Thursday, October 28, 2004

The incumbent rule

"The incumbent rule" says that in an election with an incumbent, the voters who wait to the very end to decide will usually vote for the challenger. The historical evidence for this rule is overwhelming: "analysis of 155 polls reveals that, in races that include an incumbent, the traditional answers are wrong. Over 80% of the time, most or all of the undecideds voted for the challenger." Therefore, my working assumption is that this Presidential election will follow the incumbent rule, and that Bush's situation in key states is therefore quite precarious.

That being said, over at Slate there's a wonderful "Consumer's Guide to the Polls" which has some data relevant to the incumbent rule's applicability to this presidential election. Some polls press respondents to decide on their vote for President, and for the polls that press and make the relevant data available, the authors tell us what the results of the pressing were, averaged over the last three samples:
Poll Bush boost Kerry boost
AP 0.78 1.23
Battleground 1.33 1
Democracy Corps 1.33 1.67
Gallup 0.75 1
ICR 2.63 1.97
Time 0.33 0.67
My quick and dirty unweighted average of the above data gives Bush a boost of 1.19 and Kerry one of 1.42. Now, one shouldn't leap to conclusions. These voters who only decided upon being pressed may be mere leaners, not "true" undecideds; perhaps they've really already decided but just have trouble admitting it to pollsters and/or themselves. Anyhow, it's something to think about.

A good sport

I think it's very big of William Tucker to write this right now:
I have one plea to conservatives. If the Democrats do manage to squeak through next week, I have one request: Let's let this guy govern. I know it's going to be tempting to scream foul or to start making fun of his daughters or to put under the microscope whether John Kerry really was in Cambodia for Christmas 1968.

But let's at least give him the chance to pull the country together.
I don't trust at all Kerry on foreign policy, and I don't want him nominating Supreme Court justices, and on and on. If Kerry wins, I hope to see incessant Republican criticism and opposition, but I also hope it to be civil. Republicans don't need their own Michael Moore.

Future government growth

I'd really like to think that if Kerry wins, then the Republican House will decide it's time to hold down spending again, thus making lemons into lemonade. However, I think the odds are against it. As Ramesh Ponnuru argues,
The dumb case against Bush regards him as having betrayed the historic Republican commitment to keep spending down from year to year. This history stretches all the way back to January 1995, and all the way forward until the fall of 1996. But the dumb case against Bush doesn't pause to acknowledge that Reagan increased spending, too, especially in a first term of recession and defense build-up, or that the Gingrich Congress cut a big-spending budget deal with Clinton in 1997. The case, in its dumbest form, assumes that anti-spenders can, by denying Bush reelection, cause Republicans to return to the true path. Maybe this would even make sense — if it were not the case that much of the country likes increased federal spending just fine, and far more people like than dislike increased spending for any given program. That's why spending has gone up, after all, and not just under this president.
It really comes down to the fact that there aren't enough congressmen who want a small federal government. Thus, those of us who believe in federalism and/or small government in general have a bigger problem than just our poor options in the presidential election. (Which is why I'm a member of the Club for Growth. Yeah, I couldn't resist making a plug.) Perhaps Republicans in congress, out of a desire to hurt Kerry politically, will stymie everything he does regarding the budget, resulting in lower spending. But I think it more likely that they'll compromise, like they did in Clinton's second term, afraid to repeat, say, the government shutdown of 1995. I don't expect the rate of change of federal spending to depend much on whether Bush or Kerry wins the election. Alas, I just expect the rate to be positive.

More on faithless electors

I see I'm not the only one cooking up scenarios.

Wednesday, October 27, 2004

More on Hawaii

No less than Michael Barone says Hawaii is "in play" and the betters at Tradesports now put the odds of a Bush victory at 19% (bid price). I still can't bring myself to believe that Bush can win Hawaii, just like I can't bring myself to believe Kerry can win Arkansas. Of course, it's possible everything I think I know about these states is wrong. (For what it's worth, I lived in Arkansas for only two years when still in elementary school, and I've never been to Hawaii.) We'll find out next week.

Sunday, October 24, 2004


Now I just feel sorry for the guy.

I never thought I'd see the day....

