New Stanford Elections Scandal

The Stanford Daily reports that the Black Student Union donated nearly $200 to one of the slates in the just-past ASSU election, and the NAACP also planned to reimburse them for costs, though it did not get around to it. All of this because candidate Ajani Husbands, who is black, "is closely tied" to the BSU and NAACP.

Husbands confirmed that his campaign accepted money from the BSU, but continued that this was not the first time a student group had given special-fees funds to a candidate. In the past, he said, candidates endorsed by the Students of Color Coalition “have gone to groups for fundraising.”

“This is not out of the blue,” added Husbands, who is a member and former program coordinator with the BSU.

Schwartz agreed, stating that she was led to believe that this practice was not new.

“I didn’t want to take money from groups since it’s student money and they shouldn’t be paying for ASSU campaigns, but we were under the impression that SOCC-VSOs had been paying for campaigns for years,” she said.

[...]

According to Schwartz, legal considerations did not factor into the discussion.

“No one seemed concerned about the ethics of it,” she said.

What's especially amusing is Mikey Lee's reaction. (For those who haven't followed the elections, Mikey Lee and Dylan Mefford won the election, but their election was not certified by the Senate due to "abuse of power" when Mikey Lee sent out e-mails after the election had already begun. Instead, a special election was called and the winners were certified as the new ASSU executive slate.)

Lee said he found hypocrisy in the Senate’s decision not to certify the original election results in light of the BSU’s donation, given Husbands’ connections with the BSU and Stringer’s involvement in recent ASSU election issues.

“It’s ironic that the grounds upon which our election was not certified was ‘abuse of power’ when that pales in comparison to this corruption, which is a clear abuse of power and students’ money,” he said.

In addition to authorizing the BSU’s reimbursement to Schwartz, Stringer was the Senate’s representative at a recent Constitutional Council meeting, at which he argued that senior Mikey Lee’s use of the Class of 2004 e-mail list for campaigning represented an abuse of power.

Senator Omar Shakir, a freshman, expressed concern when informed of the donation.

“This news troubles me, especially in light of the standard that the Senate set that an abuse of power cannot be tolerated in an election,” he said. “Special fees . . . can never, without exception, be used to do anything other than directly benefit the student body. While I hold Dan, Alyssa, Ajani, the NAACP and the BSU in the highest regards, it is my duty to the student body, based on the seriousness of these allegations, to ask for a full and complete investigation.”

Lee also objected to the Husband / Schwartz slate’s use of special-fees money.

“This is egregiously unfair to students who support these groups’ special fees,” he said. “I pay my special fee to support both BSU and NAACP, and to think that my money went towards supporting a slate that I was running against is mind-blowing.”

UPDATE: Read the Daily's editorial:

It is worrisome enough that the ASSU failed to catch or reprimand the Husbands / Schwartz executive slate for improperly accepting money from at least one student special-fees group. But far more disturbing is the fact that some of the organization’s leaders knew about or even participated in the Husbands / Schwartz slate’s improprieties — and yet staked out a moral high ground in censuring and punishing opposing candidates Mikey Lee and Dylan Mefford for less egregious violations.

The Restaurant Choosing Problem

Tim Groseclose and Jeff Milyo published a paper on the Fair Restaurant Choosing problem, in which a group of eaters have to jointly choose a restaurant in a way such that they are all equally likely to end up at their favorite restaurant:

Suppose a group must choose a restaurant. We propose a method where, similar to the I-choose-you-cut rule for dividing a cake, individuals in the group take turns restricting the set of choices for the group. Specifically, under our method the first person restricts the set of restaurants to a certain number; the second person restricts the set to a smaller number; and so on until the last person in the group selects one restaurant. We derive a formula for choosing these numbers such that—under a natural assumption about individual preferences—the probability that the group will choose any individual’s favorite restaurant is equal for each individual. For the case where there are only two people in the group and there are n restaurants, under our method the first person selects the square root of n restaurants. The second person then chooses one restaurant from this set. When there are k individuals, our method requires the first person to select n(k-1)/k restaurants. From this set the second person selects n(k-2)/k restaurants, and so on until the final person selects one restaurant.

