In reality, winning a nomination fight elevates the stature of the victor, who quickly brings partisans into the fold (especially during conventions), offsetting any damage to party loyalty or unity that the primary might seem to have incurred. By the time of the general election, the state of the economy plays a dominant role in determining who wins and loses, not whether one party’s candidates were mean to one another at a time when relatively few people were paying attention.

Moreover, while the winning candidate may have to spend more money or campaign harder to win in a divisive primary, he or she can also benefit from the organizational efforts required to win a tough primary fight. President Obama, for instance, seemed to perform slightly better during the 2008 general election in states that were more competitive during his nomination fight against Mrs. Clinton.

Why, then, is belief in the theory of divisive primaries so pervasive? One factor is the seeming correlation between divisive presidential primaries and general election losses. But vulnerable incumbents tend to attract credible challengers, whereas strong incumbents do not. When researchers take the state of the economy and the approval ratings of the president into account, the relationship disappears.