With Republicans gathered in New York, the lackluster job market must dishearten President Bush. He had hoped that a strong economic recovery would favor his reelection, and in some ways, he's gotten his wish. Gross domestic product (the economy's output) is almost 9 percent higher than at the peak of the 1990s boom, and business investment -- which had dropped sharply -- is up 14 percent from its low point. Jobs remain an obstinate exception. Monthly increases in payroll employment improved earlier this year, averaging 242,000 from February to May, but have since slowed. They were 78,000 in June and 32,000 in July.

Soft job markets in some swing states must especially worry Bush. The unemployment rate is 5.5 percent nationally but 5.9 percent in Ohio and 6.8 percent in Michigan. Beyond that, weak job growth casts a broader pall. Consumer confidence, though well above recession levels, has retreated from recent peaks. Ditto for stocks; the Dow (as of Aug. 27) was off 5 percent from its 2004 high. Wage gains have been modest, in part because there's surplus labor. For hourly workers, wages -- after inflation -- fell about 1 percent from July 2003 to July 2004, says Jared Bernstein of the liberal Economic Policy Institute.

Explaining the slow job growth isn't easy, particularly now. Everything is politicized. John Kerry slams Bush's policies -- and Bush boasts that they're working. Where they agree is in the presumption that presidents can deliver prosperity. Politicians, the media and the public embrace this notion. Unfortunately, it isn't even a half-truth. More like a sixteenth. A president's policies do affect the economy. But they're one of many influences. The others (including the business cycle, technology and the Federal Reserve) usually dominate.

Let's start with two popular theories of slow job growth: Bush's tax cuts and global outsourcing.

Over the long term, budgets should be balanced. But in an economic downturn, they should move toward deficit to stimulate private spending. Well, you can't fault Bush there. In fiscal 2000, the surplus was $236 billion; for fiscal 2004 the Congressional Budget Office projects a $422 billion deficit. It's possible to condemn (as many Democrats do) Bush's pro-rich tax cuts. A more middle-class tilt might have translated into more consumer spending. It's also possible to retort (as many Republicans do) that Democrats would have moved more slowly toward providing a stimulus. Regardless, the tax cuts bolstered private spending. But the resulting economic growth produced fewer jobs than expected. Why?

Although outsourcing could be the reason, it probably isn't. The stories about software jobs and call centers moving to India aren't make-believe. But the numbers are small. Charles Schultze of the Brookings Institution concludes that perhaps 155,000 to 215,000 business-service jobs shifted abroad between late 2000 and 2003. Similarly, Schultze reports that government surveys attribute only about 4 percent of mass layoffs in the past two years to "import competition" and "relocation overseas." Even if these estimates are too low, they suggest that the impact of job loss abroad is exaggerated, Schultze writes.

The bigger cause of slow job growth, he contends, is higher productivity. Companies and workers became more efficient. That's ultimately good; it raises living standards. But higher productivity can temporarily lower employment. Fewer people are needed to do the same work, and new jobs don't instantly materialize. From late 1995 to late 2000, productivity (output per hour worked) grew 2.6 percent annually. During the next three years, annual growth averaged 4.1 percent. If it had stayed at the lower level, there'd be 2 million more jobs, Schultze estimates. Unemployment would be about 5 percent.

No one knows what caused faster productivity growth. It may be technology. Or it could be cautious corporate managers. "Business confidence remains fragile," says Mark Zandi of Economy.com. Companies try to make better use of existing workers; they'll hire only when they're convinced that higher sales justify it, he says. That could explain the recent slowdown. Consumer spending -- possibly hurt by higher gasoline prices and the fading effects of tax cuts -- weakened in the spring; job growth soon subsided. Whatever is happening, presidents can't control either productivity or job creation. These are the amalgams of actions by countless firms and workers.

The actual jobs story and the campaign's stories are barely connected. The first is complicated and uncertain. Indeed, government employment surveys disagree on the extent of job creation. By contrast, the stories told by both campaigns are simplistic. Kerry and Bush both pledge policies that will produce more jobs. Their promises sound reassuring even if they aren't particularly plausible. But they do represent the agreed-on terms of debate. The Labor Department releases its next employment report on Friday. Whatever the numbers say, they will entitle one -- or maybe both -- of the candidates to claim vindication.