(Link via Marginal Revolution)

Of course, a problem of much more practical importance is the Feasible Restaurant Choosing problem, in which a group of n people jointly decide to choose a restaurant they are all willing to eat at. The constraints are of the form "I had Persian food yesterday", "I don't like pizza", "Spicy food gives me headaches", and "I'm a vegetarian and the only vegetarian food they serve there is eggplant, but I'd be willing to go to the Cheesecake Factory instead because they have good chicken caesar salad that I would make an exception for and I love their desserts."

Empirically, I've found the Feasible Restaurant Choosing problem to be usually impossible for n > 3, though strangely the probability of success increases with size whenever n is at least eight.

Sex Discrimination Is Not The Same As Racial Discrimination

Today in New Jersey, it's illegal to offer Ladies' Nights in a bar: special discounts for women are discrimination against men. Similar rulings hold in Pennsylvania and Iowa.

Taken literally, it's clearly true: higher prices for men than for women are discriminatory. One might equally add that larger bathrooms for women are discrimination against men. For that matter, different-sex bathrooms are "separate but equal" treatment and so discriminatory, especially when one considers that men's bathrooms can be often much dirtier than women's - or, alternatively, that the wait for women's bathrooms can be more than twice as long.

On the other hand, all these examples clearly make everyone better off. After all, men are more inclined to go to bars when it's ladies' night than not. Generally, most men and women (with the possible exception of West Coast college students) prefer segregated bathrooms to multi-person unisex bathrooms.

Racial discrimination is bad as a general principle because there aren't important systematic differences between the races that would justify differing treatment. On the other hand, sexual discrimination, which is often barred by the same laws that bar racial discrimination, is often justified when it is done to take into account the real and important differences between the sexes. Obviously, such as when it was done paternalistically, this can be taken too far. But it's unreasonable to bar all sexual discrimination the way racial discrimination is barred when so much of what remains is useful and valuable. That's a lesson that's recognized by most states - just not New Jersey.

Volokh Site Snatched

The Volokh Conspiracy's old site, volokh.blogspot.com, has been taken over by a search engine optimization company (read link-spammer). It makes sense, of course: Volokh's site has PageRank 6 because of all the stale links to it, and that will probably make the search engine optimization company's site have very high PageRank as well.

Taking over old blogs at BlogSpot - just another way the link spammers manipulate Google to increase their own PageRank.

Oh, by the way - if you have a stale link to volokh.blogspot.com, please update your links to the new site at volokh.com.

Musicians on Filesharing

The Pew Internet Project has a new study about how musicians of various sorts feel about filesharing. Normally we hear only from record executives, famous multimillion-selling musicians, and listeners. But their study primarily includes musicians who make a fairly small amount of their money from music, while having to hold down other jobs as well (as one would imagine is the case for a majority of musicians, however defined).

An overwhelming majority of them think of filesharing as having at least some benefits for their musical careers. Even among the small percentage that earn 60% or more of their income from music, more think that the RIAA efforts to curb filesharing will do more to hurt them than help them.

And I think they're actually right about filesharing helping them. New technology that improves music distribution has always damaged the music industry but then dramatically increased the number of people that can actually make a living off of music, and increased the amount of music available to consume. For instance, the entire phenomenon of indie rock requires some sort of flexible scheme of music distribution, like records and CDs and such. Without these recordings, I think it would be extremely likely that bands like Mogwai, and Belle and Sebastian, would probably still be in Scotland playing shows to a few hundred kids, instead of tens of thousands of Americans in between dozens of other successful shows. Never mind how they would have been able to come up with their own music without being exposed to the music of anyone who wasn't able to physically come to Scotland to play.

With the rise of mp3's, there is now a significant percentage of college students (and probably other populations) that can actively comment not just on their few favorite bands, or even two or three disparate genres, but actually dozens of different genres, including several unrelated groups of genres. All in all, it seems that even unfettered filesharing of mp3's and other music files can only in the end help the music-listening public, and even the majority of musicians, even if it decreases the chances of becoming a multimillionaire through one's music.

Polls that Matter

Zogby has started a "battleground states" poll. Right now it looks like Kerry is headed for a pretty good win, and is even above the margin of error in states like Michigan, Pennsylvania, and even New Hampshire. Of course, since I assume these are 95% confidence intervals, and there are 16 of them, there's probably about a 2/3 chance that one of them is actually outside the margin of error. That fact, together with the large number of voters necessary for a survey to have reasonably small margins of error in each state is probably why no one else has been doing this sort of thing so far. I think they should have included Arizona and Maine though, given how much both those states are being targeted. (Colorado, Louisiana, Delaware, and New Jersey are probably too much of a stretch though.)

