A Quick Guide to Election-Night Punditry

(Bloomberg Opinion) -- Believe it or not, the quality of political punditry has improved dramatically over the last 20 years. That's especially true in print (paper and electronic) journalism. It's even true on TV — overall. Still, the cable networks in particular have hours to fill and little reason to be rigorous, and the demands of live television tend to turn even the best analysts into glib, cliche-spouting know-it-alls. So here are some things to look out for while you're watching the returns Tuesday. 

Ignore exit polls. At some point in the late afternoon, early exit polls will leak. I know it's hard, but ignore them. There's information there, but it's only a piece of the puzzle and shouldn't be taken at face value. Then people on TV will start to tell you what's happening and why based on exit polls. Although those surveys may not be entirely useless, they have never really been all that reliable, and more careful survey research often leads to very different conclusions.

By the way, while most of the people on TV will be using traditional exit polls to figure out who won, that's only one of three methods this year:

All have strengths and weaknesses, but especially if you're following the election on Twitter, be aware that the network calls, the AP calls, and the Decision Desk HQ calls may differ from one another for a while (and the networks can each come to different conclusions, too; they work off the same data, but they each have their own analysts.)

Every election matters. There are thousands of elections being contested. It's true that the two most important results are the majorities in the U.S. House and Senate. But everything is important. The margin of victory matters a lot in the Senate, where individual senators can have a lot of influence. And it matters in the House: a party with a 10-seat majority won't have the same kind of control as a party with a 30-seat majority. And that's just the national level. Every governor and every state legislative chamber matters a lot for public policy in that state. We've had a lesson this year about how important state secretaries of state can be. There's also been a series of lawsuits filed over the last decade by state attorneys general that attempt to drive national policy. 

It's not a preview of 2020. We can't help it; even though we have a government where separate institutions share powers, we all tend to think of the presidency first, even when it's not on the ballot. And it's not wrong to think of the 2018 elections as a reaction to Donald Trump's presidency. Still, the midterms' results won't predict the 2020 election. President George H.W. Bush had pretty good midterms in 1990 and lost in 1992. Presidents Bill Clinton and Barack Obama were crushed in their first midterms and won second terms fairly easily. Democrats are probably going to do OK or better in Wisconsin, Ohio, Michigan and Pennsylvania, but that doesn't say anything about what will happen in those states two years from now. There's plenty of time for the economy to change and for Trump's approval ratings to rise or fall. 

Be skeptical of those who claim to know what “the people” are asking for. For one thing, it's quite possible that Democrats will win the majority of the vote in the House but fail to pick up a majority of seats. In the Senate, Democrats will win a majority of the contests, and the vast majority of votes, but may lose a couple of seats. Even without those complications, it's never easy to know what “the people” are saying with their ballots; the truth is that elections don't do what we want them to do, and the ballot is a lousy method for giving specific instructions to politicians. 

This also applies to what people say about their own votes. Social scientists find that we are pretty bad at understanding the reasons for our own actions. Most people, after all, are voting with their political party, but they'll tell you why this thing that Trump did or that thing that Nancy Pelosi did made them change their minds. Nope; it was party, all along.

And be extra careful about claims of fraud. It's just about certain that Trump will flat-out lie about the results in one way or another. He's been misstating the facts about the 2016 election for two years. We should also expect him to claim again, without any evidence, that Republicans were hurt by voter fraud. As usual, this presents difficulties in reporting on what the president says. On the other side, especially if they fall short in the House, Democrats are likely to complain about gerrymandering, rules that made it harder for Democratic constituencies to vote and other skullduggery. Some of these complaints will be accurate; some will be exaggerated; and some may well be phony; just because the Republican president is friendly to false conspiracy theories doesn't mean that Democrats are immune to them.

To contact the editor responsible for this story: Max Berley at mberley@bloomberg.net

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Jonathan Bernstein is a Bloomberg Opinion columnist covering politics and policy. He taught political science at the University of Texas at San Antonio and DePauw University and wrote A Plain Blog About Politics.

©2018 Bloomberg L.P.