While Monday’s vote in Iowa wasn’t exactly a vindication for public opinion researchers, the caucus results — the narrowest of wins for Democrat Hillary Clinton, a more decisive victory for Republican Ted Cruz over Donald Trump — lay well within the expected universe of possible outcomes. More importantly for pollsters, an embattled profession suddenly got a much better grip on what it’s dealing with in 2016 than it had even 24 hours before.
The run-up to the Iowa caucuses was an unsettling time for opinion researchers. Developer Donald Trump’s long reign atop the GOP presidential field was so unexpected that it made producers and consumers of poll statistics doubt what they were seeing. That came after pollsters had failed to predict resounding victories last year by Britain’s David Cameron and Israel’s Benjamin Netanyahu. “The world may have a polling problem,’’ declared statistics guru Nate Silver.
Writing in The New Yorker in November, Harvard historian Jill Lepore argued that changes in consumer behavior — she noted that more 40 percent of American adults now lack landlines, while cellphone numbers legally can’t be auto-dialed by pollsters — had damaged the reliability of traditional polling techniques. Lepore also raised the possibility that the constant measurement of public opinion by ham-handed means was undermining American politics.
As 2015 wore on, Trump’s constant recitation of his most favorable poll results hardly seemed like the sign of a healthy system. Earnest political sages wondered, what on earth is going on? A common refrain in recent months went like this: For Trump to defy both GOP leaders and all electoral precedent, everything experts know about presidential politics must be wrong.
The Iowa outcome suggests that pollsters are facing two separate issues. Yes, there are some genuine methodological obstacles. But the other complication is that, amid deeper shifts in the electorate, this year’s campaign climate is unusually volatile. Lots of voters, both Democrats and Republicans, are rebelling against their party establishments in ways that American political scientists don’t often see and can’t easily measure, and the unusually large number of candidates on the GOP side gave voters more opportunities to switch allegiance.
Iowa filled in some important details. Yes, the tide against the Republican Party brass was powerful. More than 60 percent of GOP caucus participants supported Trump, political novice Ben Carson, or Cruz, a freshman senator whom other leading Republicans openly despise. But Trump’s underperformance in Iowa doesn’t necessarily mean the polls are unreliable; it may just mean that he is. Come to find out, a celebrity candidate with a limited field operation didn’t live up to his poll numbers after petulantly skipping a debate in Des Moines just days before the Iowa caucuses. Who knew?
In retrospect, people who conduct and consume polls were freaking out the most at the moment of greatest fluidity and murkiness in the presidential race. But especially as more GOP candidates drop out, polling the race will likely get easier.
Some questions still linger: After Trump’s weaker-than-expected performance in Iowa, will he live up to his current numbers in the Granite State? (Hillary Clinton’s 2008 comeback suggests that a loser in Iowa can rally; Howard Dean’s loss in 2004 does not.) But these are known unknowns. On the more cosmic question of whether traditional polls have outlived their usefulness, Iowa has an answer: Not yet. Certain norms still apply.
Dante Ramos can be reached at dante.ramos@globe.com. Follow him on Facebook: facebook.com/danteramos or on Twitter: @danteramos.
