Dan Drezner, subbing for Andrew Sullivan, discusses problems with forecasting models and the media members who latch onto them. One notable oversight in forecasting: virtually all of the existing models predict the nationwide vote, rather than the outcomes of state elections to the electoral college—a particularly problematic consideration when dealing with close elections, like that in 2000. The ones that do make state-level predictions are rather dated.

More to the point, as Matt Yglesias points out, aggregate-level models are often inherently problematic. The problem that Yglesias calls “specification searching”—or what I’d call atheoretical modelling, with a healthy dose of stepwise regression to boot—is endemic to the whole class of forecasting models, because fundamentally they are inductive exercises, focused on finding the best combination of variables to predict the observed outcome. Most good social science (or science in general, for that matter), by contrast, is deductive: establish a truly explanatory theory, develop specific hypotheses, and operationalize and test them.

That isn’t to say, however, that unemployment doesn’t belong in the model at all; it may, for example, be the best available indicator of a theoretical construct like “voters’ perceptions of the national economy.” But as someone whose research interests are more centered on individual-level explanations of behavior, rather than attempting to explain aggregate outcomes, I sometimes wonder if aggregate-level models trade too much scientific value for their parsimony.

See also James Joyner, who points out that small sample sizes aren’t necessarily problematic when the universe is also small. However, in a small sample the good social scientist will be particularly attentive to the potential issue of outliers—atypical observations that can lead one to make conclusions that aren’t justified on the basis of the data as a whole.

Daniel Drezner, fresh off his 300-comment-inducing disagreement with blogosphere folk hero James Lileks, notes both progress and regress on the trade front by the administration, with regress apparently beating out progress quite handily.

Mike van Winkle of The Chicago Report is hosting a discussion on the rationality of voting. As I note in the comments, I don’t think Downs’ conception of rational voting is quite inclusive enough to explain why most people vote in the United States and other democracies where voting isn’t compulsory.

Virginia Postrel comments on a WaPo interview with Howard Dean that gives her the impression that Dean is “the thinking man’s Cruz Bustamante” (which may actually be an oxymoron). It’s fairly clear that Dean’s still tacking left; quoth Virginia:

Dean is running as a guy who wants to control the economy from Washington and who sees business as fundamentally bad. “Any business that offers stock options” covers a lot of companies, including some of the economy’s most promising and dynamic.

Regulation tends to be relatively invisible to the general public, in part because it’s mind-numbingly technical. That makes it much more difficult to reverse, much easier for interest groups to manipulate, and much more dangerous to the general health of the economy than the taxing and spending that attract attention from pundits.

She also has a challenge for the so-called “libertarians for Dean.” Ultimately (assuming Dean gets the nod) they’re going to have to decide whether being pissed off because Bush knocked off Saddam Hussein is sufficient reason to hand the keys of the economy—namely the federal regulatory apparatus—over to a man who barely pays lip service to capitalism.

One of the classic quotes of politics comes from French neo-fascist Jean-Marie Le Pen. Asked, after losing in the first round of a French presidential contest (in 1988, I believe) to Jacques Chirac and François Mitterand, who he would back in the second round, he described the choice as one between “bad or worse”*; in 2004, hardcore libertarians are going to have to decide which is worse, but for now that honor seems squarely to belong to Dean.

Jacob Levy approvingly notes Joe Lieberman’s response to this nonsense. Like Jacob, I’d hate to see the Democrats return to their bad old protectionist ways, but outside a few DLCers like Clinton, Lieberman, and 1990s-Gore I don’t think the party ever really shed its protectionist bent; when Clinton spearheaded expansion of NAFTA to include Mexico in 1994, he did it with mostly Republican backing on the Hill.

Daniel Drezner comments as well, as does Andrew Sullivan; this is my entry in James’ inaugural Beltway Traffic Jam.

