Thursday, August 26, 2004

This is dumb

I just found out that the IOC is preventing athletes from posting any online diaries or blogss from the olympics:
The IOC's rationale for the restrictions is that athletes and their coaches should not serve as journalists — and that the interests of broadcast rightsholders and accredited media come first.

Tuesday, August 24, 2004

Overtime - What They're Saying

There has been a fair amount of press coverage of the new overtime rules from the Department of Labor that took effect this week, but the range of opinions has been strangely limited. I checked the New York Times, Washington Post, USA Today, and the LA Times for stories on the topic. I found a few quotes from politicians, labor leaders, and some others, and not a single op-ed piece. The Kerry campaign blasted the changes as being a "a shameful assault on the paychecks of hard-working Americans." The biggest disagreement seems to be over how many workers will be affected by the change (somewhere between 107,000 and 6 million).

I found virtually nothing acknowledging that the labor market is, in fact, a market. Most of the opinions expressed seem imply that empoyers will just take away the overtime premiums and pocket them, with no compensating effects on wages or hours, although there are a few mentions of the idea that overtime pay encourages employers to create new jobs rather than increasing worker hours. In fact, according to this story in the LA Times, even such a suggestion might be politically incorrect:

Many provisions in the draft plan that had become political lightning rods were removed or changed.... Gone is language suggesting that employers could reduce overtime costs by cutting the hourly wages of newly eligible workers and adding back the overtime to equal the original salary.
The lone opinon piece I found is this "Ask an Expert" column from Steve Strauss in USA Today's "Money" section. He also takes a dim view of the changes:
Overtime pay serves two useful functions. First, it allows applicable employees to supplement their income. According to an analysis by the Economic Policy Institute, overtime pay accounts for up to 25% of weekly earnings of eligible employees, with the average amount being $161.

Secondly, overtime pay serves a useful position in the overall economy. By requiring overtime, the federal government crafts an incentive for busy employers to create new jobs; rather than pay overtime, a new employee can be brought on. So the old system helped in many ways.

I found almost no quotes in favor of the change, although the very end of this Washington Post article did have this:
But the Department of Labor disputes that analysis and says it changed the rules because they desperately needed updating. The old rules did not mesh with the new economy, causing increased litigation, officials argued. Overtime class-action suits have doubled since 1997 and outnumber discrimination suits for the first time. Labor recovered $212 million in back wages in fiscal 2003
Most interesting for me as an economist, I found not a single quote from an economist. Economists seem to have surprisingly little to say on this topic, for reason's I'll comment on in a future post.

Posner poses a question

The brilliant, controversial, and unbelievably prolific Judge Richard Posner is guest blogging over at Larry Lessig's blog. In addition to a lot of interesting posts about copyright, check out his views on terrorism:
Although the 9/11 Commission’s report is a good read, and has other virtues as well, one of its greatest weaknesses is its failure to address, other than in passing, terrorist risks that are even greater than that of another 9/11: in particular the risks of bioterrorism, nuclear terrorism, and cyberterrorism. The Commission’s recommendations are concerned essentially with preventing a more or less exact repetition of the 9/11 attacks, which are any event the least likely form of a future terrorist attack, since surprise has been lost. We give our adversaries little credit if we suppose that the only attack they can launch is the one we’ve anticipated. They didn’t make that mistake on 9/11; why should they now?

Friday, August 20, 2004


The issue of how to interpret statistical margins of error seems to come up on various blogs all the time, the latest being this post from Matthew Yglesias. It seems to be very difficult, even for intelligent people, to understand why it is not quite right to say things like "there's a 95 percent chance that Kerry is in the lead." Perhaps the easiest way to see why this is wrong is to think about how you would interpret multiple polls.

Suppose you take two polls on the same day with two different samples, asking each person if they plan to vote for Kerry. You are interested in knowing the true percentage of people in the sampled population who would say they are planning to vote for Kerry on that day. Let's say in your first poll, 53% of the people say they are going to vote for Kerry, and in the second poll 47% say they are going to vote for Kerry. If these polls have a 3% margin of error, then the first poll gives you a 95% confidence interval of [50%,56%], and the second one gives you a confidence 95% interval of [44%,50%]. It would be unusual for the polls to come out that far apart, but these things can happen with random sampling (just like you sometimes might toss a coin and get 8 straight heads.)

Now, how do you interpret the results? It makes no sense, from either a Bayesian or frequentist perspective, to say both that there is a 95% chance that Kerry is in the lead, and there is a 95% chance that Kerry is behind. What could such a statement possibly mean?

The correct (although a bit cumbersome) interpretation is to say that both intervals are "constructed using a procedure such that the interval will contain the true value 95% of the time." In other words, if we were to take a million samples and construct confidence intervals for each, we would expect about 950,000 of the intervals to contain the true K, whatever that might be.

In this example we can see that at least one of the polls failed to contain the true value, so we know we've observed a somewhat unusual sample (or two). But the correct interpretation still makes sense, while the incorrect one doesn't.

(BTW, I agree with Matt that statements about "a statistical tie" are usually hogwash.)

A Bold Prediction

I've been away from the blog for a while, so let me return with a bold prediction. I predict that Bush will win the election, but lose the popular vote by an even greater margin than in 2000. I'm surprised I haven't seen this prediction anywhere else (yet).

Actually I'd put the likelihood of this occurring at somewhat less than 50%, but I do think it's a serious possibility. Bush seems to be really disliked in the blue states, even by libertarian leaning moderates like me who voted for him last time. I can't imagine that he won't lose by a much bigger margin in most of these states than he did in 2000, and I can't imagine he is inspiring much new enthusiasm in the red states among those who didn't vote for him last time. So I think it's very likely he'll lose the popular vote. On the other hand, while Bush may not win a lot of new supporters, I think he inspires loyalty in a lot of his previous supporters, and I don't think Kerry is the right candidate to pry many of these Red state voters away from Bush. If Bush can manage to just win Florida and a couple of other swing states, he'll likely get the election.

The prediction futures markets seem to agree with me: on the Iowa Electronic Markets, the latest prices give Bush about a 50.5% to 51% chance of winning the election (despite Kerry's lead in current polls), but the vote share market has him dead even with Kerry with 49% of the vote each. The differences in the markets at are even more extreme: the latest prices give Bush about a 51.5% chance of winning the election, but only a 46% chance of winning the popular vote.

Prediction number two: If this happens, a lot of people will be very angry. There will be efforts to amend the constitution to get rid of the electoral college, but they will fail. I don't think the republic will crumble, but I think it will be pretty bad for public confidence in our system.