Tuesday, April 18, 2006

John McCain: It's Shiller Time!

Item One: McCain presses the flesh with Katherine Harris.

Item Two: McCain's to speak at Jerry Falwell's Liberty University; he didn't have much to say when (previously) Tim Russert questioned him, and he's got scarcely more to say to Jon Stewart. Notice, however, his repetition of the same, lame talking point about schools that disallow military recruiting. Note to Sen. McCain: there's a difference between speaking someplace where you disagree with a single policy (don't we all disagree with at least one policy of any given school?), and speaking at an institution whose entire raison d'être is promulgating fundamentalist offal.

But of course, both of these events are merely symptoms of McCain's infection by a single deeper disease: the virulent sickness of the modern Republican Party, which marries corruption, fundamentalism, and autocracy into a veritable holy Trinity of political evil. Americans (and especially the press) find McCain appealing in part because his existence appears to tell a particular fictional story that resonates with Americans: that of individual integrity, wherein (cue the deep-voiced movie-trailer voice-over and portentous pseudo-symphonic backing track) one man's struggle to overcome enables him to transcend all labels or institutions. People like McCain because this fairy tale permits them to indulge in an illusion of government without or beyond politics --- politics being, of course, that grubby business whereby groups of people bang out actual law that actually influences people's lives.

(Actual politics and policymaking is distasteful to pundits in particular because (a) it involves messy compromises among competing interest groups, (b) it requires serious thinking, often involving attention to detail and some mathematical competence, and (c) one can only devise coherent policy when guided by some coherent set of ideas about what animates and justifies social relations, i.e. an ideology. And, of course, the vast majority of pundits (a) hate any compromise involving concessions to interest groups other than fiscally conservative upper-middle-class professionals like themselves, (b) hate having to think carefully about anything other than the scintillating intricacies of their own self-important insights, and (c) hate any ideology other than the dominant center-right consensus, which pretends to be non-ideological.)

So, then, McCain's behavior of late merely demonstrates the truism that effective institutions are greater than individuals. And, of course, the Republican Party is an exceedingly effective institution, from its ruling clique's point of view. Deep structural features of modern Republicanism --- the admixture of influence-peddling, toxic ideology, and willful ignorance that knits together the Party of Reaction --- hammer aspirants to national office upon the hard and ruthless anvil of national politics, molding them into ass-kissing panderers to the worst tendencies of Homo americanus. McCain would not be where he is today if he hadn't conformed to the Party's needs when it really counted, and he will rise in power to the extent that he continues to deform under the blows of that same chastening hammer. Anyone who believes otherwise is deceived.

Sunday, April 16, 2006

Fedora make Cog mad! Cog smash! Cog... buy stuff... (sigh).

Warning: Navel-gazing techie whining follows. Skip if you value your time.

So, I thought I'd spend a lazy Easter afternoon upgrading my Fedora installation to Core 5. I've been running Fedora on various machines since the first stable release, and on my current weapon of choice since Core 3, and seldom have I encountered even a faint sigh of protest when upgrading. Usually, getting the upgrade to work involved freeing up a few hundred MB of disk space, but nothing too outrageous.

Well, Fedora's upgrade footprint has finally bloated to the point where it required excruciating and unreasonable effort on my part. If my disk were more capacious, it would have been no problem; but I've only got 4.3GB on my root partition, of which I seldom have more than 1GB free. Time and again, Fedora sadistically mocked my aspirations, promising an impending upgrade...

...only to snatch the football away at the last moment:

Notice that Fedora doesn't tell you how much disk space would make it happy. So, I went and deleted a few packages and re-ran the installer, and... anybody who's read Peanuts can complete the strip here.

At long last, after exerting heroic efforts, including erasing all my office software, half my programming tools, and most of my multimedia apps, I finally got 1.75GB of disk space free, whereupon the upgrade process went through without a hitch. Grrr.

Let me make clear that there's no good reason for the upgrade to need that much space. After the upgrade went through, I had about 1.9GB free. Fedora's installer simply insisted on copying a ridiculous mass of data (RPMs, I assume) from the DVD-ROM to the hard drive prior to installing them, so that I effectively needed enough disk space for two copies of the operating system. Which doesn't make any sense. Before upgrade: one operating system; after upgrade: one operating system. At no point was it really necessary to have two operating systems on the hard disk.