The last two polls in Hawaii show Bush up by a point! I tentatively agree with the folks at Tradesports and give Bush a 10% percent chance of carrying the state, though that feels like an awfully high probability. On the other hand, the prospect of prolonging election-night parties until after the Hawaiian exit polls come in is fiendishly delightful to me.

Faithless electors

Don't get too excited about that potential faithless Bush elector in West Virginia. Robb says it's "highly unlikely" he'd vote for Kerry, but that he might vote for a Republican other than Bush. I claim it is therefore highly unlikely he will effect the outcome of the presidential election, even if we assume the electoral college comes within one vote of a tie.

According to the Twelfth Amendment, if a candidate gets "a majority of the whole number of Electors appointed" (currently 270 out of 538), then he wins the presidency. If neither gets a majority, then the House chooses the president. If the voters choose more than 270 or fewer than 269 electors pledged to Bush, or if Bush loses West Virginia, then Robb can't make a difference. Let's consider the interesting scenarios: a 270-268 or 269-269 Bush-Kerry electoral split in which Bush wins West Virginia. If Robb doesn't vote for Bush or Kerry, then we get a 269-268 or 268-269 split, respectively. In either case, neither Bush nor Kerry would have "a majority of the whole number of Electors appointed," and the election would get thrown to the House. Thus, Robb only changes things in the event of an original 270-268 split, by making a Bush victory contingent on a House vote. However, as I will explain below, Bush would almost certainly win such a House vote. Thus, Robb's vote still doesn't matter.

The House would choose the President in a slightly complicated manner. Each state's delegation would get one vote, with the choices being Bush, Kerry, whoever Robb voted for, and abstention, the rule being that only the top three scorers in the electoral college may be chosen from. To become President, a candidate needs to win 26 delegations ("a majority of all the states"). (The quorum is "a member or members from two-thirds of the states," a requirement that neither party could deny the other.) Bush has the advantage due to the numerous small Red states. There are 30 delegations with a Republican majority, 16 with a Democratic majority, and 4 tied (derived from data here). In order for Kerry to get 26 votes, he would need to flip ten states. The only reason I can imagine a Republican representative voting against his party is to match his vote with his state's popular vote. Could Kerry win enough states' popular votes to make this work without also winning a majority of the electoral votes? Let's use Slate's very inclusive list of 19 states in which the election outcome is at all in doubt. Of the safe states not on this list, only CT, DE, and IL are safe Kerry states that have Republican-majority House delegations - all by a single representative. Let's assume Kerry gets a Republican defector from each of these states - three states down, seven to go. The states ME, NJ, OR, WV, and WA already have Democratic-majority House delegations; hence, Kerry would need to win the popular vote in 7 of the 14 other states on Slate's list: VA, MO, MI, CO, AZ, WI, PA, OH, NM, NH, NV, MN, IA, FL.

Kerry will surely win the popular vote in NJ, WA, and ME's more populous southern Congressional district, putting him at 182 electoral votes without winning any of the above 14 states. If Bush wins OR, then Bush almost certainly will win more than a majority of electoral votes, so nudge Kerry up to 189, still with seven more states to go. Kerry is even more likely to win MI than OR, so put Kerry at 206 with six more states to go. If Bush wins all of the big three battleground states - FL, OH, and PA - then he'll surely win more than a majority of electoral votes, so put Kerry at 226 (OH), 227 (PA), or 233 (FL), with five more states to go.

Suppose Kerry wins OH and PA, putting him at 247 with four more states to go. That leaves one remotely possible scenario for Robb to very indirectly give Kerry the presidency: Kerry wins NH, NM, NV, IA - and none of the remaining states, hitting 268, and popular pressure causes enough Republican representatives from Kerry states to defect. Kerry winning NV but losing MN is a real long shot.

Suppose Kerry wins only one of the big three. Then Kerry will certainly also lose AZ, CO, MO, VA, and, by hypothesis, WV. That leaves IA, MN, NH, NM, NV, WI, ME's northern district, and perhaps four of CO's electoral votes (if that ballot measure passes and survives legal challenges), for a total of 46 electoral votes. If Kerry wins OH of the big three and all of these 46 EV's, then he's at 272 - too high. Take away NH or the 4 EV's from CO to put him at 268. If Kerry wins PA of the big three, then we could also take away ME's northern district to stay at 268, or equivalently give Kerry NH and the 4 CO EV's but take away NM or (more likely) NV. If Kerry wins FL of the big three, then we need to trim those 46 EV's down to 36, which can be done in several ways I'm not going to list. The bottom line: the sequence of states NV, IA, FL, OH, PA is in descending order of Bush's current strength in the polls, so all of these scenarios are possible but improbable.