Women in Blogs

Matthew Yglesias recently posted about the lack of women who read political blogs. He says, "At first glance, one might think of this as an internet issue, related to hardy perennials regarding women and technology in general, but I think it's a manifestation of the broader fact that women don't seem very interested in politics." This argument is even stronger when you realize that in fact the internet is majority female. (By late 2000, 50.6% of internet users were femals according to the Pew Internet Project. Back in '97 it was only 42%, and it looks like it was even lower a few years earlier. So if this trend can really be projected, it seems likely that the internet is now definitively female-majority.) In fact, it seems to be even more so, when you remember that various newspaper articles that have discussed the growing participation of women in the internet attribute it to the growth of both online shopping and blogs. (Though they probably mean more the livejournal kind of blog than the political discussion blog.)

However, even this "fact" that "women don't seem interested in politics" is not entirely true. As one commenter points out, most political MeetUps and other activities he's been to are about 50/50, and I can confirm that this seemed approximately true of the Dean and Kerry MeetUps I went to. (Though I didn't count too closely, it's clear that they were much closer to 50/50 than political blog commenters seem to be.)

One commenter says, "I think many women are a little turned off by politics because it tends to be a testosterone-laced blood sport. Like most women, I think, I abhor the verbal pissing contests that seem to delight many men." This is exactly the sort of reason that has been given for the dearth of women in philosophy. We tend to think of it as a bit of a travesty that the senate is only 14% female, but this is no different from the Formal Epistemology Workshop I attended a week ago, where there were never more than about five or six women present, though the audience had about 35 people. And in workshop audiences, we don't have to worry about biased voters keeping the female population artificially low. I don't know if brutality and competitiveness is the reason why women tend to avoid politics and philosophy (in particular, I'm not sure how brutal and competitive politics and philosophy are compared to other academic disciplines with a more even gender balance, or how competitiveness differs across branches of philosophy with varying gender balances) but it has a sort of surface plausibility.

Matthew Yglesias' explanation of political blog commenting as a sort of surrogate for sports fandom is an interesting new explanation of the gender disparity, though it doesn't address the senate or philosophy.

(For comparison at least 26 of 187 math graduate students at Berkeley are female, and 15 of 48 philosophy grad students are female. So I suppose math is a bit more male-biased than philosophy. But not as much as particular disciplines in philosophy. The count of female math grad students is a lower bound because there are a lot of Asian names whose gender I couldn't place. And also several names like "Aubrey", "Tracy", and "Eli" that in many cases turned out to be men, but I wasn't sure in all cases.)

EDIT: Here's some statistics on blog readerships, including gender. Apparently two prominent political blogs have 15-20% female readership, which puts them in the same league as the Senate, and slightly above the Berkeley math department and below the philosophy department.

And choosing the University of Michigan as another school with fairly large (for statistical significance) and highly-ranked (for relevance to the field as a profession) departments in both math and philosophy, I find that 12 of 38 in philosophy are female and at least 25 of 127 in mathematics are female.

More on Deontology

Bob responds to my comments attacking deontology, saying "If everyone prefers action A to action B [sic], then it would clearly be treating everyone as an end to take action B - you are helping everyone achieve their preferences." I think this isn't exactly true. For instance, sometimes there are times when the outcome of telling the truth to someone or some group ends up making them worse off than by letting them remain ignorant of something.

Say, there's a happily married couple with two children, one of whom is secretly a child the wife had from an affair that she got over a long time ago and regrets seriously enough that she'd never consider infidelity again. Telling the child and the husband the truth here would just lead to resentment on all sides and won't make things better for anyone. (Other cases arise constantly in fiction - and someone always goes and tells everything to people and messes stuff up.)

On Kant's view (in my understanding) it's never right to lie to someone, even to protect them from some awful truth, because such a treatment is paternalistic and ignores their dignity as a rational agent that can decide what information it should receive. Thus, for Kant, the only right thing to do is to tell this horrible secret that just makes things worse for everyone. Thus, a happy family's life will be ruined just because someone's dignity requires the truth. It seems to me that this dignity is on a par with the Cartesian mind, which is independent from the body and doesn't causally interact with anything, and yet is somehow essential to the person. And thus should be ignored, because of its extreme causal inefficacy and implausibility, in a certain sense.