As it happens, I've been feeling the pinch in other ways w.r.t. hard drive space anyway, especially since I bought my first real digital camera last year, so I decided that it was finally time to buy a new hard drive. (There goes my tax refund.) Whereupon I realized that (1) I dual boot Linux and Windows, (2) I don't trust disk tranfer utilities to deal with my dual boot situation, and (3) IBM does not ship a Windows reinstallation disk by default with its Thinkpads, so I had to order one. As my warranty expired in February, this involves paying $45 for another copy of an operating system I already own. Grrr.

Anyway, I hereby offer the following advice, so that this post won't only be a useless whining session:

  • When buying a notebook computer, if you are a demanding user, buy the biggest hard drive available. Every single time I've owned a notebook, I've filled up the hard drive before I've outgrown any other part of the computer (video card, CPU, RAM, whatever). This has been true even when I've been initially amazed at the spectacular --- nay, intimidating --- size of my hard drive.
  • When buying a notebook computer, make sure that you obtain an installation CD for your OS right after you buy it, so that it's still in warranty. I'm not sure if IBM/Lenovo is the only vendor who regularly ships notebooks without an OS install disk, but be aware of the issue.
  • If you ever want to repartition a disk, use a GParted bootable LiveCD. It's amazing to me how utterly superior GParted is to, say, Partition Magic. I mean, it's not that amazing, because I've always wondered what makes repartitioning so hard anyway. (It seems exactly like a mixture of copying and compacting garbage collection, which have been solved problems for decades.) Anyway, Partition Magic and its commercial cousins are basically obsolete now.

Brand new/you're retro

Recent late-night IM conversation...

Thursday, April 13, 2006

Sandra Tsing Loh reviews "Mommy Wars"

If you somehow managed the patience to read my previous neverending post, and you feel like washing the icky feel of Catharine MacKinnon out of your system, I suggest Sandra Tsing Loh's astute review (from the Atlantic Monthly, via Powell's Review-a-Day) of the recent essay collection "Mommy Wars". In stark contrast to MacKinnon's bizarre, reality-free jumble of innuendo and overheated rhetoric, Tsing Loh's sensibility is firmly moored in real problems faced by women. Also, (1) she writes with a light and witty touch, and (2) when she chides someone for intellectual sloppiness, she does it firmly yet gently --- two qualities that, erm, I possess in somewhat lesser degree.

Which isn't to say that the central conceit of her review --- i.e., that a collection of essays by privileged upper-class women evinces a few blind spots w.r.t. race and class --- is a particularly subtle or penetrating insight. But I believe there's some value in marshalling evidence in the service of obvious conclusions sometimes. At least, I'd better believe that, because the vast majority of what I write on this blog basically consists of pointing out the obvious.

Wednesday, April 12, 2006

Catharine MacKinnon's bag of dead leaves (an extended rumination)

Bookslut, which I normally prize for the Suck-esque consistency of its witty snark, today points approvingly to Catharine MacKinnon's interview by the Guardian's Stuart Jeffries.

Now, I went into this interview with no preconceived ideas about MacKinnon, as I had forgotten who she was. To me, for the moment, she was not that infamous writer who characterizes pornography as "sex discrimination" and "violence against women" (ideas usually attributed to Andrea Dworkin, but shared by MacKinnon). For the moment, she was just some random feminist that Jessa Crispin found interesting. Then, in Jeffries's first paragraph I read the following:

Of all the provocative passages in Catharine MacKinnon's new book Are Women Human? the following hit me hardest. She writes: "[T]he fact that the law of rape protects rapists and is written from their point of view to guarantee impunity for most rapes is officially regarded as a violation of the law of sex equality, national or international, by virtually nobody."

The first thing that struck me about this passage was that MacKinnon writes terribly. The sentence completely dissipates its force in a thicket of subordinate clauses. Also, note the bracketed initial in "[T]he", indicating that only a fragment was quoted: in all likelihood, some equally verbose preceding clause obscured the passage even further.

