The excesses of political correctness

Responding to Micha.

But seriously, do you really think any actual attempts to support liberty were hobbled when people criticized politicians for using Southern strategy-type language? Is it really that hard for a politician to say: I won't take away your guns or raise your taxes to pay for welfare without resorting to Us vs. Them ingroup outgroup squabbles regarding how hard someone works and how many laws they choose to abide? Seriously?

You seem to be missing a point that advocates of regulation frequently miss. While it is inexpensive for a company that happens to already be doing the allowed thing to do the allowed thing - and this is all you have pointed out here - it is somewhat more expensive for a company that would have done the prohibited thing to switch and train itself to do the allowed thing, and it is even more expensive for a company to constantly police its practices to make sure that it never does the prohibited thing and always does the allowed thing. Regulations impose the significant burden on business of discovering and remembering those regulations, in addition to the burden of evolving its business practices around those regulations and vigilantly policing its own activities.

Furthermore is the objection only to these words? From previous comments, it seems concepts were prohibited. Rad Geek wrote:

Try thinking about it in reverse, if that helps. "Hard-working" and "law-abiding" are deliberate contrast terms for "lazy," "shiftless," and "criminal." These terms were all deployed with pretty clear racial dimensions during the political debates in question.

The relationship between "hard-working" and "lazy" is conceptual, not terminological. Conceptual links being important, then suddenly the very idea of a hard worker, if it enters into your speech, becomes a potential liability. The chilling effect on speech can potentially be broad and deep.

Chilling speech is of course not necessarily what happens. Non-leftists are apt to ignore the lunatic ravings of leftists, and therefore to go ahead and speak however they feel like speaking, and in doing so constantly step on one left wing land mine after another. While the repercussions exist mainly in the mind of leftists, they do at least in that respect prevent leftists from legitimately dealing with the actual arguments presented. It's scary when leftists lust after prohibiting speech (which they do). It's funny when leftists are shell-shocked by the explosions that go off in their own echo chambers.

I forgot to mention previously that the point I raised also goes toward doubting the story independently of any motivation of the story tellers. As I pointed out, these are important ideas. Criminality? Very important. Earning a living? Yes, extremely important. Therefore it is likely that they were in fact used to mean exactly what they seem to mean rather than as code for racism. Furthermore, even if some people used them as racial code, and even if some audience took them that way, we're talking about large populations here. Not everybody was a member of the secret society. Not everybody had the decoder ring. It seems likely that a lot of people both heard, and made, speech with those terms meaning exactly what they literally meant. In all likelihood the speeches, even if they could be interpreted racially, were in fact sometimes, maybe usually, entirely valid and persuasive points when understood completely literally. To summarize my reasoning: if people have a propensity to do something (e.g. be concerned about criminality), then chances are good that that's what they did. If some leftist comes along and says that when Southerners talk about criminality they're really talking about blacks, I find that impossible to believe as a blanket statement about all or even most Southern talk about criminality and the law. And even if a leftist manages to present evidence that somebody at some time used it as a code, well, it takes a lot more evidence than that to convince me that Southerners didn't care about criminality but only cared about blacks.

If the case for liberty rests or falls on a politician's ability to sort his or her supporters into law-abidin', hard-workin' 'mericans rallied against that "other" group of undesirables, the case for liberty is already lost.

Thanks for adding the twang. That really underlines the contemptibility of the statements you are portraying. If it's said with a Southern twang, then it must really be contemptible. Because, you see, Southerners are that other group of undesirables.

Meanwhile, on the matter of whether "the case" is "already lost" - how? The distinctions are neither invalid nor unpopular. Granted, they are unpopular among the left. As Evan Sayet has pointed out (the Youtube version is best):

There's a brilliant book out there called The Closing of the American Mind by Professor Allan Bloom. Professor Bloom was trying to figure out in the 1980s why his students were suddenly so stupid, and what he came to was the realization, the recognition, that they'd been raised to believe that indis­criminateness is a moral imperative because its opposite is the evil of having discriminated. I para phrase this in my own works: "In order to eliminate discrimination, the Modern Liberal has opted to become utterly indiscriminate."

