The Special Election Fallacy

We have recently experienced another set of special elections, which once more reaffirm the nation’s repudiation of Obama’s policies / have little or no bearing on the national political scene (pick one).
See how they can be spun? Special elections can both symbolize a fundamental turning of the tide in the national political scene, or a particular candidate’s particular triumph in a particular part of the country.
In a country eager to find benchmarks to quantify abstract notions such as a candidate’s or administration’s “approval” or “favorability,” special elections have a rather unique place in the discourse. They provide nations with a definitive numbers fix at times other than the nationwide elections every two years, allowing us to get a snapshot of how the nation feels about its leadership in between.
They’re like polls, only better. Special elections give concrete, hard and fast data, that, unlike polls, reflects actual votes for actual candidates. Under the assumption that individual parts of America are microcosms of the whole, special elections give us insight into how people who voted in general elections now see things at different points in a presidential or congressional term. For a party that triumphs in a special election, a well-needed morale boost gives them confidence moving forward – they seem to believe that special elections signify how not just a district, but the country has changed its mind towards the current governing party. People vote, they claim, to repudiate the president (and they do have a basis in truth when they say this; in the 2009 gubernatorial race in New Jersey, which, though it was not a special election, was at an unusual time, 38% voted in part either to support or oppose Obama, according to exit polls). The media, always almost obsessively eager to find a narrative, jumps on special elections as “referendums” and “gauges” on a Congress or Presidency: the NPR, the Houston Chronicle, the Boston Globe, Fox News, and MSNBC are all guilty of this.
They’re like polls, only worse. They are inaccurate predictors of national election results in terms of turnout, demographics, and voter concerns. Remember the “real America”, “not-so-real America” line of reasoning brought up in the 2008 election? Just as you would not trust the people of Oklahoma to give a resounding endorsement of the Obama presidency, or only college students in a vote to decriminalize banned substances, a vote in New York’s 9th District, which was in the aftermath of a scandal, did not represent typical voter turnout, and happened in, well, one of the nation’s 435 districts. It is not, therefore, a statement on Obama’s presidency. Polls, doubt them as we might (which, if we are behind, we are wont to do), are generally statistically significant, use reasonably representative sample sizes, and provide a sense of how the nation is feeling.
The two recent Republican wins in New York and Nevada should not have, as the NPR claims, “sent a wave of gloom through the Democratic Party.” They are not indicative of voters’ views on Obama.
Except that they are.
Blame the media.
In 2010, Scott Brown’s campaign was seen as a clear sign of a Republican comeback, though the reality was that Coakley, his opponent, ran a rather weak campaign (notably, she went on vacation within a month of the election.) But it was the beginning of a Republican surge – they won back the House last November. Special elections turn the tide, because they give individuals an excuse to claim a change in the mythical “momentum.” It’s one of those things where just because everyone says something exists, it does. (It doesn’t – if it did, Hillary Clinton, after long strings of Obama victories, would not have kept winning elections by ginormous margins near the end of the campaign in places like West Virginia and Kentucky).
Studies have shown that special elections in which a seat flips do offer reliable indications of where a general election is heading. But even then, the insight they provide comes with numerous asterisks. One of them, for example, is that the morale effects of special elections could be more “predictive” than the actual results themselves. Nate Silver of FiveThirtyEight has a brilliant post which helps cut through the “spin.” The media, over-covering the results of special elections, and framing them within the narrative of comebacks, surges, and repudiations, successfully convinces voters – and candidates – that these small, non-representative elections are accurate gauges of national sentiment. Because of this, special elections do play a huge role in the direction the political narrative is taking. Because they can be interpreted very differently based on how they are contextualized, there is too much room for political maneuvering. Even Nate Silver isn’t sold on the overarching meaning of special elections; as he says, “there are always idiosyncrasies.” So let’s not take the giant leaps in reasoning made by the media and the party press offices and remember that though special elections are definitely relevant, they aren’t the most trustworthy tea leaves.
In essence, what happens in NY-9 stays in NY-9. (Unless, of course, you’re Anthony Weiner.)
For a bit on the history of special elections, check out Chad Pergram’s article “The Specialness of Special Elections” at FoxNews.com.
Image Credit: Tom Arthur, from Flickr, via Wikimedia Commons, http://commons.wikimedia.org/wiki/File:Voting_United_States.jpg.