The second thing that struck me was that the claims make no sense. Disentangle the above into more lucid English prose, and you get the following:

Rape law is written from rapists' point of view. As a result, rape law protects rapists to such a great extent that it guarantees impunity for most rapes. Almost nobody officially regards this fact as a violation of the national or international law of sexual equality.

The first claim is false; the second claim is a misleading spin on a kernel of truth; and the third sentence is meaningless if the "fact" in question is not a fact.

In what sense are rape laws written from rapists' point of view? Jeffries describes MacKinnon's answer:

Are you suggesting that rape law enshrines rapists' points of view, I ask MacKinnon? "Yes, in a couple of senses. The most obvious sense is that most rapists are men and most legislators are men and most judges are men and the law of rape was created when women weren't even allowed to vote. So that means not that all the people who wrote it were rapists, but that they are a member of the group who do [rape] and who do for reasons that they share in common even with those who don't, namely masculinity and their identification with masculine norms and in particular being the people who initiate sex and being the people who socially experience themselves as being affirmed by aggressive initiation of sexual interaction." She takes a well-earned breath.

Jeffries seems (or pretends to be) thoroughly impressed at MacKinnon's verbal avalanche. I'm pretty sure this is some kind of dry British sarcasm on Jeffries's part, but I'm an ex-New Yorker, so I'm just going to call the fast-talking b.s. by its proper name. MacKinnon's saying that when she wrote that "rape law is written from rapists' point of view", she meant that rape law is written by men. Since rapists are men, the reasoning goes, it necessarily follows that anything that men do must reflect the point of view of rapists. This is a comically unsound chain of inference [0]. By the same logic, I can demonstrate that since Hitler was a vegetarian, vegetarian cookbooks reflect a secret program to exterminate the Jews.

Okay, maybe that's not entirely fair. The link between "aggressive initiation of sexual interaction" and rape is somewhat more plausible than the link between non-consumption of meat and genocide. But it's still vague guilt-by-association. Merely pointing out that men tend to aggressively initiate sex does not constitute any kind of argument that rape law reflects the views of rapists.

A sound argument supporting this allegation would have to do two things:

  1. Identify a view held by rapists, but not by non-rapists [1]. I emphasize the latter clause because it's trivially obvious that rapists and non-rapists share many views. For example, I believe that the sky is blue, and I'd venture to guess that virtually all rapists also believe that the sky is blue. In some sense, therefore, the observation that "the sky is blue" is a view held by rapists; but it is not "a rapist's point of view" in any meaningful way.
  2. Identify some aspect of rape law reflecting that view.

Does MacKinnon do either of these things? Well, not really. Jeffries lets her off the hook for a while, but returns to the subject later in the article:

Otherwise, she contends, unenlightened men still write the laws. And when, for instance, they write laws on rape they make what she believes are grotesquely sexist assumptions. "The assumption," she says, "is that women can be unequal to men economically, socially, culturally, politically, and in religion, but the moment they have sexual interactions, they are free and equal. That's the assumption - and I think it ought to be thought about, and in particular what consent then means. It means acquiescence. It means passivity. You can be semi-knocked out. You can be dead in some jurisdictions."

I almost choke on my mineral water. Dead and giving consent? "Sex with a dead body is necrophilia but it isn't regarded as rape." Oh, I see. "You can be semi-comatose, not to mention married in many places, and be regarded as consenting whenever sex takes place."

MacKinnon thinks consent in rape cases should be irrelevant. Women are so unfree that even if a woman is shown to have given consent to sex, that should never be enough to secure an acquittal. Why? "My view is that when there is force or substantially coercive circumstances between the parties, individual consent is beside the point; that if someone is forced into sex, that ought to be enough. The British common law approach has tended to be that you need both force and absence of consent. If we didn't have so much pornography in society and people actually believed women when they said they didn't consent, that would be one thing. But that isn't what we've got."

