Yea, Nay, or Meh?

November 3 was not a great day for Kentuckians who struggle with health insurance costs, Mississippians looking for a real alternative to their incumbent governor, or people vulnerable to discrimination in Houston. Republican Matt Bevin’s election to the governorship of the Bluegrass State means the possible dismantling of its health insurance marketplace set up under the provisions of the Affordable Care Act. Bevin also wants to make Medicaid recipients enrolled as a result of the law pay more. Remarkably, many Kentuckians who would be most affected by their new governor’s policies gave him their votes, because they were tired of “career politicians” like his Democratic opponent.

In Mississippi, the problems started in the Democratic primary, when trucker and political neophyte Robert Gray won his party’s nomination despite not campaigning. Possible explanations for his victory include his “common name” and his gender. As expected, incumbent Republican Phil Bryant—another foe of health reform, despite his state’s self-evident need for it—won easily in the general election. Meanwhile, 61 percent of Houston voters rejected a non-discrimination ordinance because of the specter of transgender bathroom use—an ugly example of demagoguery and trans-phobia carrying the day, and an unwelcome pre-Veterans Day tribute to those service members who also would have had legal recourse under the measure.

In addition to voters with obvious interests at stake, another group had some troubles on Tuesday: pollsters. The failings of polls were most obvious in Kentucky, where most had Conway winning narrowly instead of losing by almost nine percentage points. Polls in Houston also tended to overestimate support for the equality measure, though they were more accurate. Only in Mississippi, where the Democratic nominee hardly inspired confidence, did polling accurately predict the outcome.

These pitfalls during off-year state and municipal elections might seem unremarkable, but they fit into a wider pattern of polling snafus around the world. In Britain, Poland, and Israel, polls were wide of the mark in major elections earlier this year.

Such concerns were paramount for Harvard History professor Jill Lepore, when she delivered the 26th Theodore H. White Lecture on Press and Politics at the Kennedy School last Friday. As Professor Lepore pointed out in that talk and in a New Yorker article on the same subject, polls have always faced credibility issues, such as in the aftermath of the infamous 1948 election, when President Truman’s reelection was so unexpected that it caught even major newspapers flat-footed. Though the 2016 presidential race has become more poll-dependent, polls are coming under increasing scrutiny as their response rate withers and the emergence of online alternatives raises graver questions of accuracy.

Beyond technical considerations, Lepore argues that our continued infatuation with polls even as their accuracy becomes dubious is a “paradox” that requires explanation. Perhaps most striking, Lepore asks whether polls have a tendency to move politics towards a kind of “direct democracy,” as at least one congressman complained in the 1940s. If a member of Congress supports a bill that she knows (thanks to perfect polling) her constituents do not, should she vote for it? Such is the uncomfortable hypothetical Lepore poses.

Of course, these issues have a long lineage. In his famous 1774 “Speech to the Electors of Bristol,” British conservative politician and political philosopher Edmund Burke articulated what today would be a decidedly anti-polling argument. “If,” Burke argued, “the local constituent should have an interest, or should form an hasty opinion, evidently opposite to the real good of the rest of the community, the member [of Parliament] for that place ought to be as far, as any other, from any endeavor to give it effect.” Translation: A representative’s duty is to his own “judgment” and “conscience,” not to the momentary foibles of constituents.

More recent American figures have voiced similar convictions. John F. Kennedy ’40, whom Lepore cites as having an interest in polling, had a certain Burkean streak. In his 1956 book “Profiles in Courage,” he profiled eight senators who had ignored their constituents in favor of what they saw as the common good. “The true democracy,” Kennedy wrote, “ puts faith in the people—faith that the people will not simply elect men who will represent their views…but also elect men who will exercise their conscientious judgment.”

What do these philosophical meditations on true democracy and representation have to do with last Tuesday, the future of polling, and the upcoming presidential campaign? Most obviously, the results of Houston’s referendum give some idea of why direct democracy, in practice, is often a poor idea. When minority rights are up for grabs in an election with 26.9 percent turnout, those “hasty opinions” of which Burke was rightly wary are at their most influential. In Mississippi, meanwhile, the strange results of the Democratic primary might make a living JFK question his “faith in the people.” Similarly, the Republican victory in Kentucky, driven by a lack of trust in “career politicians” regardless of their policies, also points to a tendency to value vague labels over “conscientious judgment.”

In a world where populists like Donald Trump and Ben Carson are serious contenders for the Presidency, what Lepore calls the increasingly “Trumpian” nature of our politics is especially disturbing. Electing leaders “who will exercise their conscientious judgment” to protect “the real good of the rest of the community” remains the basic imperative of American democracy. If last Tuesday is any bell-weather, however, reality remains far removed from that aspiration. Just don’t expect the latest poll to tell you that.

Nelson L. Barrette '17, a Crimson editorial executive, is a history concentrator in Winthrop House. His column appears on alternate Fridays.