We appreciate the thoughtful comments of Michael Dawson, Richard Johnston and Emily Thorson, Rick Perlstein, and Mark Schmitt. They add new insights about 2008 and push us to reflect on our own analysis and interpretations. We can organize our reflections around three questions. How do we explain the outcome of the 2008 election? How should we interpret the outcome? And how can we improve media coverage of campaigns and elections?

A key concept in explaining elections is “structure”—to use a term from both Perlstein and Schmitt—or “fundamentals”—to use the term that we as well as Johnston and Thorson employ. These terms imply the same logic: election outcomes depend a great deal on factors that candidates largely cannot control, especially the country’s economic health. The fundamental conditions can be translated into electoral predictions, as election-forecasting models do. Johnson and Thorson are right to note that these models are imperfect: in 2008 the average prediction was spot-on, but in 2000 and 2004, it overestimated by about two points Gore’s and Bush’s respective vote shares. We emphasize “the fundamentals”  as an antidote to a campaign narrative fixated on candidate strategy and personality, the sort of thinking that leads some people to believe that Michael Dukakis would have been president had he only stayed out of that tank.

Of course, campaigns can and do matter. After all, the candidates’ poll numbers often fluctuate during the campaign, even as the fundamental conditions do not change much. The question is where these fluctuations come from and what they add up to on election day. In one prominent political science theory, the campaign makes the conditions in the country more salient to voters, moving them in line with the prediction derived from the fundamentals. As Lynn Vavreck discusses in her recent book, The Message Matters, one of the jobs of a good campaign is to focus the voters’ concerns. The 2008 election seems illustrative: Obama steadily gained support after the Republican convention as he continued to emphasize the weak economy. The financial crisis may have made this message more resonant.

Johnston and Thorson pour cold water on this interpretation, drawing on the National Annenberg Election Survey (a dataset that we hope will soon be released to the public). During the fall of 2008, Americans indeed became more pessimistic about the economy, but those views became less, not more, related to their vote intentions. The fluctuations in polls appeared to derive instead from the public’s changing assessments of Sarah Palin. If further analysis confirms this finding, then clearly the story of 2008 will turn not only on a weak economy and an unpopular president, but also on candidate strategy, in this case McCain’s surprise decision to select Palin. This would not imply that the media’s wall-to-wall coverage of Palin was somehow prescient: the media may have magnified, rather than anticipated, her importance—a hypothesis that later research can test. Ultimately, Johnston and Thorson’s preliminary conclusion should lead us to recalibrate the balance between structure and contingency in 2008.

Michael Dawson brings contingency even more to the fore. He emphasizes an important fact that also figured in our account: the increased support and turnout of minority voters were crucial to Obama’s win. In 2008 the turnout of black voters increased by five percentage points over its 2004 level, according to the Current Population Survey. Hispanic turnout increased by 2.7 points. White turnout actually declined by one point. We agree with Dawson that the campaign helped persuade and mobilize minority voters. But so, perhaps, did the fundamentals. A weak economy and an unpopular Republican may have predisposed some minority voters, especially Latinos, toward a Democratic candidate. Future research will help us know how much Obama’s coalition was shaped—in its size and composition—by both structure and contingency. We doubt that the answer will render the campaign irrelevant, as Dawson fears. Instead campaign mobilization may prove one means by which election outcomes are brought in line with the fundamentals or, less often but even more importantly, made to deviate from the fundamentals.

Schmitt interprets the 2008 election as a harbinger of Democratic dominance. But are the Republicans truly unable to “form anything close to a majority”?

How should we interpret the 2008 election? Schmitt’s reply, like Perlstein’s, emphasizes historical context—except that Perlstein compliments us on our attention to history and Schmitt thinks that we are not attentive enough. Schmitt rightly notes that interpretations of an election depend on the particular previous elections to which it is compared. In this case, Schmitt suggests we are too focused on comparing 2008 and 2004; to him, the magnitude of Obama’s victory (along with the Democrats’ congressional victories two years earlier) only becomes apparent over a wider span of time.

We agree. One of the misdemeanors of election commentary is to pick a time frame to support a conclusion. Want to emphasize Republican dominance in presidential elections? Pretend history starts after FDR. Et cetera. Like Schmitt, we continue to believe that the 2006 and 2008 elections constitute a historically notable shift in party power.

Schmitt goes on to interpret the 2008 election as a harbinger of Democratic dominance. His interpretation, rooted in observations about the party coalitions, is plausible, given an electorate that is becoming younger and more ethnically diverse. But we are less certain. Obama did rack up impressive wins in many states, but is Wisconsin out of play? Is the Republican Party just a Southern party? Are the Republicans truly unable to “form anything close to a majority”?

It is difficult to say, at least at this stage. The structure of party coalitions can be buffeted by big changes in fundamentals, that other sense of “structure.”  A six percent swing in the national vote would bring the Republicans to parity in congressional voting and a presidential victory. From this perspective, it is not so important exactly which states they win, just that there is enough dissatisfaction with Obama to provoke such a swing. We are not making any sort of prediction. Readers of a certain age may remember that Ronald Reagan dropped in popularity, and his party got whacked in the midterm elections of 1982, but then the economy improved and his reelection was relatively easy (a template that we expect Obama is well aware of).

The broader point is this: in looking at national partisan control, we can go surprisingly far by studying national vote. The translation from popular vote to congressional and Electoral College majorities is not perfect—see, for example, the 2000 election—but, by and large, swings in representation come from corresponding national swings in votes.

That said, analyses of states and districts are absolutely necessary for political strategists who want to know where the competitive races will be, where resources should be spent, and which members of Congress are most vulnerable to challenge from the left or right. Schmitt makes the crucial point that political change comes not only from successful campaigning but also from successful governing, which requires an understanding of individual political players, not just national trends.

The final question is whether and how conventional wisdom about campaigns and elections can be improved. In his gonzo fantasia—and we mean that as a compliment—Perlstein singles out a “pundit-industrial complex” that traffics in “gossip about celebrities.”  We confess to a weakness for gossip—when reading for pleasure, we certainly do not pick up the American Political Science Review—but, like Perlstein, we think that political science helps make sense of American elections. Bringing the classroom to the public, however, has its challenges: our research cycles—from the genesis of an article to its publication—are typically three years long, or approximately two years, 364 days, and some hours more than the standard news cycle. Our efforts in this magazine and on our respective blogs, as well as the like contributions of other scholars, are all attempts to shorten the cycle.

Will these efforts get political scientists invited to Joe Scarborough’s kaffeeklatsch? Probably not. The media ecology fetishizes novelty in reporting and certainty in commentary. And yet the academic study of elections shows that what is certain is almost never new, and what is new is almost never certain. We might only bore Fox & Friends with our scholarly qualifications and caveats, or simply look foolish trying to present our research in soundbites. (Even Boston Review, in the interests of readability and space, asked us to reduce the number of graphs accompanying our article from thirteen to one.) That said, we are willing to risk it if we can take airtime away from yet another discussion of the Bradley effect. Wolf Blitzer: have your people call our people.