Saturday, August 28, 2004

An Exegesis of Feminist Research on Women in Information Technology
The important computer science organization ACM has a thrice-weekly posting called TechNews of news items about the world of computing and technology. When the posters are not shilling for Linux or actually reporting on something interesting, hardly a third-of-a-week goes by where they fail to discuss what every nonthinking IT professional knows is a crisis of monumental proportions: the small number of Women In IT.

I have known many feminist computer scientists for a long time, and they have always known the cause of this problem. In the seventies and early eighties they knew that the problem lay with the way that parents and grade school teachers treat very young girls and boys: these parents and teachers (perhaps unintentionally) steer girls away from an interest in computer science. Studies were done, and this explanation was proven correct. Of course, these feminists were young adults with no children of their own. Later they had children and their friends had children and they saw differences between girls and boys, so they no longer offered this up as the explanation; in fact, they had never offered this up as the explanation, so no apologies were necessary.

New explanations were needed. This is a problem that has to be explained and fixed because ...

Well for one thing, we need a huge number of additional IT employees (or at least we did before the "bubble" burst), and where else can we find them but amongst women? I suppose we could look for more men, since most men are not IT employees, but presumably the men who aren't in IT have a good reason for not being there: they aren't good at it, or they aren't interested in it, or they are otherwise gainfully employed. Women who aren't in tech, on the other hand, are leading useless lives and need to be saved. For another thing, women have a lot to offer IT that men just don't, and presumably these special skills aren't as important in that other stuff (law, medicine, motherhood, etc.) we want to take them away from.

Here, chronologically from almost five years of the ACM TechNews archives, are some explanations that have been offered for why there are so few women in IT and for why we need them so badly, as well as discussions of how to solve the problem. I've only chosen a small number of articles that I find especially amusing. ACM is obsessed with this issue (in a totally one-sided way), and this obsession reflects the incredibly totalitarian atmosphere in the IT industry, and especially in university computer science departments. (Remark: My host, Dan Simon, objects to this slur upon the IT industry.)

I've put (links to) these explanations on a separate page. If you read them you'll learn that sexism is to blame for the large number of female entrepreneurs; or that the real problem is that women don't like introverted white male geeks; that video games are the problem; that men do social networking better than women; that IT is too time consuming; that women are differently abled; that the rise of dot-coms hurt women; that the fall of dot-coms hurt women; and much, much, more. Each of these articles is idiotic and most are self-contradictory, but the sum of them together totally destroys any illusion that the ACM takes this subject seriously.

Here is a challenge for my reader(s): Find a feminist oriented article about women and IT from the last four years that was deemed too stupid to make it into ACM TechNews.

Here are the links.

Wednesday, August 25, 2004

It should be obvious by now that the battle between John Kerry and the organization known as "Swift Boat Veterans for Truth" (SBVT) has nothing to do with what, exactly, John Kerry did or did not do in Vietnam. The actual accusations about his service record are for the most part exceedingly minor--at worst, that Kerry wasn't quite as heroic as he has claimed, and may have exaggerated his bravery a touch. The sole accusation with any potential substance is that Kerry later claimed to have been in Cambodia, when in fact it's doubtful that he ever was.

But the reason for this accusation's significance is precisely that it calls into question Kerry's political record, not his military one. For while Kerry may have used his claimed military heroism to burnish his character credentials, he later used his claimed runs into Cambodia to buttress his political positions as an antiwar activist and politician. Similarly, most of the SBVT animus towards Kerry, judging by their public pronouncements, is a product of his antiwar activism, not the petty details of the episodes that won him his medals.

However, SBVT dare not make Kerry's antiwar history the issue itself, for the simple reason that their staunch belief in the justice and honor of that war is far from unanimously shared among voters. Were they to assail Kerry as a hippie pinko radical traitor, they would run up against a great deal of opposition from people who considered Kerry's antiwar stance a principled and justifiable one. Indeed, there's no guarantee that such a debate wouldn't earn Kerry more support than it loses him. So instead, SBVT call Kerry a liar--and who can argue in defense of lying?