What does she mean - how does pornography affect this? "Pornography affects people's belief in rape myths. So for example if a woman says 'I didn't consent' and people have been viewing pornography, they believe rape myths and believe the woman did consent no matter what she said. That when she said no, she meant yes. When she said she didn't want to, that meant more beer. When she said she would prefer to go home, that means she's a lesbian who needs to be given a good corrective experience. Pornography promotes these rape myths and desensitises people to violence against women so that you need more violence to become sexually aroused if you're a pornography consumer. This is very well documented."

At this point, I simply must digress a bit, because I find the above paragraphs almost physically painful to read. Reading involves a form of ego-submission: while you read, one of your mind's interior spaces actually recapitulates the text. In some sense, while you read, the writer's voice becomes your voice; the writer's thoughts become your thoughts. In order to read the above passage, I must force myself to emulate the hazy, confused, babbling fugue of free-associative nonsense that passes for thought in MacKinnon's mind. And it is painful. It's the mental analogue of wearing a hairshirt.

The best way to describe MacKinnon is this: her brain is broken. Broken rules of inference are embedded in her tactics of thought. Jeffries writes that MacKinnon's widely regarded as "bright" and that she completed a Ph.D. in political science at Yale. So she's probably smart in the limited sense that, given some seed premises, her brain can rapidly sprout a collection of thoughts related to those premises. The problem is that the thoughts have no connective tissue of logic binding them together. MacKinnon's reasoning resembles a valid argument in much the same way that a trash bag full of dead leaves resembles a tree.

Okay, enough digression. Now I'll talk about her actual statements.

First, let us quickly dispose of the bizarre digression on necrophilia. Our society has customs and laws regarding the treatment of corpses. For example, Chapter 9A.44 (sex offenses), Section 105 of the Revised Code of Washington classifies "sexual violation of human remains" as a class C felony. But corpses are not human beings, and so of course performing sexual acts on a corpse is not "rape". I honestly cannot imagine what MacKinnon hopes to accomplish by conflating corpse desecration with rape. MacKinnon seems to think the distinction illuminates society's misogyny, but that strikes me as a non sequitur so extreme that it borders on paranoid schizophrenia [2].

Next, consider the issue of "force" versus "consent". This is a little sticky, because Jeffries does not give MacKinnon's exact words in full, but judging by his rendering, MacKinnon's analysis is utterly confused. She claims that society makes it impossible for a woman not to consent, since even non-consent will be interpreted as consent. Therefore, laws should be crafted to ignore consent, and to examine instead whether any element of "force" was involved.

But that's nonsense in at least two different ways. First, it is possible for a woman to signal consent, and it is possible for the absence of consent to be discerned. If these things were not possible, then the rate of rape convictions would be zero. For various reasons, rape reporting and conviction rates are low (this is the "kernel of truth" that I alluded to above), but the solution is not to eliminate the notion of consent; it's to educate people more effectively about consent. Which brings me to my second point: it's conceptually incoherent to invoke the notion of "force" while discarding the notion of consent.

The word "force" carries an implication of coercion, or the absence of choice. But unless one is willing to deny that choice exists at all [3], coercion can only be defined in opposition to consent. The negative is meaningless without the affirmative. How, then, can legislators, judges, or juries possibly make a judgment that "force" occurred, without considering whether consent occurred? If we discard consent, there's literally no possible way of distinguishing between "force" and non-"force". (And, incidentally, no distinction between an accusation and the evidentiary burden required for conviction.)

On the other hand, although force requires the absence of consent, the converse is not true: absence of consent does not require force. A sex act involving someone who is incompetent to consent --- for example, someone who is mentally retarded --- need not involve physical force, or even the threat of emotional abuse.

In other words, "absence of consent" actually describes a strict superset of the set of situations involving "force". Furthermore, under modern law, the absence of consent is broadly construed. Not that you'd know this from MacKinnon's interview.

MacKinnon would like us to believe that "You can be semi-knocked out" and still be judged to have consented. Well, I admit that judges and juries in America can be highly imperfect in enforcing the law, and that laws were once pretty bad. However, taking Washington State again as an example, 9A.44.050 of the Revised Code of Washington currently states that if a person's "incapable of consent by reason of being physically helpless or mentally incapacitated", then engaging in sexual intercourse with that person is rape in the second degree, and hence a class A felony. Penalties for class A felonies range up to life imprisonment and a fine of fifty thousand dollars.