Making the relevancy of Robb's vote even more unlikely is the high probability that, if some Republicans representatives are pressured to defect so as to match their state's popular vote, then the same will be true for Democrats: Red states AR, ND, SD, TN, and WV all have delegations with a Democratic majority. (Update: I forgot to add that Red states TX and MI have evenly split delegations.) And we haven't even considered how much pressure there would be on the House to vote for Bush simply because he should have won 270 EV's, not 269.

Well, I think that's enough obsessive electoral math for now. (If I get into this state of mind again, then perhaps I'll figure out the most likely scenarios for a 269-269 electoral split.) The moral of the story is that your vote probably won't matter, even if you're an elector.

Update: Michael Barone also thinks Bush would win if the election went to the House. He also notes an important wrinkle I left out of my analysis: the next Congress would do the voting, not this one. However, Barone argues that "Republicans are very unlikely to lose more than two delegations," "Democrats could easily lose one delegation," and that the currently tied Texas delegation is "overwhelmingly likely to go Republican." Though I won't go into details this time, I think at best Kerry could end up having one fewer delegation to flip. I still think it is highly unlikely that Robb will swing the election, but I leave the details of updating my original analysis using Barone's House election predictions as an exercise for the reader.

Tuesday, October 19, 2004


Ah, the great pastime of making fun of philosophers. (Hat tip to Josh Chafetz.) Since I do math and not philosophy for a living, I can safely laugh, holding my nose up when among the unrigorous, at least until someone asks me whether sets exist, or worse, what a set is. (My respective answers are "Yes," and "Next question.")

Wednesday, October 13, 2004

Debate III

In VP debate and the second presidential debate, I found the domestic policy half distinctly less interesting. So, I unsurprisingly had a hard paying attention to every word of this debate. In the process of giving some live instant-messenger commentary on the debate, I got into a much more interesting debate of my own about immigration policy with some libertarian pals of mine back at MIT. Therefore, I've no business saying who won. I did pay sufficient attention to know that neither candidate won decisively, and that Bush was much more relaxed than in the past two debates.

Update: The consensus (among my favorite blogs) appears to be that Bush won. To me, neither side won decisively because there were many conflicting claims about policy details thrown about, and often neither guy's case was more convincing than the others based on his words alone. I agreed with Bush most of the time, but how would the sought-after undecided voter decide between the two candidates' claims if they weren't intimately familiar with the details? I'd presume such a decision would be based on who looks and sounds more trustworthy. I don't trust myself to say who that would be because I have such a strong preexisting opinion about the matter.

Another update: A slightly wider survey of online commentary reveals a lot of folks calling it a draw. Alas, the most influential commentators are those on television, and I don't watch them, so don't take my survey too seriously.

On blockades

In commenting on the second presidential debate, I mentioned in passing a naval blockade as an intermediate option for how the U.S. might "get tough" on Iran: it lies between sanctions and military force. My good neighbor and pal Brian Ulrich writes that a naval blockade wouldn't work without the "cooperation of the international community" and that sanctions would be just as effective as a blockade in the presence of such cooperation.

I disagree on both points. If others nations besides Iran don't like our hypothetical blockade, then all they will be able to do is scream about it at the U.N. and, if they're really mad, hurt themselves and us by refusing to trade with us. No one has a navy that could break our blockade by force. Thus, the blockade would successfully stop naval trade with Iran, even if every other nation in the world was against it. Now it is true that a naval blockade would not stop overland trade with Iran. However, it would still significantly damage Iran's economy. For one thing, Iran mostly exports its oil by sea. As for Brian's second point, even if we could convince every government in the world to refuse to trade with Iran, that would not stop black market trade. A naval blockade of some form would still be needed to enforce the world's embargo on the seas.

