didn’t quite get to where he needed to.

  87. 87 87 Daniel

    Ugh, I’ve made an error, furthering my point that you need to be very careful. My assumption about white women was correct (about 13.6% were raped by someone outside their own race), but the percentage for Black women was incorrect, because I subtracted in the wrong direction. Black women could be raped by only black men, but I still think there’s too many complications to come to a definitive conclusion either way because of the multiple races. I apologize for the error.

  88. 88 88 Daniel

    Another problem with using this data to try to answer this question is that Hispanics are not included as a possible perceived category of a rapist, which forces them to be lumped into one of the other categories, and thus possibly inflating either the white or black perceived category depending on what others perceive them as being. This complication is also probably confounded by the race of the victim.

  89. 89 89 Harold

    Daniel: I agree that any such analysis is very crude and almost useless on its own, and that these self reported numbers are not reliable, as indeed any figures on rapes are not due to under-reporting etc. I think from the figures that black men are over-represented in the attacker figures, and there are more white women than men, it is at least likely that more white women are attacked by black men than black women by white men. This is not at all the same as the chances of each occuring, just the absolute numbers. I am not at all confidant that my analysis is correct – apart from the differences in numbers you point out. I wanted to have a stab at the figures, I will be honest, in the hope of showing Ken’s to be inaccurate. I think the figures support the direction of his claim, but not the magnitude.

    \83. Well, yes, I was flippant. I did actually consider whether to use thousands or millions. I selected thousands in the end because I felt that that more accurately represented the number of credible potential attackers if Salman stayed in the West.

  90. 90 90 Harold

    More white women than men? What was I thinking? That should read more white women than black women.

    A further comment – most attacks are from people known to the victim. thus is seems likely the majority of attacks will be white on white and black on black.

  91. 91 91 Todd

    @Daniel 82

    My point is that government has not been a friend to minorities. Your argument seems to be that government was necessary to undo previous government intervention. I don’t see this as a counter to my claim that government should have stayed away all along. If we had never acknowledged race as a meaningful distinction in law, excplicitly or otherwise, we would be much further removed from our racist history today.

    And yes, I do think that today’s well-intentioned race legislation and social organizations serve to perpetuate the problem which began with Jim Crow, though obviously not to the same degree. But we have probably carried this off-topic discussion far enough, don’t you think?

  92. 92 92 Daniel

    @ Todd,

    A response requires a response. We can’t go back in time and right the harms of past state governments, and you should realize that slavery was not something that came to us through fault of democratic government alone. There were many market and colonial reasons for the start of slavery, which you can review on your own time. My point is that the federal government had to respond to the state government’s bigoted policies and the response in this case required them to define race. What would you have preferred the Federal government to do to right these state government policies? They could’ve seized state government sovereignty but that would have been in direct violation of the constitution.

    You still haven’t admitted that your original statement was in error, that affirmative action and hate crime legislation somehow treats different races differently. This was the major part of your statement that I disagreed with.

  93. 93 93 Ken B

    @92:
    Actually I am not sure that this required the federal government to define race so much as to react to the states’ definition and regulation thereof. I think one of the high points of American history was when Eisenhower sent the National Guard to Little Rock. Absolutely right to do so. So I agree there certainly was a time when the federal government needed to step in, but it stepped in in the name of the 14th amendment, not to impose a “definition” of race.

  94. 94 94 Daniel

    @93,

    We’re agreed that Eisenhower did the right thing. The definition of race was to get around the fact that State governments were using criteria not based on race in their language to discriminate based on race. As in, at the time AA had much lower reading rates, and so could be prevented from voting because by requiring reading and this was entirely the intention of the law. How would you have suggested the federal government solved this problem without defining race? This was the language of the law:

    to prevent “voting qualification or prerequisite to voting, or standard, practice, or procedure … to deny or abridge the right of any citizen of the United States to vote on account of race or color.”

    This allowed the US government to look into and stop cases of laws that were not explicitly about race, but just the same disenfranchised voters of a single race. How would you suggest they have dealt with this problem without defining race?

  95. 95 95 Ken B

    @94
    And you think that constitutes a *definition* of race? Because it sounds to me just like what I said: a straight forward application of the equal protection clause.