Both Matt Stinson and Robert Garcia Tagorda note George Will’s Sunday WaPo column on the politics of the steel tariffs—and of the European Union response to them. Like Robert, I hope this development gives the administration the final push it needs to abandon the tariffs, before this escalates to a trade war which neither the U.S. nor foreign states can win.

Daniel Drezner is displeased at the news that the administration may try to evade the WTO ruling against the steel tariffs. The adminstration plans to maintain these protectionist barriers despite evidence that the steel tariffs cost many more American jobs—in many industries that use steel—than they saved.

The open political question is whether the tariffs are causing enough damage to the overall economy, including the economic recovery, that their marginal benefit in states like West Virginia is offset. The trouble here is that the marginal benefit from the tariffs is easy to quantify, because it is concentrated, while the damage is diffuse—thousands of jobs spread across perhaps two dozen states. And that damage could get far worse if it leads to a trade war with the European Union, who—in this case, at least—are clearly in the right.

Like Dan, I hope the administration will come to its senses. But I can’t be optimistic, especially since the dynamics of the Democratic campaign preclude almost any criticism of Bush from that quarter for not being enough of a free trader.

I thought the only dilemma I was going to be faced with this week was figuring out which side I despised more in the Colonel Reb Foundation versus Richard Barrett dispute (it’s Barrett, by a hair, although I have to give mad props to the Foundation for giving Barrett a new excuse to come to Ole Miss in the first place). Now, however, comes word that Donald Luskin is allegedly siccing a lawyer on Atrios; in this one, I think I have to feel sorry for the lawyer.

They’ve apparently kissed and made up. How sweet…

The Commercial Appeal on Sunday extracted its head out of the buttocks of the Turley-Belz-Lightman Memphis land-speculation elite just long enough to take a look at the city’s abuse of eminent domain as part of the massive, taxpayer-subsidzed Uptown redevelopment project—a project that wouldn’t exist without said land-speculation elite—near St. Jude. Money graf: a quote from Henry Turley, one-third of the prop-spec Axis of Evil, which wouldn’t have looked out of place in the early 20th century “slum clearance” movement:

Henry M. Turley Jr., one of the private developer partners in the Uptown project, said there’s a clear public interest in clearing out blighted areas, and it’s imperative that municipalities use the legal tools available to them. He believes that governments aren’t using eminent domain enough in consolidating tracts large enough for redevelopment.

Collaborating in this shameful exercise are everyone’s favorite Memphians, the Memphis Housing Authority (slogan: “Nobody found guilty of corruption in 7 days!”). Quoth MHA executive director Robert Lipscomb:

Lipscomb said the authority is careful to protect individual property rights while at the same time not unduly enriching those who might try to stall and make a windfall.

Damn straight, Robert; the only people allowed to make a “windfall” in this are Turley and Belz. Heaven forbid any poor bastard who actually had to live in Uptown before the city decided to clear the place out benefits from the exercise.

Dan Drezner is mildly in favor of lifting the trade embargo on Cuba. While I think he slightly overestimates how totalitarian the Cuban regime is—I think it’s done a very effective job of brainwashing much of its populace, and it is almost as brutally oppressive toward political dissidents as the North Korean (DPRK) regime, but I don’t think it has as effective a repression apparatus as North Korea has or some of the old Soviet client states (most notably Romania) had, and by all accounts there’s a degree of economic freedom at the margins absent in the DPRK—I agree that simply removing the embargo won’t lead to miraculous political change. However, it will deprive Castro and his Hollywood apologists of their one legitimate grievance against the United States government—and, for that reason alone, the sanctions regime should be removed.

More thoughts on this are at YankeeBlog and OxBlog.

Warning for those offended by “France-bashing”: the extension of this post contains some.

Two months ago, Daniel Drezner noted the split over whether the European Union is an international organization or a supranational authority among IR scholars (my answer, when asked to provide one when I took an International Organizations course in the Spring of 1999, was “Yes and Yes“), and that upcoming events in France and Germany would help settle that question—in particular, whether those countries would be punished for violating E.U. treaty commitments.