Of course, SBVT hardly invented this tactic. Anti-Bush partisans prefer to call him a liar, a fool or a corrupt dictator rather than discuss the nuances of Iraq policy--on which Kerry has a hard time defensibly differentiating himself from Bush. Similarly, former president Clinton's haters accused him of corruption, dishonesty, immorality, even murder--anything but bad policy choices, since his policy decisions were for the most part highly popular. These days, only unsuccessful politicians get the honor of being attacked for their unpopular political positions. The rest are subjected instead to withering personal attacks, calculated to rally the faithful and perhaps win a few stray fence-sitters, while alienating nobody but the hard-core partisans who are already invested in the attacked candidate's manifold virtues.

The success of such attacks, however, is far from certain, and ultimately depends on the strength of the attacked candidate's political support. Bill Clinton, for example, was pummelled by a series of far more devastating shots--alleged draft evasion, Gennifer Flowers, and others--yet survived to beat an incumbent president basking in the triumph of a spectacularly successful war. His secret: a detailed, brilliantly crafted policy program that allowed him to dismiss personal attacks as insubstantial compared to the "real issues".

Kerry, on the other hand, has been light on policy proposals, choosing instead to run largely on the theme, "I'm a war hero"--and, more generally, "I am not George W. Bush". And because the politician who lives by "character" dies by character, the SBVT attacks have exposed a clear vulnerability in Kerry's campaign strategy. If I were Kerry, I'd be instructing my advisors to come up with some surefire, voter-pleasing new campaign promises--and quick.

Saturday, August 21, 2004

We need to have a sense of substantivity and proportion about "privacy".
Substantivity means that our discussions about loss of privacy should actually be about loss of privacy, and proportion means that stronger concerns should actually reflect greater loss of privacy.

Although I've been interested in privacy issues for some time, the immediate spark for these remarks is a discussion by Eugene Volokh about a decision of judge Kozinski objecting to (and presumably asserting the illegality of) the collection of DNA from people who are on probation after having been convicted of a crime. Kozinski's argument is that this will lead, via a slippery slope, to the collection of DNA from many more people, and perhaps from all people. He points out that a similar expansion has already happened with the collection of (searchable) fingerprints, and I think he is right that there is the "danger" this will happen with DNA. Volokh also feels that the slippery slope argument is well made (oddly, he doesn't explain why such an argument is appropriate to a judicial -- as opposed to a legislative -- decision), although he chooses not to discuss whether or not the "bottom of the slope" is truly a bad thing. Kozinski deplores the bottom of this slope, saying that we will all come to accept it because "the fishbowl will look like home".

What exactly is so bad about the government having all our DNA fingerprints on record? Unless I've missed something, Kozinski (see page 11530 here) thinks this is not much worse than the government having all our finger fingerprints on record, and he laments about how close we've come to that. He recognizes the huge help these databases provide in solving crimes, and is sad that the public doesn't realize ... what exactly? What is the substantive harm of having our fingerprints on record? You'd think he would give some examples. Perhaps the government goes to legal brothels and collects fingerprints to match with political opponents? Would this be the best way to get dirt on people? At least some people claim that DNA fingerprints are special, in that they could, conceivably, wind up containing health information about people, that can then be used by the government to ...

The government collects a huge amount of financial information on all of us; both the invasiveness of the collection of this data and the potential for its misuse dwarf by orders of magnitude anything we have to fear from the collection of (DNA) fingerprints. Similarly, the substantive invasion of privacy and potential for abuse from searching my home or person dwarfs that of swabbing my cheek. Why should we be so sure that those "framers" of the constitution would think otherwise if they were brought technologically up to date?

People generally have nonsubstantive and/or disproportional attitudes towards "privacy". They are horrified at the possibility that Microsoft might collect some information about their computer hardware while doing an "update". But they have no concern whatsoever that the phone company keeps records of every cell phone call they make or receive, including who they call, when, and for how long. (I know these records exist because I receive a copy of them every month with my bill.) And Mastercard knows about all my travels and purchases and the movie rental store knows about my kinky tastes in film, but these are not typically our concerns.

One of the big scandals of the modern day is that our medical records are not computerized and networked so that any doctor with a legitimate need can have access to them. In addition to giving timely and accurate access to the most important aspects of a person's medical history, this would allow automatic checking of many things including adverse drug interactions. The one and only reason given for not doing this is "privacy concerns". These concerns are substantive, but greatly disproportional to the enormous loss of life that happens every day due to the nonexistence of this system. (I hope to write more sometime about the pros and cons of such a medical system.)