Of course, this sort of intuitive example I'd like to present only has its full force when things are worse off for some people and better off for no one as a result of doing "the right thing". Surely, it seems to me that any reasonable ethical system should avoid such cases. Thus, it's grounds at least for revising Kant's positions, if not abandoning them. But there are other, less completely compelling, cases in which some people are much worse off and a few people are slightly better off as a result of doing "the right thing" on some deontological grounds. If one is willing to take all such cases into account, then even if one sticks with a deontological system, there is a sense in which it is equivalent to some utilitarian system, in that it sanctions exactly the same actions. It seems to me that the best defense against this argument would be to suggest (rightly) that there's no realistic way to combine utility functions for different people into one aggregate function (seems reminiscent of Arrow's Impossibility Theorem now, somehow) and thus that this argument only has force when doing "the right thing" leaves some people worse off and no one better off. Which wouldn't be as difficult a task for a deontological system to meet, but still hard enough.

Note that none of these arguments work against a rule-based utilitarianism. Even if one admits certain rules into the ethical system, it can be best to follow these rules even when they result in greater aggregate harm than good in particular instances, because the general fact of having the rule can itself add up to much more good than harm. For instance, even though the FDA occasionally holds up the approval of certain beneficial drugs that could save many lives, it (I assume, though more research would be necessary) more than makes up for it by catching harmful and potentially fatal drugs before they are released onto the market. If we just avoided it whenever we had a drug that we knew was beneficial and had no potential harms, then we would damage its power as a regulatory body, and thus weaken its effectiveness for stopping bad drugs. And thus, when the costs of weakening the rule are factored in, things actually are better off overall when we follow the rule than when we break it, even though in the particular case, taking the action prescribed by the rule looks like it makes things worse off overall.

Of course, finding these rules is the hard problem. And then there's the other problem of deciding how ethical rules should relate to law and public policy, which is another debate entirely. But I think it should be mentioned, because some of the arguments Kilroy makes here against deontology seem to apply more to a deontological-style legal system than ethics per se.

Periodic Table of Operators

The Perl scripting language has so many operators that someone decided to put them all together in a periodic table. (Link via Educated Guesswork.) Perl's design philosophy was to add every operator that seemed useful. Taking this philosophy to an extreme would give you the language APL, which requires using a special symbol keyboard in order to code in it.

C++, on the other hand, has very few operators and keywords, but each of them is overloaded in five different ways depending on the context. Taking this philosophy to an extreme would give you only a single operator, with an incredibly rich texture of meanings depending entirely on the context. (The lack of parentheses would also require that the operator be used with a prefix or postfix notation.) Imagine the joys of writing a compiler for such a language...

One More Post On Deontology vs. Utilitarianism

Kenny notes, in the comments to my last post on Kantian ethics:

My real problem with utilitarianism is the fact that it posits an objective value scheme that really can't be known. There's obviously no way to calibrate different people's utility scales against one another, and no way to set any baselines. I remember going to a talk when I visited UC Irvine in which the speaker claimed to show that even under _convergent_ systems by which two people interact and try to figure out how each person's value scales lie relative to their own, there is a radical path-dependency in how they converge, so no objective comparison is possible.

But I think the most damning objection to any sort of deontology (or non-consequentialist ethics of any sort) is that if the "right thing to do" leaves everyone worse off than some "wrong thing to do", then there is clearly a sense in which it is "better" for someone to do this "wrong thing" than the "right thing". ie, it seems to me that consequences can trump deontology, because consequences are real. You can approximate deontology with a sort of rule-based utilitarianism, which I think is probably the best ethical system. But it still has the insurmountable problem of all utilitarianism, unless we apply some clearly false simplifying assumption, like the idea that utility can be measured in dollars.

I'd have to agree with Kenny that weighing different people's utilities against one another is the worst problem for utilitarianism from a philosophical standpoint. (Though not necessarily from the standpoint of our moral intuitions.) There's simply no philosophically reasonable way of doing it. On the other hand, I think the argument he makes against deontology is flawed, from a philosophical standpoint, for precisely that reason.

