Republicans tend to boast that markets like them more than Democrats, but until recently there hasn’t been a lot of solid evidence for that argument.

In fact, equity prices have generally performed better under Democratic presidents over the past three-quarters of a century. Identifying the validity of that Republican conviction requires isolating investor partisan preference from other news, which is difficult to do.

There’s a certain irritating circularity to the discussion about the current Republican incumbency. When the economy grows, people say, Bush family members get re-elected, and the greater that growth, the greater the chance of reelection. But which came first: the growth or the politics?

What’s more, many professional investors nowadays muddy the picture intentionally by refusing to say which party’s victory would make them more bullish. The famous efficient-markets hypothesis says the market already discounts for everything, and many participants hide behind that theory like a naked man behind a black umbrella.

Now, three economists have managed to capture the elusive political investor in action. In a recent report, Justin Wolfers at the Wharton School of Business and Erik Snowberg and Eric Zitzewitz at Stanford University identified a moment when it is almost impossible for the economy to affect people’s votes: the intense period when elections are actually taking place.

In that brief time frame, the influence flows only one way: Political news moves the market. The authors found that markets, particularly equities, in aggregate prefer Republicans. Statistics going back to President James Garfield in the early 1880s show that surprise Republican victories moved equity prices as much as 3% higher than where they would have stood had a Democratic candidate won as expected. Their time frame was the 48-hour period that begins on the eve of an election.

Yet it is the experiment, rather than the conclusion, that is worth looking at first. Elections that are hard to call may annoy or humiliate most of us. Few Americans on either side remember with pleasure the weeks of indecision following Election 2000. Still, Messrs. Wolfers, Snowberg, and Zitzewitz recognize that such elections yield the most information about honest market behavior precisely because their unpredictability forces the investor to leap about so.

In 2004, for example, the authors note that exit polls at 3 p.m. on the day of the election suggested that President Bush was going to lose against Democratic nominee John Kerry. Markets tend to follow exit polls, even if they are flawed. The value of $10 futures contracts betting on Mr. Bush’s re-election dropped to $3 from $5.50, and other markets moved down, too.

Later in the evening, as Mr. Bush’s chances looked better, the $10 contracts became more valuable and markets also recovered. News of Mr. Bush’s gains raised the price of the dollar, oil, and equities. The Bush victory also pushed up both real and nominal interest rates.

But what if all this happened merely because markets like the continuity of re-electing an incumbent? The report’s authors sought to control for this possibility by focusing on the 2000 election, another Bush presidential contest and one in which there was no incumbent.

This cliffhanger yielded plenty of action behind the umbrella.The authors found that the prospects of a Bush presidency were worth an increase that day of about 1.5% in the Standard & Poor’s 500 futures, and a 3.5% increase in Nasdaq 100 futures. These figures would suggest it was Mr. Bush’s political philosophy that pleased markets.

What about bonds? Until Ronald Reagan, the authors found, “bond yields were historically quite unresponsive to political shocks.” In 1980, the year Mr. Reagan was elected, there was a near instant increase of 15 basis points on the yield of the 10-year Treasury bill. Since then, Republican presidential victories have tended to raise bond yields.

One explanation for this might be that Republicans are big spenders, and so are likely to inflate, driving down the value of bonds. That certainly would be the sort of argument made by the former treasury secretary, Robert Rubin.

The authors delved into the meaning of the interest-rate increases, at least in recent elections. They used the iShares Lehman TIPS Fund as a meter of comparison between inflation-protected bonds and traditional ones. If the market in 2004 expected inflation to accelerate after a Bush victory, the spread between inflation-protected bonds and others would have diverged as it became clear that Mr. Bush was winning. But yields marched together, suggesting traders bet that growth would make cash scarce.

It’s odd that election-time markets prefer the GOP so strongly. Not even the nastiest Republican can claim that Democrats are always worse for returns on capital than Republicans are. To do so would be to ignore, say, Secretary Rubin’s capitalgains tax cuts of the late 1990s.

More importantly, there is that abiding fact: Markets over the very long term have done just fine under Democrats. The authors have come up with a few explanations: that “past Democratic presidents have pursued policies that were more beneficial for equity returns, but investors have not noticed.”

They also posit that “past Democratic presidents have pursued beneficial policies but investors do not expect future ones to do so.” These theories are improbable.Their third suggestion — that Democrats are lucky — is more likely.

More people believe Republicans are good for the economy than are willing to admit it. It’s a story, all the better because it is hidden most of the time behind an umbrella.

Miss Shlaes is a columnist for Bloomberg News.