Today, Glenn Reynolds notes that France is getting a free pass for violating the “growth and stability pact” that members of the single European currency agreed to; as Pieter Dorsman at Peaktalk noted yesterday, this isn’t exactly popular with smaller countries like the Netherlands who actually abided by their commitments to the pact.

Alex Tabarrok notes that the Senate has passed, 95-0, a bill that would make it illegal for employers to use a person’s genetic information in hiring or firing. That’s good.

But the bill would also make it illegal for insurance companies to use genetic testing to deny coverage or set premiums. As Alex points out, that’s bad. (I can’t tell from the story whether this applies only to health insurance, or if it applies to life insurance as well.) This can only exacerbate the problem of adverse selection, leading to higher insurance rates for everyone, and pricing more and more people out of the insurance market.

Like Alex and Tyler’s guest-blogger, Lloyd Cohen, I’m skeptical about Alex’s optimistic suggestion that the problem can be solved through “genetic insurance“, which one would purchase before taking a genetic test, and which would cover one’s raised premiums in case one had, e.g., a genetic predisposition toward heart disease. How could the genetic insurance provider be sure that the purchaser had not already taken the genetic test ahead of time, to determine whether it would be worthwhile to purchase the genetic insurance? Again, we have the problem of adverse selection.

Advances in genetic research hold out a great deal of promise for improved health care in the long run, but in the short run they will inevitably lead to a collapse of the insurance market (both life and health) for diseases that have a genetic component. Ultimately, I believe this will force us into a single-payer government-run health system, which can pool risk by making insurance mandatory. Unlike many other liberals, I’m not sure this is a good thing. But good or bad, it appears inevitable.

Update: corrected the permalink to Lloyd Cohen's post.

Alex Tabarrok says he wants to move to Switzerland to take advantage of the low taxes and the declining marginal income tax rates. (Declining as a function of income, that is.)

He may want to reconsider, after the xenophobic Swiss People’s Party gained 11 seats in the recent Parlimentary elections, making them the largest bloc in the governing coalition.

On the subject of marginal tax rates, could one of the econo-bloggers perhaps explain the argument in this paper that declining marginal rates are a property of “optimal tax systems“? The paper completely lost me, so you’ll need to explain it in short words that a mere philosopher turned computer geek would understand.

I’m willing to reconsider my view that an income tax with increasing marginal rates strikes the best balance between fairness and minimizing economic disincentives.

My co-blogger has equated Richard Stallman’s proposed abolition of copyright with slavery. Kevin Aylward has equated Stallman’s agenda with Communism.

Both are being unfair to Stallman.

First, let’s look at Aylward’s charge of Communism. Aylward writes:

Stealing the product, regardless of the extreme moral relativism employed by Stallman, is wrong. And he’s not just talking about teenagers downloading copyrighted materials on Kazaa, he wants the remove the rights of the content producers as well. Your output as an artist (or programmer) belongs to EVERYONE. Replace the word EVERYONE with STATE and what do you get?

Communism

Who owns the air we breathe? “No one” would be the best answer. “Everyone” might be just as good. But that’s hardly the same as the air being owned by the state, and it does not make the USA a Communist nation.

Next, let’s look at Chris’s charge of slavery. Chris writes:

Taking away that choice by requiring them to give away their work—Stallman’s ultimate utopia—is morally indistinguishable from telling programmers they are slaves. That Stallman would have the state feed and clothe the authors of software and other works makes it no less slavery than if the system were operated by rich white plantation owners.

Let’s just set aside the fact that the vast majority of software development is not creation of software for sale. Part of my job is software development, but the stuff I develop would not be of the slightest interest to anyone but my employer. (As a matter of fact, the software I develop for my employer is in the public domain.)

In most countries, the state claims a monopoly on law enforcement. If you want to be a cop, you have to work for the state, and accept the state’s terms of employment. Cops are fed and clothed by the state. Does this make them slaves? No, because they have the option of getting some other job.