  96. 96 96 Daniel

    @95,
    All words that are used in law must be defined somewhere when the meaning is not obvious. Federal law in 1965 considers race to be self-defined through census. Therefore when the law mentions race it is referring to this definition. Without some kind of definition the law wouldn’t have been able to operate.

    The whole point was that I was asking Todd how these laws define race in a way that treats different race’s differently as his whole supposition stood upon. He’s yet to provide any answer to this question. It seems to me that the voting rights act and these other laws rely on the same definition and do not do so in order to create differences but to prevent other parties from discriminating based on race. Does this alone justify their existence as laws, no. And I take the thought crimes line of reasoning quite seriously. Its just that these laws rely specifically on treating all races the same and Todd was suggesting quite the opposite.

  97. 97 97 Ken B

    I have 2 points Daniel.
    1. You said the govt had to define race for things like Little Rock. No.
    2. What the federal govt really has to is vitiate any state law bashed on ANY defn of race. Quite different. It’s about rejecting the criterion not specifying it.

    You only need to define race when you want to use race to treat some people better than others based on their membership in your defined groups. Need that for affirmative action, don’t need it to enforce the 14th amendment.

  98. 98 98 Ken B

    To follow up on 97. Octoroon is a word you might need if you want to enforce race preferences, but not to forbid them. It is liberals boosting AA who are bringing such shabby notions back into fashion.

  99. 99 99 Daniel

    1. Except I never said that about Little Rock. That’s an example that you brought up, not me.

    2.Except that State governments aren’t stupid and can craft laws that aren’t explicitly based on race but are de facto racist. You just keep rambling and pretending that there’s a way around this that doesn’t rely on the federal government defining race but you’re not saying how. You say that the government can simply vitiate laws. How do they assert jurisdiction unless they can prove that these non-explicit laws are racist.

    3. Affirmative Action doesn’t treat certain groups better than other’s. Also the fact that you think African Americans have been the biggest beneficiaries of the law proves that you haven’t done your research at all. White women were the overriding winner with affirmative action because they had the most connections but were still under represented when the law passed. The point is that, certain groups might make out better than other’s under the law but they’re still all treated then same under law. There are many laws that de facto advantage whites, like our criminal justice but they’re still technically treated the same under law.

  100. 100 100 Daniel

     ”It is liberals boosting AA who are bringing such shabby notions back into fashion.”

    How are liberals boosting AA? Explain what you mean by this?

  101. 101 101 Ken B

    Daniel you are ridiculous. Cite one comment from me on this or any other site where I said African Americans benefit more than others from affirmative action. That’s just crap you made up. In fact cite any other comment from me except that one where I discuss the issue at all.

    The 14th amendment does not define race does it? It does though invalidate racist laws.

    I am not replying to you any more. You make up falsehoods and argue incoherently.

  102. 102 102 Daniel

    @Ken B

    okay, maybe I misunderstood what you were saying about affirmative action. But regardless, you never answered the most important point about how the federal government can fight against racist laws without defining what race means so I’m going to assume you don’t have one.

  103. 103 103 Ken B

    Daniel, here’s an example of what I mean by your incoherence. You say I cited no examples of how the federal govt can fight racist laws without defining race. I cited two, the 14th and Little Rock. You missed the 14th entirely but look at your comment in 99. Are you asserting that that involved a defn of race, or that it didn’t work against a racist state law? It has to be one or the other if it doesn’t count as my citing an example.

  104. 104 104 Daniel

    @ Ken B,

    I think we may have just been talking past each other needlessly. What I meant was that there are some State laws that are not explicitly racist that can’t be dealt with directly through the 14th amendment. The Voting rights law was designed to address this and so needed to use the definition of race explicitly.

    So let me rephrase my question to make it more coherent with all arguments you’ve made so far.

    How can the Federal government fight against laws which are not explicitly racist, such as segregation, but rather implicitly racist, such as Jim Crow laws, without defining race?

  105. 105 105 Ken B

    @104:
    By noting that they are racially motivated and passing laws and passing laws to vitiate the effects. I commend to your attention for example section 2 of the 15 am

    Section 2. The Congress shall have power to enforce this article by appropriate legislation.