MacKinnon would like us to believe that the law doesn't recognize that "force or substantially coercive circumstances between the two parties" implies rape. Well, look again at 9A.44.050 RCW, and you discover that "forcible compulsion" is sufficient to establish rape. 9A.44.010 RCW defines "forcible compulsion" as "physical force which overcomes resistance, or a threat, express or implied, that places a person in fear of death or physical injury to herself or himself or another person, or in fear that she or he or another person will be kidnapped." If somebody threatens you or anybody else with physical injury, no matter how slight the injury or indirect the threat, and sexual intercourse results, the law calls it rape. [4]

MacKinnon would like us to believe that the legal definition of consent is infinitely elastic --- that the law would affirm arguments like "when she said no, she meant yes" or "when she said she would prefer to go home, that means she's a lesbian who needs to be given a good corrective experience". Again, that's simply ridiculous. By 9A.44.060 RCW, rape in the third degree, a class C felony, includes any form of sexual intercourse not involving consent. Consent is defined by RCW 9A.44.010 as "actual words or conduct indicating freely given agreement to have sexual intercourse or sexual contact". Coerced agreement is not consent; the absence of physical resistance is not consent; only affirmative "words or conduct indicating freely given agreement" constitute consent. Saying no, or that you'd like to go home, handily establishes the absence of consent.

So, what about this "sexual intercourse" business? Does that phrase leave weasel room for a rapist to argue that some act was not rape because he didn't go "all the way"? Once again, RCW 9A.44.010 says:

(a) "Sexual intercourse" has its ordinary meaning and occurs upon any penetration, however slight, and

(b) Also means any penetration of the vagina or anus however slight, by an object, when committed on one person by another, whether such persons are of the same or opposite sex, except when such penetration is accomplished for medically recognized treatment or diagnostic purposes, and

(c) Also means any act of sexual contact between persons involving the sex organs of one person and the mouth or anus of another whether such persons are of the same or opposite sex.

Admittedly, I can think of sexual acts that aren't covered by the above, but I can also appreciate the difficulty in writing a law that covers every conceivable sex act. And other sections of RCW proscribe assaults, sexual or otherwise, which do not satisfy the definition.

Basically, in the State of Washington, the law entitles you to be physically secure against any violation of your body. Modern rape law is relatively comprehensive and well-written, due in no small part to the actions of the feminists who helped craft it. It is slander against both feminism and the legal profession to give an impression as misleading as the one MacKinnon gives.

Are the laws perfect? No. Do they "guarantee impunity for most rapes" and constitute "a violation of the law of sex equality"? Pffft.

Anyway, this post has gone on too long. And MacKinnon has still failed to articulate a single way that rapists' views are reflected in rape law. I don't really have time to get into MacKinnon's deeply confused and factually mistaken discussion of pornography. So, I'll conclude by considering the wince-inducing attempt at emotional manipulation that Jeffries documents at the end of his article:

MacKinnon's book ends with a wonderful rhetorical essay called Women's September 11. It points out that roughly the same number of women are murdered by men in the US each year as were killed in the Twin Towers (between 2,800 and 3,000). But those killings provoked no parallel war on terror.

So what does MacKinnon think should be done? She writes that violence against women "qualifies as a casus belli and a form of terrorism every bit as much as the events of September 11 do". Is she serious that violence against women should be treated as a war? "I think only because it's men doing it against women that it isn't seen as a war." I feel another twinge of vertigo.

I've only read a tiny bit of the scholarship on terrorism, so I'm no expert. But I'm pretty sure no serious definition of terrorism would include random murders, committed in civil society, by a massively dispersed group of individuals (i.e., men), without any unifying organization or motive, against an incredibly broad variety of unrelated targets (i.e., women), without any self-conscious intention to manipulate the media or body politic.

So why make the comparison? MacKinnon has a bad case of Rhetorical 9/11 Parasitism.

