Evaluating the Bush administration's international economic policy is the political equivalent of diagnosing a schizophrenic. Every step forward in Robert Zoelllick's grand strategy for trade liberalization--getting fast-track authority, launching the Doha round of world trade talks--is matched by a blatantly protectionist measure contained in Karl Rove's master plan for reelection, such as the steel tariffs and the farm bill.

The past week was no exception. On the one hand, U.S. trade negotiators reached a crucial deal with less developed countries on generic pharmaceuticals. The deal, which permits developing countries to import generic knock-offs of patented pharmaceuticals in cases of national medical emergencies, gives the cause of trade liberalization much-needed momentum in the run-up to WTO meetings in Cancun this month. Combined with the emergence of an agreement between the United States and European Union on agricultural issues, the deal puts the latest round of WTO negotiations, launched in Doha in 2001, back on track.

On the other hand, George W. Bush's Labor Day speech suggests he has no intention of abandoning his ongoing flirtation with protectionism. The president used the occasion of the speech to announce the creation of a "manufacturing czar" to deal with the perceived rise in outsourcing of jobs to the developing world, observing that:

One way to make sure that the manufacturing sector does well is to send a message overseas, [to] say, look, we expect there to be a fair playing field when it comes to trade ... See, we in America believe we can compete with anybody, just so long as the rules are fair, and we intend to keep the rules fair.

When Bush starts sounding like Howard Dean on the campaign stump, it's time to go short on firms with significant overseas investments.

The administration seems for the moment to be teetering on the precipice between further liberalization and further protectionism. Is there any way to predict which way it will fall?

In a sense, that depends on whether Bush looks at the trade issue as a national security question or a vital reelection tool. The national security argument for further trade liberalization is quite straightforward; the Bush administration itself has made it countless times. For example, the fall 2001 National Security Strategy plainly observed that:

[T]he United States will use this moment of opportunity to extend the benefits of freedom across the globe. We will actively work to bring the hope of democracy, development, free markets, and free trade to every corner of the world. The events of September 11, 2001, taught us that weak states, like Afghanistan, can pose as great a danger to our national interests as strong states. Poverty does not make poor people into terrorists and murderers. Yet poverty, weak institutions, and corruption can make weak states vulnerable to terrorist networks and drug cartels within their borders.

The security rationale is behind the administration's exertions on behalf of the Doha round, as well as President Bush's announcement in May to create a U.S.-Middle East free trade area by 2013. It also helps explain the linkage developing countries perceive between political cooperation with the Bush administration (say, in Iraq and the war on terror) and negotiations over bilateral trade agreements with the United States.

But as strong as the national security pull is in favor of liberalization, the domestic politics of the 2004 election are constantly pushing the White House in the opposite direction. In particular, there's an increasing hue and cry across the land about the phenomenon of "outsourcing," in which firms relocate operations in countries with lower labor costs.

The market pressures of outsourcing cut into two constituencies Bush cannot afford to alienate. The first group is rural voters--part of Bush's base. The small towns in these regions are far more dependent on local manufacturing as an economic engine than the rest of the country. Rural areas suffer proportionately greater economic and social dislocations from these plant shutdowns than areas with more employers. Bush targeted this group in his Labor Day address and in his measures to protect farmers and steelworkers.

The second group is more amorphous but also more dangerous--affluent professionals in the service sector. Doctors, lawyers, engineers, and professors have avoided competitive pressures because their services were, to this point, not thought to be tradeable. The Internet has changed all this. As Raghuram Rajan and Luigi Zingales point out in their recent book, Saving Capitalism from the Capitalists:

For centuries, technology has created new products and new ways of making them that render workers and their skills redundant. While the dislocation stemming from technological change is not new, its pace has increased tremendously. Moreover, it is now affecting the professions that have not much changed their way of doing business over the centuries [emphasis added].

Steelworkers losing their jobs is one thing. But hell hath no fury like a spurned white-collar worker.

Stimulating the domestic economy--not erecting barriers to international exchange--is the best way to help groups affected by economic globalization. Protectionism does little to solve employment problems, although it does succeed in providing massively inefficient subsidies to coveted interest groups. The dynamic benefits of free trade to consumers--through cheaper goods and services--is significant. It far outweighs the costs of lost jobs due to trade (a small yet highly disputed number). As for the increased global competition for white-collar jobs, the law of comparative advantage does not stop working just because 401(k) plans are involved. Protectionists like to focus on the demise of uncompetitive sectors, but the logic of open markets is that greater competition permits the reallocation of labor and capital to the more profitable sectors of the economy. Cushioning this process for displaced workers makes sense--halting the process through protection is a recipe for economic decline.

The one appeal of protectionist measures is political--protecting domestic markets gives the appearance of leaders taking direct and decisive action to help the economy. In an economy in which interest rates are approaching zero, taxes have been cut three times in three years, and the dollar has fallen more than 35 percent against the euro over the past two years, protection is one of the last overt actions the administration has left.

So which impulse will the current administration act on: the need to help out weak foreign countries or the need to help out concerned voters? The golden rule of political science is that politicians' desires to stay in office trumps all. This does not bode well for an open global economy.

The most likely outcome for the next 18 months is a policy of "hypocritical liberalization." The Doha round will proceed, as will the Middle East Free Trade Area. But the administration will take advantage of every exception, escape clause, and loophole at its disposal to protect vital constituencies from the vicissitudes of the global market. This will hurt the broad majority of American consumers and a healthy share of producers that rely on imported raw materials. But hey, there's a rosy future awaiting West Virginia steelworkers.