Poll Dancing

Will the industry redeem itself for the mid-terms?

Illustration by Peter Noonan

The outcome of the 2016 presidential election was shocking to nearly every American. Reportedly, even Donald Trump himself didn’t expect to win the contest. But then he began winning state after state that he was supposed to lose.

There wasn’t much that Republicans and Democrats agreed on during that election, but they did agree on the collective wisdom of the polls, and those polls said that it was a sure bet that Hillary Clinton would defeat Trump.

And as we head into the homestretch of the 2018 midterm election, there are two points all voters should keep in mind: Polls will still drive the converstation, and there is no reason to believe that these polls are any better than the ones from two years ago.

You’d think that the 2016 election would serve as a crisis moment for the polling industry, but instead of considering what went wrong and how to improve methods, literally nothing has changed in the way polls are being conducted.

While there has been a noted skepticism over polling in the press and less coverage of polls, bad polls are still being put out, still being reported on, and still frame how everyday people, candidates and political donors see their races.

Specifically regarding New Hampshire, in the two years since the 2016 contest, there have been two developments on local polling, but neither has been a fundamental shift. The first is that, after years of criticism, WMUR-TV ended its official relationship with the University of New Hampshire Survey Center. This came after the last WMUR/UNH Granite State Poll before the 2016 election showed Clinton winning the state by 15 percentage points. She did win the state, but only by three-tenths of a percent.

Meanwhile, the Saint Anselm College Institute of Politics has now begun polling in a regular, but infrequent way. Their poll, it should be noted, is outsourced and conducted by a third party. It’s a new, untested poll, so it’s still unclear if their numbers should be seen as credible either.

For at least a decade, the best polls of New Hampshire are conducted by firms from outside of the state. There is no secret sauce involved, but organizations such as Suffolk University, CNN and NBC actually spend serious money to do serious polls here.

This isn’t a presidential year with the state’s contested Electoral College votes up for grabs, so there is likely to be very little polling of the Granite State. However, what polling does come out will drive press coverage, decide who will get to participate in debates, and impact how candidates campaign. All this in spite of the fact that there is little reason for these polls to have that kind of impact.

Where does that leave the voter?

It means that voters shouldn’t vote based on who is expected to win, but should pick who they think will do the best job. Because, really, who knows who is actually winning until the votes are counted?