I believe a naval blockade would be quite effective at causing Iran economic pain. However, that doesn't mean it's the best option for stopping Iran from getting nuclear weapons. If we know what buildings to bomb, then destroying Iran's nuclear infrastructure is our best bet. The original reason I thought of a blockade is that I don't think the major world powers will ever agree to stop buying oil from Iran. A blockade would destroy Iran's oil revenues without directly killing anyone or blowing up anything. However, if Iran decides to put up with the economic pain and build a bomb anyway, then we'll be worse off than before the blockade: Iran will be closer to getting nuclear weapons while we'll have paid great economic and diplomatic costs for the blockade. There is no easy solution to this problem, but if force really is to be our last resort, and we really don't want Iran to get the bomb, then a blockade may be in the cards.

For what it's worth, I think the most probable scenario is that we'll still be talking with Europe about sanctions on the day Iran announces it has a nuclear weapon.

Monday, October 11, 2004


Timothy Noah thinks Bush's invocation of Dred Scott in the last debate was code to the Christian right for Roe v. Wade. That's possible - Bush would surely rather talk in terms of promoting a "culture of life" than in terms of judges' opinions on Roe v. Wade. However, the supposed allusion was completely lost on me. (Why didn't I get the email from Rove!?) Anybody who interpreted it as an unambiguous pledge to only appoint Supreme Court justices opposed to Roe v. Wade already trusts Bush to be solidly pro-life, so why bother to make such an obscure allusion? The other religious allusions Noah mentions, "culture of life" and "power, wonder-working power," are unambiguous and unobscure. When I first read Bush's use of "wonder-working power," I immediately heard the chorus from the appropriate hymn in my head. "Culture of life" is a term most familiar to Catholics, but I've seen it plenty of times. I think the most likely explanation is just that Bush wasn't bold enough to explicity mention Roe v. Wade, and that the Dred Scott reference was just another botched attempt of Bush to speak extemporaneously.

Not success

In the spirit of my last depressing post on demography, I link to this Atlantic piece on Russia. If other European peoples are dying out, then Russians are dying out a lot faster:
Heart disease, alcohol consumption, and tuberculosis are epidemic. So is addiction to nicotine. You won’t see many pregnant women on the streets; Russia has one of the lowest peacetime birth rates in modern history. Long life is one of the central characteristics of an advanced society; in Russia, men often die too young to collect a pension. In the United States, even during the Great Depression mortality rates continued to drop, and the same has been true for all other developed countries. Except Russia. In the past decade, life expectancy has fallen so drastically that a boy born in Russia today can expect to live just to the age of fifty-eight, younger than if he were born in Bangladesh.
If Russia is lucky, by 2050 the population will have fallen by only a third, to a hundred million. That is the most optimistic government scenario. More realistic predictions suggest that the number will be closer to seventy-five or eighty million—a little more than half the current population.


It seems that everyone in Afghanistan went to the polls on Saturday, except the Taliban. The biggest problem was not blood on the streets, but ink off the thumbs. Every election involving millions of voters has fraud, but there's fraud and then there's massive fraud that changes the outcome. Exit polling unsurprisingly suggests Karzai won in a landslide.

Even better, Karzai's strongest opponent in the race, Qanooni, has backed off from boycotting the results to supporting the UN inquiry into fraud allegations. I don't think Qanooni had much choice, given the public's overwhemling happiness about their first election. Democracy is a fragile thing, but it is also a strongly coveted thing.

Friday, October 08, 2004

Debate II

My unpolluted opinion is that it was too close to call. (I also thought the first debate was too close to call.) The major change in the dynamics was that Bush spent a lot more time talking about Kerry's Senate record. This factor put Kerry on the defensive more than he was in the first debate. Moreover, the questions weren't all about Bush's record, like they were in the first debate, which further put Kerry on the defensive. However, Kerry skillfully rose to meet the challenge, fighting Bush to a draw, even though Bush did better than in the first debate.

Miscellaneous remarks:

(1) Bush still didn't keep his cool as well as Kerry; he seemed irritated most of the time.

(2) It will be hard to assess the effect of this debate on the polls because it comes only three days after the VP debate and five days before the last presidential debate. Therefore, any poll trend predications are only marginally scientific in the Popperian sense. That being said, if all you care about is foreign policy, then the second debate won't change your vote if the first one didn't - there was little new here. Likewise, if your vote swings on stylistic aspects of the candidates performances, then you won't be swung by the second debate if you weren't by the first. Furthermore, I didn't discern a clear winner on domestic policy, so I wouldn't expect much change in the polls on that basis either, though I'm less confident about this last point.

