The opening ad salvo against Donald Trump’s presidential campaign was launched last week and I’ve found myself wondering if he’ll crumble under the eventual advertising onslaught.
Of course, it’s possible he’ll still be top of the polls when the snow flies because his negatives are “built-in,” or largely irrelevant given the unusual nature of his appeal. Which will it be? Nobody knows. The predictions are all bluster and guessing.
In fact, we don’t have a clue what share of the vote Trump would win if the early states voted tomorrow. There has been one public poll to date that was based on a list of registered voters and gave any indication as to the preferences of actual likely Republican primary voters. That survey was conducted by the Democratic data firm Civis Analytics and it’s the only poll I could find that connects primary vote frequency to actual voter history. And having that information is the only thing that can give us some certainty about which supporters will actually turn out in the nominating contests.
There’s been about 20 other Republican primary polls just since July included in the RealClearPolitics average, most of which provide little useful information. They report the preferences for all respondents who say they’re Republican, or who say they might vote in a Republican primary. But the vast majority of these respondents will never vote in a primary.
The polling firm PPP calls a list of registered voters, but doesn’t appear to weigh its results by vote probability. Moreover, straight polls can never give us a glimpse into the future. These polls are talking-head feed: Pointless grist for the horserace article mills. The news-reading public deserves better data. We can know what the future most likely holds with a bit of work.
Back in 2011, my firm, Evolving Strategies, conducted a first-of-its-kind primary message experiment testing positive and attack ads on then-presidential candidates Newt Gingrich, Mitt Romney and Rick Perry. Just like now, pundits battled over how robust Gingrich’s surge really was. Some claimed Gingrich’s long career meant the voters knew the candidate and his baggage was priced into his stock. Others argued voters didn’t really know Gingrich, what he stands for, or what he’s been doing since leaving office.
We ran the message experiment Dec. 8-10, right at the time Gingrich peaked in the RealClearPolitics averages for Iowa at around 31 percent. Our data showed Gingrich collapsing over 15 points under the weight of just a single attack ad. We found that Romney’s negatives looked priced-in, and he maintained support.
Back in Iowa, Gingrich collapsed 18 points over the next few weeks of saturation-level attack ads. Romney, who began that summer with an average of 22 percent support, ended the race with a little over 24 percent.
In other words, we had solid evidence of Gingrich’s glass jaw and Romney’s plodding consistency a month out from the election and before the onslaught of attack ads. The collapse we observed in our experiment came to fruition on the ground in Iowa when the caucuses concluded.
Fast forward to 2015. Trump is certainly different from Gingrich in a host of ways. We don’t know if Trump has a glass jaw like Gingrich, but the pundits’ views are pretty much the same. It’s as if they’ve dusted off the “he’s doomed” and “he’s for real” articles from the turbulent 2011 Republican primary archives, dropped in Trump’s name but otherwise kept the same arguments.
The fact is it’s easy to spin a plausible argument for any particular side when we don’t have solid data one way or the other. Without an experiment that tests how he holds up under a serious attack, we’re just adrift on a sea of rough speculation.
Jeb Bush’s “The Real Donald Trump” video raised the question: Will Trump’s legs be cut out from under him by a flood of news clips featuring the businessman’s self-proclaimed love for socialized medicine and the Clintons? On the flip side, would Bush be in trouble if the airwaves are full of ads referencing murderous illegal immigrants?
I have a guess. I could even marshall a pretty compelling case for each side. But I’d much rather turn those guesses into hypotheses and scientifically test them. Short of that, I suppose we’ll all have to struggle to survive the upcoming months-long flow of punditry pap.
Adam B. Schaeffer, Ph.D., is the Director of Research and co-founder of Evolving Strategies, a data and analytics firm dedicated to understanding human behavior through the creative application of randomized-controlled experiments.