I'll give you an example. At the airports, in order not to discriminate, we have to intentionally make ourselves stupid. We have to pretend we don't know things we do know, and we have to pretend that the next person who is likely to blow up an airplane is as much the 87-year-old Swedish great-great-grand mother as those four 27-year-old imams newly arrived from Syria screaming "Allahu Akbar!" just before they board the plane. In order to eliminate discrimination, the Modern Liberal has opted to become utterly indiscriminate.

The problem is, of course, that the ability to dis criminate, to thoughtfully choose the better of the available options--as in "she's a discriminating shopper"--is the essence of rational thought; thus, the whole of Western Europe and today's Democrat ic Party, dominated as it is by this philosophy, rejects rational thought as a hate crime.

Granted, distinctions such as the distinction between black and white, between foreigner and American, are often abused, leading to injustices. However, what we can see among leftists is a tendency to broaden the fight against discrimination. Now it is not enough that we refuse to discriminate between black and white, between American and foreigner. Now we must refuse to discriminate even between criminals and non-criminals!

What at first seems a reasonable idea (that we should not, e.g., discriminate by race), quickly balloons into a sickness (the refusal to discriminate even between criminals and non-criminals).

Share this

In this specific case Sayet

In this specific case Sayet is actually wrong:

We have to pretend we don't know things we do know, and we have to pretend that the next person who is likely to blow up an airplane is as much the 87-year-old Swedish great-great-grand mother as those four 27-year-old imams newly arrived from Syria screaming "Allahu Akbar!" just before they board the plane.

Racial profiling doesn't work because it's very easy to game. Random testing - which does mean, yes, treating the 87 year old Gran equally to the muslim hothead - is harder to foil.

You know that for sure?

The supply of terrorists ready and able to immolate themselves in American planes, or anywhere else, seems to be pretty slim these past several years. You can't game the system without (a) a plentiful supply of terrorists, from which you can pick all shapes and sizes, and (b) a vast secret organization coordinating all these terrorists and selecting the ones that best defeat the profiling. In short, while one can easily imagine in theory how profiling might be defeated by a certain kind of enemy, the actuality of terrorism is another matter. Also, the difference between a grandmother and a young man is not racial, so calling it "racial profiling" is just incorrect. Call it, instead, "profiling", and then argue whether profiling (which actually is used in places, e.g. in Israel) is actually (not just theoretically) gamed to the point of being self-defeating. It's not enough to point to anecdotes, since the question is a matter of numbers. Schneier, who seems to think a lot of profiling a bad idea, nevertheless concedes that "it makes real security sense for El Al to spend more time investigating young Arab males than it does for them to investigate Israeli families," and, "done correctly, this intuition-based sort of profiling can be an excellent security countermeasure".

Aside from that, my application of Sayet's point was not to profiling.

You answered your own

You answered your own question. If the supply of terrorists ready and able to immolate themselves in American planes, or anywhere else, seems to be pretty slim these past several years, then it doesn't really matter whether you favor a policy of profiling or no-profiling - the result will be the same either way. What does profiling get us? Nothing. What does not-profiling get us? Nothing. It's a stupid argument.

Let's think it through

Let's consider two possibilities.

Possibility 1: terrorists are few in number but are able to inflict tremendous damage on victim populations. In this case, even if the supply of terrorists is slim, then contrary to what you conclude, it does really matter.

Possibility 2: terrorists are are many but are able to inflict only minuscule damage each. In this case, if you implement profiling that stops the vast majority of terrorists who fit the profile but turns a blind eye to the tiny minority of terrorists who do not fit the profile, then by greatly restricting the number of effective terrorists you have still had a massive impact on the amount of terror that terrorists are able to commit.

Your own argument - that the United States is (in contrast to, say Israel) virtually free of terrorism and therefore (presumably you mean to say) it's a waste of time to do anything at all to stop terrorism - is a separate argument from the specific, targeted argument that profiling per se is useless.