I am not denying that we have much to fear from inappropriate behaviors of our government. Just to give one example, I am horrified that merely because I am suspected of using my car to deal drugs, the government can take that car away from me. But we should deal with substantive abuses in a direct way, and not by arbitrarily denying the government access to the data it should have. If by consenting to have my cheek swabbed I can greatly reduce the probability of my house being broken into and my family assaulted, this is not an example of living in a fishbowl, but rather an example of my privacy being greatly improved in every substantive way. If Mr. Kozinski lives in an apartment building, we can be reasonably certain that he insists on having a doorman who monitors everyone who comes and goes and who learns a great deal about the occupants of the building. He can afford this, and no doubt he also has access to whatever (very intrusive) bodyguards he needs to follow him around and protect him. I would guess that when it comes to his own life, he chooses substance in privacy over whatever he imagines the framers would want for him.

Friday, August 06, 2004

The New Republic apparently believes that economic redistributionism is due for a big comeback. This past week, it published two articles on successive days advocating the return of taxing the rich to pay the poor (or at least, the not-rich). Now, I'm anything but a fanatical libertarian, and it's entirely possible, for all I know, that current levels of government-mediated wealth redistribution in America are less than optimal for maximizing collective happiness. But these two articles are so nonsensical in their reasoning that they cast more doubt than light on their common position.

The first, by economics professor Barry Schwartz, argues as follows: studies have shown that people who are overwhelmed by an excessive number of choices are less happy than those with a more moderate range of options. Wealthy people, because of their wealth, have inordinately many options in life, and are therefore most likely overwhelmed by them. Hence, taxing them and giving the proceeds to the poor (who have far fewer options) will tend to make both groups much happier.

It's, as they say, a creative argument, to be sure. And no doubt some rich people would actually be happier with simpler, more spartan existences. But such people have an uncanny ability to find means--drug addiction, religious discipline, self-impoverishment through profligacy--by which to reduce their choices in life. It's hard to imagine that the failure of the rest to find such means is due to a lack of resourcefulness, rather than a lack of inclination.

The second article on the subject is somewhat less obviously ridiculous than the first. In it, political science professor Jacob Hacker argues that public opinion about the economy is pessimistic these days--despite the economy's fairly robust performance--because family income has become more and more unstable over the years. Hacker's solution: massive government-mediated income redistribution in the form of "insurance" against income instability, in the spirit of the Social Security program. By increasing the safety of the average family's income, Hacker argues, such programs would increase public confidence and encourage economically productive risk, in the same way other forms of insurance make otherwise excessively risky activities feasible.

Now, I'm all in favor of risk reduction through insurance, and I'm not at all averse to government-provided insurance, in those cases where it's valuable for it to be universal. Retirement and medical care are two such cases, in my opinion, and while I acknowledge the problems that have beset government-run health care and retirement insurance plans lately, I'm far from convinced that these problems are inherently insoluble.

But Prof. Hacker's argument suffers from two glaring flaws. First of all, he rests his case on the premise that today's pessimism about the economy is a product of rising income instability. But his own figures indicate that income instability in America rose massively from the 1970's to the 1990's--the latter period being one of widespread economic optimism such as the nation hadn't seen for decades. Apparently, gloomy predictions are less well correlated with widespread income instability than he implies.

Secondly, Prof. Hacker writes as though economic insecurity is crowding out risk-taking on the part of anxious Americans. Yet American household debt--as pure an indicator of risk-taking as one can imagine--is at an all-time high today. (Indeed, this kind of risk-taking is likely more potent than income instability at generating feelings of economic insecurity.) If Hacker's goal is to correct the level of risk-taking among households, then he'd do better to recommend ways of increasing the costs of risk to today's credit card-bingeing, zero-down mortgage-signing American households.

There was a time when appeals for income redistribution were straightforward: one would cite the plight of some miserably disadvantaged group, and declare it a shame that they should suffer so terribly in a society where the privileged enjoyed such wealth. Apparently, it's much harder than it used to be to find such people, because today's advocates of redistributionism have to resort instead to absurdly specious arguments in making their case. We are, indeed, living in fortunate times.