If, indeed, there was some situation in which every person was better off if action A was taken, but deontology required action B, then it would be right to call that deontology perverse. But this situation doesn't really come up, at least with respect to reasonable interpretations of Kant's deontology. Remember that Kant's commandment is to treat all persons as ends only, and not as means - usually understood as at least requiring that actions be voluntary on the part of all involved. If everyone prefers action A to action B, then it would clearly be treating everyone as an end to take action B - you are helping everyone achieve their preferences.

This, however, might not be exactly what Kenny meant. Suppose we change the situation such that everyone is better off in the aggregate if action B is taken - but some people are worse off than if action A is taken, and those people are being used as a means. (Sometimes its permissible to take an action that makes someone worse off, so long as that person isn't being used as a mere means.) In order to say that deontology here is perverse, we must say that the benefit to those who are helped is greater than the harm to those who are hurt - but this is precisely the problem of weighing interpersonal utilities again. So I think this objection - at least stated this way - doesn't really work. On the other hand, I think any reasonable moral system has to take consequences into account somewhere.

I think the real problems with deontology have to do more with some things that Kilroy mentioned in the comments to this post (though this isn't precisely his view): where do these moral rules come from? If they come from reason, then we have the problem that any deontological system derived from reason must be incomplete. (Perhaps not necessarily, but there don't seem to be any that aren't.) If we appeal to our moral intuitions, we have to remember that our moral intuitions are formed by genetic evolution and cultural evolution, and we have no philosophically good argument as to why they should be in accord with morality. In fact, often they are in contradiction with each other. And, of course, some people's moral intuitions support genocide.

Reason #43 Why I'm A Libertarian

No one else stands up for the good old-fashioned vices anymore.

Fun Ways to Play with eBay

If you have a Gmail account, you might be interested to see that the invites are selling for $50 apiece on eBay.

(Thanks, Bob Schafer.)

Hackers and Kantian Ethics

Kenny mentions below that it might be perfectly compatible with Kantianism to execute hackers. I don't think it is, but it's a somewhat complicated argument.

Kant had two forumlations of the Categorical Imperative that he considered equivalent. (Three, actually, but we'll just focus on two). The one I quoted a few days ago was

Act in such a way that you always treat humanity, whether in your own person or in the person of any other, never simply as a means, but always at the same time as an end.

"Act only according to that maxim by which you can at the same time will that it should become a universal law."

The cornerstone of Kantian ethics is that moral rules are universal - equally binding on all people at all times - and that this fact is a requirement of rationality. Furthermore, to unify the first and second formulations, Kant argues that no one could possible want a universal maxim applied that treats people as a means only and not as an end, simply because such a maxim would be binding on everyone.

Now, this has an interesting twist in terms of Kant's theory of punishment. When Kant talks about respect for peoples' dignity as free and rational actors here, he means respect for their choices about what moral rules are universal. Specifically, we must apply to the actor the moral rules he follows. In a sense, this is a sort of better version of the Golden Rule - treat people the way that they show, by their actions, how they think people ought to be treated.

Thus, from a Kantian point of view, it is not the consequences of an act that is relevant to punishment, but the moral maxim it follows. A hacker essentially commits trespass, or maybe vandalism. He follows the rule that property rights can be violated. A murderer commits, well, murder. He follows the rule that people's lives can be violated.

Kant would insist on an eye-for-an-eye punishment - the hacker having his computers destroyed or made useless and the murderer being executed. I don't think that's a requirement of justice. But I do believe that the amount of punishment should be related not to the consequences of the acts but to the intentions involved. And this is the requirement of justice that utilitarianism disregards in the specific case we discussed and in general.

On the other hand - to put in a good word for utilitarianism - from a public policy perspective, one often has to consider how many dollars one should spent on police protection, knowing that too few dollars spent will result in too few criminals punished, but too many dollars spent will require an exorbitant amout of taking involuntarily from individuals. I don't think that Kantianism has much to say about this. Utilitarianism seems to be the only reasonable way to go here. That's why, again, I think the true philosophy must be some sort of synthesis of these two views - but, then, I don't really believe any coherent philosophy can be in accord with all of our strongly held moral intuitions, so I'm happy to be incoherent for now.