Rhetorical 9/11 Parasitism is what I call the rhetorical use of the September 11 attacks, and their explosive emotional impact, to lend weight to some unrelated issue. Or, in other words, If I Don't Get My Medium-Rare Shell Steak With Roasted Vegetables In The Next 10 Minutes, The Terrorists Have Already Won. The grandest extant case of 9/11 Parasitism was, of course, the rhetorical circus held by the Bush administration before the invasion of Iraq. MacKinnon's version is rather paltry by contrast: rather than a literal war, she merely wants to undertake a metaphorical War on Violence Against Women.

Of course, when one invokes the metaphor of war, one invokes not only the expectation of all-out effort, but a whole constellation of associated concepts. To pick just a few:

  • Wars justify the suspension of ordinary legal norms and even basic, primal human taboos (like "thou shalt not kill").
  • Wars are organized by central planners against external enemies, not by individuals against the members of their own neighborhoods or households.
  • Wars end.

In short, the idea of war suggests tactical and strategic expectations that --- need I really say this? --- simply aren't appropriate to solving most problems in civil society. Hence the clichéd observation that most metaphorical wars don't work very well, as metaphors or as wars: the War on Poverty, the War on Drugs, the War on Terror. Like poverty, drug use, and terror, violence against women isn't caused by an organized enemy that can be defeated with force, and thereby induced to sign a treaty of surrender. So why would anyone believe that recasting the problem of violence against women as a war would lead to progress?

More to the point, why does Catharine MacKinnon believe this?

Well, OK, that's a rhetorical question, and you know what I'm going to say. Catharine MacKinnon believes these things because she's an idiot.

[0] More precisely, it assumes the (bogus) rule that if A and B are subsets of C, then any property of B must necessarily be shared by A.

[1] Well, reality is probabilistic, so, more precisely, one must isolate a view that is relatively correlated with being a rapist and anti-correlated with being a non-rapist.

[2] Seriously, what's the deal here? Is she worried that necrophilia's some kind of "gateway drug" to rape? Is there some hidden national epidemic of necrophilia that I haven't heard about? Worrying about necrophiliacs when you talk about rape is like worrying about lesbian Nazis when you talk about anti-Semitism. Lesbian Nazis probably exist, and they're probably anti-Semites, but they're not exactly the heart of the anti-Semitism problem, and you would have to have some pretty weird issues even to bring it up.

[3] I don't believe in free will, so in some abstract philosophical sense I would agree that choice doesn't exist. However, as a practical matter, the social fiction of free choice undergirds most of the law, including the laws governing contracts, property, and the democratic process itself. By default, the law assumes that the decisions of mentally competent adults have meaning, and that actions affirming these decisions have a binding effect. Hence, when women vote, sign contracts, buy property, choose careers, or grant interviews to newspapers, the law treats their decisions as meaningful ones.

Now, the outward signs of consent can themselves be coerced --- e.g., someone can put a gun to your head and make you sign a contract. So the law carves out specific, narrow exceptions for "false consent": when you're under physical duress, when you're mentally incapacitated, etc., the law ignores evidence of consent. But it is absolutely crucial to the rule of law that these exceptions be narrowly targeted, and based on objectively observable criteria, rather than applying categorically to a vast and vaguely specified set of situations or people.

I suspect that MacKinnon believes that women can only freely consent to sex with men in the absence of sexual power inequalities, and therefore the law should disregard evidence of their consent. This is a catastrophically bad idea. Every society since the dawn of humanity has suffered from power inequalities, and virtually every human relationship involves some element of inequality. Yet society has generally found it useful to treat adults' decisions as meaningful anyway.

Recall what Jeffries writes of MacKinnon:

"The assumption," she says, "is that women can be unequal to men economically, socially, culturally, politically, and in religion, but the moment they have sexual interactions, they are free and equal. That's the assumption . . ."

MacKinnon's utterly mistaken. The law assumes no such thing. No law governing consensual transactions assumes that the two participants in a transaction must be on absolutely equal footing. It would be nice if most people were more pairwise equal than they are, but the rule of law cannot wait for utopia. To be useful, the law must govern, and give force to people's decisions, in the imperfect societies that exist now.