    Congress need not define race to ban literacy tests or albedo tests at polling stations. (I do not know if congress has done either of these things.) It need only address the discriminatory state laws and policies.

    Segregation *was* explicitly racial. There were LAWS against integration. (Why do so many on TBQ seem unaware of this fact?) And depended upon the ability to define criteria for race categorization.

  106. 106 106 BillD

    Here’s what is possibly a more common example: many people who kill pedestrians or bicyclists with an automobile often escape serious charges that land them in jail. Most car “accidents” are avoidable and/or preventable. Why is there a higher tolerance?

    http://chi.streetsblog.org/2013/06/26/why-does-a-cta-death-lead-to-32-years-in-jail-while-a-fatal-dui-merits-none/

  107. 107 107 Daniel

    Hey Ken B,

    The phrasing of my question was poor. I meant that segregation was explicitly racist, whereas Jim Crow laws were implicitly racist. So I think we’re in agreement about that fact if I’m not mistaken right?

    The problem with just banning literacy tests or albedo tests is that this doesn’t prevent States from turning around and creating whatever other random test they can come up with to prevent some citizens from voting. It would overwhelm the Federal government to to examine every law that each state passes and to see if it is in some way implicitly racist, and then pass a law that explicitly prevents that practice. That was the whole reason for the voting rights act, was it not?

  108. 108 108 Harold

    Lets say a state had a catagory of people it wanted to stop voting. How about 18-21 year olds? Or the car-less? New voting restriction mean you can only vote if you are over 21, or arrive at the polling station by car. If this did not in any way involve race or gender, is the state allowed to do this?

    If they are, then where does the difference lie. Over age 21 affects all ethnic groups equally (pretty much – life expectancy aside). Lets hypothesise that blacks have fewer cars. The car law would be discriminatory. Does the federal Govt. have to define race in order to make such a judgement? Well, it must have a working definition of race in order to decide if the rule is breached. I do not see this as a problem. Acknowledging that there is a trivial difference that others may see as a significant difference is OK in my book. Many laws in the past have discriminated on religion. Acknowledgeing that there are such things as catholics and protestants is not itself discriminatory.

  109. 109 109 Nic

    I think a key point yet to be addressed is people’s ability (or not) to move in and out of the victim pool. If you are sufficiently afraid of becoming a road-rage victim, you can choose to stop driving a car. But no matter how afraid you are, you can’t choose not to be black/white/gay etc.

    Additionally, it is much easier for a person to feel, even though they are in the potential car-victim pool, ‘it is not going to be me, because I am a brilliant, courteous driver!’ – in fact I propose most people feel like this. And if they are self-reflective enough to realise they are not, they can make attempts to improve and be more courteous. Again, people will feel less fear about their potential victim-hood, because they will feel it is more within their control. Less fear = happier population = less lobbying for policy change. [NB, attempting to take control of ones potential victimhood here by treating others better, has positive consequences for all. If you are actually in a potential victim pool from which you can't escape, e.g. you fear rape because you are a woman, and are advised to take control of your potential victimhood with your behaviour e.g. by wearing baggy clothes and not getting drunk at parties- this is not logical and does not have helpful consequences for anybody]

    In the Rushdie example, it is not just Rushdie whose fear is increased, it is the fear of anyone who may wish to speak freely in disagreement with Islam. Threatening Rushdie could therefore be defined as a hate-crime. Interestingly, whilst the victim pool is ‘people who are anti-islam’ it is not the inverse of an ‘islam’ victim pool, because it could also include islamic people.

    To run that through my first machine, people could choose not to be in the ‘anti-islam’ pool by staying silent. So does this make it more like the car-victim pool?

    By choosing to modify your position in the car victim pool, you either stop driving or become a more courteous driver, both of which arguably make it better for others. It is too simple to say ‘by removing yourself from the pool you make the pool fractionally smaller, therefore make it fractionally more likely for everyone else that they will be chosen’- if everyone were busy trying to be more courteous or removing themselves from the road entirely, the whole road-rage scenario would change and it would become less likely that anyone would be killed at all because there would be fewer angry people.

    By choosing to modify your position as potential victim in the ‘anti-islam’ pool, by keeping silent, you contribute to the power of the group which is causing fear, thereby making it worse for others generally.

