In his victory speech Saturday night, John McCain proclaimed that the winner of the South Carolina Republican primary has always gone on to the nomination. While true, that assertion is less impressive than it sounds.
The catch is what we political scientists call a “small ‘n.’” It is hard to derive forecasts from a small number of cases, which is what we have here. The state’s Republicans have been holding presidential primaries only since 1980. Because there were no serious nomination fights in 1984 or 2004, there were only five contested primaries before this year. Five cases do not an iron law make.
What’s more, four of the five are not comparable to 2008. South Carolina Republicans picked Reagan in 1980 and the elder Bush in 1988, but both had already achieved frontrunner status. Bush and Dole beat Pat Buchanan in 1992 and 1996, but Buchanan never had a serious chance at the nomination anyway.
So all we have is the 2000 primary, in which George W. Bush recovered from a New Hampshire loss to John McCain. Even that case has only limited relevance to recent events. From the earliest days of the 2000 campaign, national polls consistently showed most Republican voters favoring Bush. McCain always lagged far behind. Bush also had crushing advantages in campaign finance and GOP establishment support.
So eight years ago, New Hampshire was an anomaly. South Carolina merely put the contest back on its natural course.
Unlike 2000, this campaign lacks a “natural course.”
No current candidate enjoys the dominance that Bush had in 2000. McCain is currently ahead in national polls, but Republican voters have been very fickle this season. He is also struggling to make up for several bad months for his campaign treasury.
Up to this point, at least, success in one state has not brought victory in the next contested event. Huckabee went from first in Iowa to third in New Hampshire and Michigan. McCain won New Hampshire, and then lost Michigan to Romney. Sensing defeat, Romney pulled out of serious contention in South Carolina. McCain has now won that state, but it would be wildly premature to suggest that he is sure to win Florida.
This discussion brings us to a larger point about the uses of campaign history. Any generalization about presidential nomination politics will suffer from the “small ‘n’” problem. The United States has had only a few dozen presidential elections, most of which do not offer usable precedents. Through 1968, party leaders largely ran the process, which is why Hubert Humphrey could become the Democratic nominee without winning any primaries. The current system, in which most convention delegates come out of primaries and caucuses, dates back only to 1972.
So from how many election cycles can we draw our lessons? Just nine.
Over those nine elections, changes in the process have confounded the search for durable patterns. Thanks to new laws, regulations, and court cases, campaign finance is fundamentally different from the days of Jimmy Carter. And with “front-loading,” it’s a whole new calendar, too. In 1980, South Carolina held its primary on March 8. Now several states are voting in January, and a big bolus of delegates will come up on February 5.
Some commentators have belittled Rudy Giuliani’s strategy of focusing on Florida and letting other candidates split the earlier states. In 1988, they note, Al Gore fell flat with a similar approach. But 20 years later, the compressed schedule could mean a different outcome. And besides, what failed for Al Gore might work for an actual human being.
In the nomination contest, history is not destiny. John McCain got a boost, not a lock.
— John J. Pitney Jr. is the Roy P. Crocker Professor of Government at Claremont McKenna College.