And, in fact, most groups have generally benefited when society treats their decisions as meaningful. People usually fight for the freedom to vote, to enter into contracts, to buy property, and --- yes --- even the right to consent to sex. Being over eighteen, I certainly appreciate the fact that I can have sex without fearing that my partner will be jailed for statutory rape. I suspect that most adult women also enjoy having that right.

[4] What about non-physical forms of coercion? Again, read the rest of 9A.44 RCW and you'll find that people in "supervisory positions" over children, the developmentally disabled, or other people dependent on assistance, cannot have sex with their charges. You'll also find that 49.60 RCW defines unfair practices for employers, which the courts have held to establish civil remedies for sexual harassment --- which, OK, does not classify the behavior as rape, but I have major problems with MacKinnon's propensity for calling everything rape, from corpse desecration to pornography to sexual harassment. Sometimes there's value in preserving distinctions among the meanings of different words.

Sunday, April 09, 2006

Sharing good deeds

AJ points to Dogooder.info, a site for sharing the good deeds you have done, or have had done unto you. This site could use more traffic, methinks. I can imagine something like this becoming a great resource for birthday/anniversary/etc. ideas.

Friday, April 07, 2006

Economists empirically analyze online dating (reactions to Hitsch et al. 2005)

So, Günter J. Hitsch, Ali Hortaçsu, and Dan Ariely have a paper that empirically analyzes correlations between posted characteristics and success rates in online dating. A lot of you have probably seen this already, because UC Berkeley economist Hal Varian had a little writeup in the Times about a year ago, but I ran across it recently somehow (blog post, perhaps? Lost in the foggy mists of my browser's history...).

Anyway, as Varian suggests, it's mostly an exercise in confirming common sense, but not entirely. Of course, it is actually valuable to have your common sense examined empirically once in a while, so I heartily encourage you to skim the paper's results yourself, but I'm going to pull out a few results that aren't entirely obvious from (my) common sense.

First, here's Hitsch et al.'s chart correlating body mass index (BMI) for men (left) and women (right) with response rate.

It's no surprise that women benefit more from being thinner, but I think it's really interesting that the response rate peaks at such an extremely low BMI. The authors report that the peak BMI is about 17, which corresponds to an underweight, "supermodel" figure. It is, of course, commonplace for feminists to remark that American society has unrealistically thin body expectations for women. A lot of men believe that this is an expectation perpetuated mostly by women's fashion magazines. Whenever you bring up the body image issue, online or off-, some man will say something like: "Well, I don't prefer stick-thin women. Slim, yes, but not bony like a fashion model." I mean, I'm a man who believes something like that. Yet I look at the chart above and I wonder.

Now, the caveats. First, it's possible that the results arise from men's misunderstanding how a numeric weight translates to an actual feminine figure. After all, the men are only reading a number, not looking at bodies (I'm assuming the profile photo's usually a head shot, rather than a full-length bikini beach photo). For example, I weigh about 155 lbs, and I assume that a slender woman my height would weigh somewhat less, but I'm not really sure whether 135 lbs would be emaciated or merely thin. Second, the economists find that women are probably lowballing their self-reported weight: "Among women, we find that the average stated weight is less than the average weight in the U.S. population. The discrepancy is about 6 lbs among 20-29 year olds, 18 lbs among 30-39 year olds, and 20 lbs among 40-49 year olds." --- so either the personals users are exceptionally fit, or they're lying. Men might know this, and compensate by mentally adding a few pounds to whatever the profile reads, hence their preference for exceptionally low (reported) BMI.

However, in spite of these caveats, I think the evidence suggests that the ideal towards which so many women strive is, to some extent, a rational judgment about men's aggregate weight preferences. It's true that people can be attracted to potential partners who don't attain that ideal. But that doesn't mean that more people wouldn't be more attracted, on average, to someone who's closer to that ideal.

OK, enough with the weight. Let's look at the ethnicity findings. In my opinion, these are the most interesting findings in the paper. Money, education, looks, weight, age? Everybody basically agrees and accepts that people discriminate based on these factors; they reflect differences in character, in lifestyle, or in plain sex appeal. But race and ethnicity are stickier issues. It's not clear whether it's really acceptable in polite society to admit that, given a choice between a good-looking white man and an equally good-looking black man, with the exact same job and income, you would really prefer to date the white man. Is that a form of racism, or not? Is that the same as men preferring blondes (an actual finding that, incidentally, also emerges in the study), or is it something different?