    The more people speak out, the larger the potential victim pool becomes, but the risk associated with being in that group does not have a linear relationship with the size of the group over time- the first person, Rushdie, is in immense danger, and those who defend him are also, but as the group gets large enough it becomes able to instigate a new norm regarding what is socially acceptable.

    During that process, the risk changes in other ways. Picture Rushdie and Ayatollah Khomeini at the centre of a circular community, say a circular city made up of smaller, local residential units of mixed beliefs, and map that to a pebble dropped in a pool with circular ripples coming off it. As a ripple passes a local community, you get smaller echoes/effects of the crime- a person may express anti-aslam views in the local cafe and a khomeini supporter may punch them, for example. Or people may become afraid of others in their community whom they previously felt unthreatened by. The social unrest spreads. i.e. during the process of transition it gets worse before it gets better. Makers of social policy can either help or hinder that transition and various points in its trajectory.

    This is different from someone seeing a criminal getting a light sentence and feeling less of an incentive not to break an existing law, because the latter applies equally whether it be road-rage or racially-motivated murder, but the former happens in addition to this. I do not believe road-rage would spread in the same way, nor have the same trajectory of ‘greater social unrest before new norms are stabilized’ therefore there is a stronger case for social policy to be created to label and deal with ‘hate violence’ differently from ‘plain old violence’ to speed up the stabilization process.

    Finally, we can apply this circular radiating model to crimes B and C, where people cannot choose to enter or leave the victim pool. This time, the local areas in our city are mixed black/white. Some of the areas will have a much higher proportion of one or the other, even if we suppose a greater number of whites over all. Say a black kills a white at the centre. As the ripples spread out, there will be different effects in the subcommunities. A white in a mostly black community will feel more afraid, a black in a mostly white community will feel more afraid, and that white community will feel disproprtionately afraid of that black; a group of blacks in a mostly white cafe will feel defensive, a group of whites in a mostly white cafe will also feel defensive when the blacks come in. Reverse the black/white act at the centre and exactly the same is true. The proportion in the city over all makes no difference- fear and defensiveness is increased over all.

    To sum up and answer the original questions:
    Which is more harmful- targetting an individual or a random person from a given group of people such as race? I believe my mechanism above shows the latter.
    If that is so, does it matter if the potential victim pool is larger or smaller? I believe my mechanism above shows no, and point 3 does not apply because all groups have an increase in negative feelings, not just the group (of whatever size) that is targetted.

    Finally, I ended up on this blog as a result of investigating what all the ‘rape fuss’ was about. I read the whole thing and I don’t think anyone directly addressed the impact of increased group fear and potential victim pool. In the first two examples, there is essentially a victim pool of one- the only way the victim pool could increase would be by lobbying by that individual, to deliberately spread their personal disgust to others, or by collecting around them others who already feel that disgust. It may be that the pools of people who already feel that disgust are large, however there are no assumptions in the problem meaning the pools must include everyone and referring to real life, we can conclude the pools do not include everyone. However, in the rape case, the potential victim pool *does* include everyone: the very definition of rape is that it is unwanted, therefore it is impossible to want it and thereby escape the potential victim pool. Although Steve hypothesises a ‘type of rape’ which it would be possible to NOT not-want, I still believe this is a logical impossibility: the definition of rape is lack of consent, so you cannot even consent that you don’t mind it happening to you should you happen not to know about it- you would simply be displaying a niche sexual preference, not ‘consenting to rape’.

    If unconscious rape were legal, everyone would have an increased fear of it. Even rapists would not be immune to being raped by *other* rapists, which by definition they must fear. Whether or not you know about it becomes irrelevant- the ‘psychic damage’ is incurred by a whole community living in fear, making it a clear cut case for policy intervention.

    This relates to the question above: in point 1, it is implied that ‘personal fear’ is not directly proportional to risk. I think it should be noted that even if personal fear *were* proportional, ‘group fear’ would still not be a proportional relationship between personal fear and number of people, because of the way people both consciously and unconsciously reinforce each others fear. The media and policy makers are well aware of this and not all interventions aimed at ‘stabilizing’ (or having control of) communities take the form of actions to decrease fear.

Comments are currently closed.