(3) I really wanted a more specific answer to the question on Iran. I think the questioner wanted to know whether military force was an option or not. Kerry said we'll get "tough" with Iran if we have to, which could mean (more) sanctions, or military force, or something in between (say, a naval blockade). Bush also failed to say whether military force was an option.

(4) Abstaining from reading debate commentary thus far has been difficult for me - it's almost like there's some chemical dependency here - like my need for Dr. Pepper.

Update: There seems to be a consensus that Bush was more animated this time - I noticed but thought the difference was minimal, probably because I didn't think Bush came off as tired in the first debate, just irritated by Kerry, as he was in this debate.

Also, Jonah Goldberg has posted to the Corner an email from a reader arguing that wonkish types like myself don't perceive the debate the same as "normal" folks, a thesis I readily endorse. The moral is to be humble about making predictions of how voters perceive the debates.

For some reason I forgot to mention the Afghan election this Saturday. Coming right after the debate, it's effect on the (American) polls will be inseparable from the effect of tonight's debate. Focus groups of debate-watchers provide some info, but it's connection to national polls must be inferred. Moreover, focus-group polls don't capture the effect of the all-important post-debate spin in the media.

Of course, the Afghan elections are vitally important in and of themselves, regardless of any political points they score over here. I pray they go well.

What geeks photograph

I found this in a computer lab at UW Memorial Library. Can we say "mad geek points"?

Wednesday, October 06, 2004

VP debate

Like Andrew Sullivan, but unlike just about everybody else, my initial reaction was to declare a huge Edwards win. I thought Cheney won on substance overall, and landed some hard rhetorical blows, but I hardly cared. Compared to Edwards, Cheney spoke very quietly, often mumbling, and I therefore subconsciously discounted all but his most forceful moments, giving Edwards victory almost by default.

Andrew Sullivan is standing by his initial assessment, despite the consensus against it. I'm not so confident - I don't know who won anymore. Furthermore, rewatching the debate wouldn't prove anything, because I've certainly been affected by all the commentary I've read. I'll just have let this remain a minor mystery.

Update: This article agrees with me and Sullivan that Edwards was far more telegenic, and notes the CBS poll of undecided voters which favors Edwards. I'm still stuck with my minor mystery: obviously perceptions are subjective, but I'd like to understand why there are such divergent perceptions of this debate.

Slavery, freedom, and the Bible

I just read an interesting argument for the Biblical condemnation of slavery. (Hat tip: Winds of Change.) It links to a much more comprehensive argument made in 1845. As for the latter, there's some assertions about Greek words that I'm not qualified to evaluate. These assertions amount to the claim that every instance of the word "slave" in the New Testament should really be translated as "servant." I'm very skeptical of this. (Hopefully there will be a follow-up post with some more research on this point.) I think both arguments I linked to above persuasively argue for a Biblical condemnation of slavery as it was practiced in the antebellum South, but not for a universal condemnation of slavery. (Even the 13th Amendment allows convicts to be enslaved.) In Ephesians, the apostle Paul tells Christian owners of slaves to be good and kind masters, but he does not tell them they must free their slaves. Of course, this doesn't mean I don't support the 13th Amendment; it just means I don't think it is a corollary of Christian doctrine.

On a related note, I think a Christian could consistently be a monarchist, for example. Jesus said love your neighbor. He wasn't too specific on how to govern your neighbor. I heartily agree that freedom is God's gift to mankind, and that we should help others become more free. (An atheist can examine human nature to reach a similar conclusion.) However, more specific proposals, like spreading democracy, are for me based on prudential arguments about how freedom can best be promoted. (Moreover, freedom is a complicated thing: for starters, it is not license, and various things we call freedoms often conflict with each other.)

Tuesday, October 05, 2004

History lesson

Did we ever get along with the French?

Monday, October 04, 2004

Truth is stranger than fiction

A tale of false intelligence in 2002 that has nothing to do with Iraq. I felt like I was reading a chapter of Tom Clancy.


