Editor's note: Julian E. Zelizer is a professor of history and public affairs at Princeton University. He is the author of "Arsenal of Democracy" and a book on former President Jimmy Carter, and is editor of a book assessing former President George W. Bush's administration to be published this fall by Princeton University Press.
Princeton, New Jersey (CNN) -- After last week's primaries, the national media tried yet again to figure out what the results mean for the national political landscape. Writing for Politico, John Harris reported that the primaries had brought good news for President Obama and Democrats.
Republicans, he wrote, "were left with several new reasons to wonder whether all the favorable national trends showing in the polls are enough to overcome local candidates who are inspiring little confidence about their readiness for the general election 12 weeks from now."
Although House Speaker Tip O'Neill once noted, "all politics is local," you wouldn't guess that from the election coverage that we have been living through over the past year and a half.
Starting with the gubernatorial elections in New Jersey and Virginia in November 2009, the news media have been looking at every set of results to find some kind of dramatic evidence about what will happen in 2010 or, more importantly, in 2012.
In an era of 24-hour, multiple channel news coverage, the scramble for stories has resulted in a permanent election night. Remember when Americans used to wait every two years for the media to start honing in on the electoral landscape?
These days, the coverage is incessant, and reporters try to place every story, no matter how small, into a dramatic storyline about where the presidency and Congress are headed.
The problem is that outside of presidential elections, local issues and the particularities of each campaign heavily influence most of these contests.
When voters make choices about representatives, senators and gubernatorial candidates, they're heavily influenced by local issues along with perceptions about the specific candidates. Each contest is not always a referendum on national issues. In fact, it's the opposite.
Even midterm elections are historically difficult to analyze in terms of what motivated voters. To be sure, there have been some midterm elections where the contests revolved around national debates, such as in 1966 when Republicans made gains in response to President Johnson's difficulties in Vietnam and the rioting in urban America.
Or, in 1994 when Republican campaigns focused on President Clinton's tax hike and his failed health care proposal. There have also been several forces that are making local contests more national, such as the decline of local newspapers as a source of information for voters.
But O'Neill's point still holds. Most social scientists have refrained from making too much of midterm elections -- other than how the outcome shapes the composition of Congress -- because of multiple factors at play. Plus, turnout for the midterms is notoriously low.
The problem is the modern media environment does not lend itself to localized and disconnected stories.
Instead, in the scramble to find the hottest news item on politics, reporters, editors and producers have been trying to fit every contest into some bigger story about President Obama's political condition or the state of the Republican Party. Good horse-race coverage depends on dramatic turning points and defining moments, so every election must mean something.
The result has been practically a parody of interpretation, with questionable storylines changing every month. The Tea Party's emergence and impact of this political movement has been one of the biggest stories of the year. Yet, thus far, the Tea Party has only been successful in electing a handful of candidates, with a list of losses that is far greater.
Following the losses of Utah Sen. Robert Bennett and Pennsylvania Sen. Arlen Specter, the media have been obsessed with stories about the anti-incumbent mood and evidence that voters are willing to throw any bum out of office.
Yet, as of now, only six incumbents have lost out of 282 federal incumbents who have faced elections. This is hardly a landslide.
"This is about average," the University of Virginia's Larry Sabato, who monitors elections, told Slate.
Last week, Sen. Michael Bennet's victory in Colorado, just like Blanche Lincoln's victory in Arkansas, contradicted the storyline about the anti-incumbent tidal wave.
It might be time to give this kind of coverage a rest. While providing a close examination of elections is a good thing for democracy, trying to fit each contest into a dramatic tale does not strengthen our understanding of the issues shaping the nation. Rather, it produces a skewed interpretation that doesn't necessarily fit the facts.
The obsession with the horse race also takes attention away from other aspects of politics -- such as public policy concerns and long-term demographic changes -- that lack the same drama but might have much more to tell us about where the nation is headed.
The opinions expressed in this commentary are solely those of Julian Zelizer.