Anyway, here's Hitsch et al. 5-10:

This figure requires a bit of explanation. The charts on the left indicate who women emailed, and the charts on the right indicate who men emailed. These are so-called "first contact" emails, meaning that these are the preferences of initiators, not responders. (This is a pretty big source of asymmetry, since men are usually initiators. In fact, 54.5% of men received zero first contacts. For men, it would be more interesting to track "first responses" to emails sent.) The labels on the vertical axis indicate the ethnicity of the person emailed. For each of the four ethnicities --- white, black, Hispanic, and Asian --- the rate at which the emailer emails his/her own ethnicity is considered the baseline rate. The labels on the horizontal axis describe how much less (or more) probable each ethnicity is to receive a first contact email, compared to the baseline. (I assume the pink bars are confidence intervals.)

So, what's the result? All men and women of all ethnicities prefer their own ethnicity. That's straightforward enough and can be explained by basic homophily or whatever, although the strength of the preference is pretty startling to me: most groups have at least a roughly 1/4 disadvantage relative to baseline, and often more. I guess I'm naïve, but I would have hoped for closer to 5% than to 50%.

More interesting than the strength of the preference, however, are the differences among the groups. Some of the confidence intervals run pretty wide, so some individual charts have to be taken with a big grain of salt, but you can still see broad trends. For example, women across the board have much stronger ethnic preferences (or biases, if you like) than men. Also, Asian men and women alike seem to prefer whites significantly to any other non-Asian ethnic group, but whites do not return the preference, ranking Asians last or next-to-last. In fact, Asian men do pretty badly overall, coming in no better than a 75% diminished preference by any other ethnic group (and apparently getting zero emails from Hispanic women). Finally, based on my completely unscientific eyeballing of the charts, for whites and Asians, men and women seem to have preference profiles with similar "shapes" --- that is, the women's ethnic preferences are stronger, but their relative preferences are similar to men's --- but for blacks and Hispanics the men and women seem to diverge.

But even more interesting than any of that is Fig. 5-11:

This is the difference between (white) people who admit they have ethnic preferences, and those who claim that they don't. Note my verbs, because although there's a statistical difference, the curves have the same similar shapes (UPDATE: see comments), and the ethnic preference remains strong in the latter group --- greater than 50% for women, and in the ballpark of 15-50% for men.

I think that chart almost speaks for itself, so I'll keep my remarks brief. Is this caused by unconscious bias? Perhaps. Or maybe it's like lying about your weight. I know that if I saw that a woman's profile had an ethnic preference --- any preference, even for my own ethnicity --- I probably wouldn't want to email her. By not listing a preference, one signals one's open-mindedness.

Finally, to end on a lighter note, here's Fig. 5-7, the result that's probably most relevant to the people who read this blog --- educational attainment and response rates:

Look at the highest data point in first contacts for men. That's right guys, chicks dig grad students.

Well, OK, back to reality. Clearly, the "in post-graduate program" results for men are a fluke arising from the fact that the category lumps together professional students with Ph.D. grad students. Law, medicine, and business students are training for lucrative careers as highly esteemed members of society. Ph.D. grad students are training to be credentialed professional nerds. As much as I wish that most women found the latter attractive, I have to admit that it is, shall we say, a distinctly minority taste.

Monday, April 03, 2006

John McCain pandering

Crooks and Liars points to Tim Russert's evisceration of John McCain's pandering ass. It's so satisfying to see this "straight shooter" try to spin himself out of the holes he's dug for himself through his words and actions.

Now all I want is for the New York Times or WaPo to point out, in a "news analysis" on McCain's Presidential ambitions, that a Senator hasn't won a Presidential election in a long time. Funny how this factoid gets whipped out within nanoseconds whenever the subject's a Democratic Senator running for President, but never for McCain. Liberal Media strikes again.