A bit irrelevant but still

Last time I took the plane at JFK, I had a black LvMi Rothbard t-shirt on reading "Enemy of the State". I got a complete search at the subway station and at the airport security. Random check they said ^^

Ignoring game theory

You seem to miss the point that in the scenario, the terrorists are aware of whatever security regime is in place. If it's random, they can't reliably game it - the only strategy against random checks is to use a surfeit of bombers and hope that a percentage slip through the net and, to say the least, there are logistical difficulties with this approach. On the other hand, if you know for sure - 100% sure - that certain types (whether that's a swedish granny, a pregnant irish girl, an italian priest) are safe from checks it's a comparatively trivial matter to suborn or imitate one of such a type for the plan to succeed.

No

You seem to miss the point that in the scenario, the terrorists are aware of whatever security regime is in place.

No, I didn't miss that point but took it as a given, and responded to it.

Try to imagine some realities that might limit the ability of the enemy to implement the optimal strategy. For example, suppose that there is a hole in profiling that lets through Swedish grandmothers. Great! Since you the terrorist are aware of this hole (by assumption), all you have to do is send your army of Swedish grandmothers through the security hole. Oh. Wait. You don't have any Swedish grandmothers in your army.

I explained this already - all I'm doing here is fleshing out my previous response with details.

You claim that "it's a comparatively trivial matter" to successfully disguise a young Saudi male as an elderly Swedish woman, or to recruit an actual Swedish grandmother. Really?

To recap, you pointed out that in theory (specifically, in game theory), if the enemy knows your strategy he can plan around it. Indeed. The more he knows, the more he can plan around it. This is a general point that weighs against any strategy, and it places a limit (but not a complete limit) on the strategies you can take. If the enemy knows your strategy, it does not follow that the most successful strategy is a fully randomized strategy, with equal probabilities of all possible moves on your part! That would make game theory utterly uninteresting, and it is not uninteresting.

But my point in response was that sometimes you can't implement the optimal strategy because you simply don't have the resources. I wrote,

The supply of terrorists ready and able to immolate themselves in American planes, or anywhere else, seems to be pretty slim these past several years. You can't game the system without (a) a plentiful supply of terrorists, from which you can pick all shapes and sizes, and (b) a vast secret organization coordinating all these terrorists and selecting the ones that best defeat the profiling. In short, while one can easily imagine in theory how profiling might be defeated by a certain kind of enemy, the actuality of terrorism is another matter.

Patton was a brilliant general, but he would not have been able to win his battles if he didn't have anybody to command.

You seem to think the resource problem is trivial. I don't think it is. Swedish grandmothers aren't exactly lining up to be recruited by Al Qaeda. And even highly paid Hollywood makeup artists have not quite yet mastered the art of disguise. Some suspension of disbelief is still required of the audience.

But please, if you know how to convincingly turn a male Saudi into a Swedish woman, please, do share. There are many trannies who could use your advice (no offense intended, ladies).

Constant, I know from

Constant, I know from reading your postings here that you're pretty intelligent but you're being very obtuse here. Do you really think it's hard to recruit someone who doesn't fit the standard terrorist profile? During "The Troubles", one IRA tactic was to hold an ordinary civilian's family hostage and force him to drive a bomb-laden car into the barracks. A bomber needn't even be a "believer" to be useful so long as there's something at stake for him or her.

Your imagination versus empirical reality

You're the one being obtuse because you keep on making points that I already replied to, which demonstrates that you're not comprehending.

Like I said before, it's a question of numbers. Can the bad guys sometimes recruit somebody who doesn't fit? Doubtless. Can they do so enough to more than compensate for the impact of profiling on the vast majority of their potential recruits? If you'll look above I raised this point and a warning against anecdotes ("one IRA tactic" - with no numbers). Your imagination is not enough to decide the issue. The Israelis seem to think that profiling works for them, and at least one widely read "security expert" seems to agree with them. So that goes against the blanket statement that the bad guys can game the system to the point where it would be better not to profile.

Here are two complications which affect the ability to use a coerced recruit:

1) Something like 9/11. I'll just leave you to think about the logistics of trying to carry that out with coerced recruits.

