The most important lesson of the 2012 presidential campaign, in my view, was not that polling-based models are foolproof ways to assess the political environment, but instead that undisciplined ways of evaluating polls and political events can lead to flawed conclusions. On several occasions during the race, news media commentators either overrated the amount of information contained in outlier polls and jumped the gun on declaring a change in momentum — or insisted that a candidate had the “momentum” in the race when there was little evidence of it.
The past year-and-a-half hasn’t made me optimistic that things are getting better. Late last year, the news media badly overrated the political consequences of the government shutdown. Just a couple of months later, it somewhat overhyped the lasting impact of the botched rollout of Obamacare. (I think that case is more debatable, but President Obama’s approval ratings have improved by about 4 percentage points from their lows in December.)
The general flaw is in overestimating the importance of recent events and assuming that short-term trends will continue indefinitely: that a candidate rising in the polls will continue to do so, for example. In fact, especially in general elections, candidates gaining in the polls see their position revert to the mean as often as they continue to gain ground.
The political news media are by no means alone in committing this mistake. It’s a close cousin of the hot-hand fallacy. This is the tendency — also evident in sports commentary — to place too much evidence on recent events, which may be idiosyncratic or essentially random compared with longer-term averages and patterns.
Still, the news media may be especially prone toward overhyping purported “game-changers” that make for snappy headlines. Two weeks ago, after Sen. Mitch McConnell beat a more conservative rival in the Republican primary in Kentucky, some in the political media were ready to declare another momentum shift, claiming that the tea party was “losing steam” to the GOP establishment. But Tuesday night in Mississippi, incumbent Sen. Thad Cochran received fewer votes than challenger Chris McDaniel, a state senator who is often associated with the tea party. (McDaniel appears as though he’ll finish with just under 50 percent of the vote, however, so the race is probably headed to a June 24 runoff.)
As I wrote after McConnell’s win, the “tea party” may no longer be a useful analytical concept. But how Republican incumbents are faring against their challengers is a more tangible measure. For example, we can chart the share of the vote received by Republican incumbents in their Senate primaries back to 2004.
The medium-term trend has been toward more competitive Republican primaries. McConnell’s race was consistent with the pattern — the 60 percent of the vote he received led to a comfortable victory but was less than most incumbents have received in the past. Here’s what our chart looked like after his victory:
And here’s what the chart looks like after Tuesday’s results, accounting for Cochran’s race in Mississippi and Sen. Jeff Sessions’s uncontested win in Alabama, which we count as 100 percent of the vote.
The charts are not that different; in fact, the trend line hasn’t really budged one bit. But the news media’s narrative about the tea party will probably shift a lot — especially if McDaniel prevails in the runoff.
The trend toward more competitive Republican primaries may eventually revert to the mean, too. Over the years, incumbent losses in the primaries have been fairly rare, but they’ve gone through cycles of occurring relatively more and less often. There were seven incumbent defeats (among both parties) in Senate primaries between 1978 and 1980, for example, but then none for 12 years. At least this is a trend that has taken shape over a decade, and not just a few weeks.