Personally, I would not be in favor of completely abolishing copyright. But Stallman has something interesting and valuable to add to the ongoing dialog about copyright protection. And unfair accusations of Communism and slavery do nothing to further that dialog.

Kevin Aylward does me the huge favor of explaining my distaste for Richard Stallman’s agenda. Indeed, in my opinion, the key reason why producing free software is morally superior to producing proprietary software is that the author is making the choice to give away the fruits of his labor for the benefit of others.

Taking away that choice by requiring them to give away their work—Stallman’s ultimate utopia—is morally indistinguishable from telling programmers they are slaves. That Stallman would have the state feed and clothe the authors of software and other works makes it no less slavery than if the system were operated by rich white plantation owners.

One of Karl Marx’s most famous aphorisms is that capitalism would eventually collapse due to its own internal contradictions. While old Karl wasn’t a very good prescriptivist (ask the Russians or the Chinese), he did come up with a useful coinage. And, today, Pieter Dorsman of Peaktalk takes up that theme in discussing the future of Canada, on the day that the leaders of Alberta and British Columbia signed an agreement on interprovincial cooperation that might be the precursor of a secessionist movement in the Canadian West. One telling reason why the provinces might cooperate:

Almost one-quarter of Canada’s population lives in the two provinces. In 2002, Alberta and B.C. produced $300-billion worth of goods and services, one-third of the national total.

In other words, the per-capita contribution to national GDP of Alberta and British Columbia is 50 percent higher than that of the rest of Canada. And now, these provinces face serious damage to that economic power in the form of Ottawa’s insistence on ratifying the Kyoto accord, which will undercut their advantages in natural resource production.

One is reminded of the situation of the American South prior to the Civil War. To say it was about slavery is both true and to miss the point; the abolition of slavery would have severely damaged the economies of the Southern states, and the leaders of the southern states saw no alternative for preserving their economies but secession. They gravely miscalculated in thinking that the rest of the country would accept that, and quite clearly were wrong to have adopted slavery as the basis of their economies in the first place, but to them secession was preferable to economic collapse.

Is this analogy perfect? Not really. Canada’s central government doesn’t have the military power to prevent secession, and probably wouldn’t be permitted by its Supreme Court (or, more likely, by the United States) to use it even if it did, and the global warming issue is not as morally unambiguous as slavery. But the fundamental lesson—that preserving a region’s economic strength may be a cause for secession—is still valid.

Matthew Stinson has the definitive word on the topic in comments at Dan Drezner’s place.

Pejman dissects Krugman’s latest. Apparently Krugman has concluded that it is truly impossible for him to be both honest and polite at the same time, at least when writing for the New York Times. Wow. Simply wow. You’d think that acquiring that skill would be a prerequisite for finishing grad school.

Tyler Cowen has the scoop on the Economics Nobel prizes, which are being given to the inventors of two time-series econometrics techniques: ARCH and cointegration. As Tyler points out, Granger is more famous (perhaps even infamous) for his contribution of the concept of “Granger causality”; the typical joke is that, by the Granger defintion, summer “Granger-causes” fall (or autumn, if you don’t live in North America).

Anyway, very cool stuff; I’m not a time-series guy myself, mainly because there isn’t all that much great cross-sectional time-series data on mass political behavior at the individual level, but ARCH and cointegration are a big deal for political scientists looking at things like presidential approval and aggregate voting behavior over time, and the Nobel is well-deserved by both.

Kevin Drum proposes a taxonomy of “poor”, “middle class”, “rich”, etc., based on income. I was surprised to discover that I’m upper middle class (albeit at the low end of it).

It was so much easier back in the fifties:

If you drove a Chevy, you were lower middle class.
If you drove an Oldsmobile, you were middle class.
If you drove a Buick, you were upper middle class.
And if you drove a Cadillac, you were well off.