2) Profiling is not just racial. Suppose terrorists try to get around racial profiling by coercively recruiting. So now you have an average Swedish grandma carrying a bomb into an airplane. Do you think she can really get by unnoticed? Close your eyes and imagine how she feels, then imagine how she behaves, then imagine how she looks to others. I mean, come on. It's trivial to tell the difference between a genuine smile and a forced smile, but it's supposed to be impossible to tell the difference between a Swedish grandmother going on vacation and a Swedish grandmother who has been forced to carry a bomb that's going to kill her along with everyone else because her family is being threatened with death?

Do you really think it's

Do you really think it's hard to recruit someone who doesn't fit the standard terrorist profile?

Yes, it is. The supply is much thinner, the cost are much higher.
Assume you have two categories of people, A and B. Let p(X|A) and p(X|B) be the probability of checking person X knowing it is of type A or B. You are suggestion to use the rule p(X|A) = p(X|B). Now assume the cost for the terrorist of finding a willing kamikaze in A is c(A) and in B is c(B), further assume that the gain to them of committing a terrorist act is G and that they only lose the cost of finding the terrorist if the plot misses. If they choose someone in A, their expected gain is (1-p(X|A))G-c(A), same for B. If terrorist maximize expected gain (let's say for now terrorists are not risk averse) the optimal probability of checking are such that p(X|B)-p(X|A) = [c(A)-c(B)]/G... p(X|B) = P(X)+(c(B)-c(A))/G.P(A) and p(X|A) = P(X)-(c(B)-c(A))/G.P(B), the optimal conditional probabilities should be equal if and only if the cost are the same to the terrorist, otherwise their should be at least some bias which means some profiling. This result remain if you make the terrorists risk averse.

You may very well argue that c(B)-c(A) << G and thus that the amount of optimal profiling to perform is almost minimal, but a priori you cannot just say "it doesn't work".

I for myself most definitely agree with Constant, there are many categories for which the cost of recruiting is so high that they should almost never be checked, while others should almost always be checked.

I'll try with an example

I'll try to do this with an example.

Suppose it costs one dollar to recruit a Saudi, and ten dollars to recruit a Swede, suppose both kinds of recruits are fully effective assuming that they are not caught, and suppose the terrorist organization has ten dollars. Suppose that over a span of ten days ten airplanes in total board with ten passengers each, one Saudi and nine Swedes (to reflect the low density of Saudis on Swedish flights).

Airport security can perform the following two kinds of check with equal cost:

1) Check all 10 Saudis and check no Swedes. If the Saudi is a terrorist the Saudi is caught.

2) Randomly check 1 out of 10 passengers regardless of nationality. If the randomly checked person is a terrorist then the randomly checked person is caught.

If Airport security does (1) then the terrorists cannot use Saudis and must recruit a Swede. They can recruit one Swede. So one airplane will explode.

If Airport security does (2) then the terrorists can use both Saudis and Swedes with equal probability of success. Because Saudis are cheaper, they recruit ten Saudis. On average, only one of these Saudis is caught. So on average, nine airplanes will explode.

In this model, profiling by nationality improves the situation nine-fold. This model takes into account:

(a) differential cost to the terrorists of different kinds of recruit
(b) fixed resources of the terrorists
(c) fixed resources of airport security

Notice the following: profiling (option 1) opens up a predictable hole in airport security. Without profiling (option 2), everyone is equally likely to be checked, so there are in a sense "no holes" in security. Nevertheless, because of the fixed resources of airport security, the result is not that everyone is checked but that one out of ten are randomly checked. This opens up a probabilistic "hole" in security.

The "hole" in option 1 is in a sense more "open" to the terrorists (since no Swede is checked at all) but it is a costly hole to step through, because Swedes are costly to recruit. The "hole" in option 2 is slightly less open because there is always a small chance that a particular terrorist will be checked, but the fixed resources of airport security means that the probability of being checked is only 1 in 10. So if you recruit ten Saudi terrorists, you can expect that nine will get through.

Got a comment tied up in

I have a comment tied up in moderation???

Is it ok to repost it or should I wait for it to appear?

My solution

The last time my comment got tied up in moderation I waited for several seconds and then lost my patience and posted that sucker as a blog entry. That's actually this blog entry.

Sorry about that, Frank. I

Sorry about that, Frank. I think I got the one you're talking about.