Polls

Washington is preoccupied with polls, which is one reason why so much election coverage rides on the release of new polling data. Political reporters love polls because they make for an easy narrative hook, and the investment in polling by large news organizations guarantees headlines. Candidates love pointing to polls to build momentum for their campaigns, and there are enough polls out there for everyone to find something to trumpet. Bernie likes to point to polls showing him stronger than Hillary in general election match-ups against Trump. Hillary likes to point to polling averages showing her leading the general election field. Trump can point to a couple of recent polls showing him ahead nationally and close in key states.

We can end up drowning in all this information if we don’t have a clear way to understand it. And often we don’t. We’re not trained to interpret polls. Neither are many journalists. And with their perpetual need to hype the story, reporters often give us a myopic interpretation of polling data that focuses only on one survey without putting it in context. Over the past few days, those polls showing Trump leading Clinton have led to breathless coverage about how Clinton’s general election lead has evaporated and, not coincidentally, to bouts of panic among some of my readers who momentarily have forgotten not to worry.

So let’s take a step back and consider what polls can do. A good poll can give us an estimate of public opinion at a fixed point in time. The closer we are to the election and the more rigorous the methodology, the more meaningful the estimate. And that’s it. At this stage of the election, polls cannot tell us who is going to vote on Election Day. They cannot tell which people will change their minds. They can only give us a snapshot of the moment, and it’s usually blurry.

Now, if we see a number of polls pointing us in the same direction, they are probably telling us something is happening out there, but that doesn’t make it meaningful. There is good evidence that the presidential horserace tightened this past week, but that doesn’t mean the race will be close on Election Day. It just means that something is happening at the moment to push things in Trump’s direction. The movement we’ve seen is likely a function of natural Republican movement toward Trump in the aftermath of the primary season. If that’s the case, we would expect to see an increase in Hillary’s support once the Democratic contest ends (and it will end eventually). But it would be just as unreasonable in that case to say Hillary is opening a lead as it is to say Trump is presently closing the gap. It will of course be interesting and consequential to see how strongly the two candidates can consolidate their support. But that story hasn’t been written yet.

It’s always better to look at an election’s fundamentals to get a less volatile sense of where things stand. There is plenty of evidence suggesting that Trump is a poor fit to the electorate and will have an enormous challenge appealing to voters outside the Republican base. He receives historically low approval ratings from women, voters of color and young people, all groups he will need to reach if he is going to win. Such fundamental attitudes do not change overnight and can be more reliable predictors of where the election is going. If we see changes in these attitudes over time as the result of Trump’s campaign efforts, then we may begin to speak of a close election. But that didn’t happen in the past week.

Polls will inevitably become a central part of the election story as we get closer to November, and at that point they will be much more useful, especially as people begin seriously tuning into the election. Until then, it’s best to remember that we still have five and one-half long months until the election, and leave the preoccupation with May polling to official Washington.