By all means, let this election be about Vietnam:
Bush obviously stands with the large contingent of Americans who are determined that, if we ever did face another Vietnam, never again would we pull out in a headlong rush and leave our allies sinking in the mud, clutching at our helicopter skids as we fly away, with the wreck of the new and better nation we had tried to build collapsing around their heads. Never again will we treat America's trustworthiness and honor, and the hopes of our friends, and the blood-sacrifices of our soldiers, like bad debts to be written off with a shudder.
Update (10/5): Cathy Young also thinks Vietnam is relevant.

Episodes IV, V, VI

"George Lucas, on the other hand, is so obsessed with airbrushing history that at the end of the day, only Jar-Jar Binks will be left seated on the couch with Lenin." That's Jonathan Last reviewing the latest version of the original trilogy. Read the rest. Then, if you enjoyed that, read this. If not satiated, then go read some old columns of Jonah Goldberg on Star Trek (perhaps this one, or my personal favorite). Google a bit, and you'll discover there are a disturbing number of posts on conservative blogs about the politics and philosophy of various sci-fi universes.

Ok, my sci-fi nerd euphoria is wearing off now. I think I realized it was time to stop after reading this. Hopefully my next post will be about a [clears throat] serious policy discussion.

Sunday, October 03, 2004


Ever since 9/11, I've read quite a bit about how Islamic civilization has handled modernity very poorly. However, in this post I'd like to examine a challenge of modernity that the West is also failing to meet. In Europe, Canada, and Australia, the level of fertility is well below the replacement level. (This is one of many places with data verifying my demograhpic assertions.) Here in the U.S., we're treading water, just barely at the replacement level of 2.1 children per woman. (Our population is expected to grow by half over the next fifty years, but that's mostly because of immigration.) Only in Latin America is the West still growing. However, I suspect their fertility will have dropped to our levels (or even European levels) by the time Latin America reaches current European levels of prosperity. Comparing Latin American fertility rates in the 1970s to those of the 1990s reinforces my suspicion.

Bluntly, half of the West is dying out, and there is a real risk that the other half is but a generation behind in this trend. Meanwhile, the Muslim and Indian civilizations are growing. (And China would grow too were it not for her tyrannical government's one-child policy. Growing fastest of all is sub-Saharan Africa, which I'm not prepared to classify as part of the West, though Christianity is the predominate religion there.) Therefore, Bernard Lewis' remark that Europe will be Islamic by the end of the century should not surprise us, given European immigration policy and the overwhelming tendency of Muslim immigrants to Europe to stay Muslim and have more children than native Europeans.

Perhaps every human civilization will eventually depopulate or at least demographically stagnate in the face of modernity for universal reasons. Perhaps all but a few backward corners of the world will have reached a harmonious demographic equilibrium by the end of the century. Even under this scenario, though the West of the New World will survive, the West of the Old World will more likely than not be just a memory. Of course, darker scenarios are also possible.

Europe is not blind to all this, and has become ever so family-friendly in its work rules, but to little effect. Having even two children is still undesirable for the vast majority of Europeans. In economic terms, they have a shortage of a public good: it is in the interests of each of them for their fellow citizens to be fertile, but not for themselves to be such. If Europeans want to preserve themselves, then they must achieve either a radical change of attitudes towards children, or a radical increase in state incentives for larger families. However, the latter is likely impossible without the former.

The enemy of my enemy....

I noticed these in front of neighbors' doors today. Keep up the good work, guys!


This site is to be the new home of my preexisting blog. For now, if you want to see my old posts, then go here. I plan to port all my old posts to this site, but that will take a while.

Update (10/5): Archives ported back through April 9th, along with some earlier linked-to posts. My Blogger profile says I've made 79 posts in one week, which is a tad misleading.

First Post

Testing 1, 2, 3.

Friday, October 01, 2004


I'd give you my take on who "won" the debate, but J. McIntyre over at RealClearPolitics already said essentially what I'd say. And if you don't like his analysis, then go to RCP's main page for links to many other debate analyses.

While I'm feeling superfluous, I may as well note that in case you like seeing pictures of Madison, Ann Althouse is by far the more prolific Madison photoblogger. (Apparently she's a law professor here. I think I walked by the law school once.)