With two months to go before the midterm elections, Republicans are staring down a scary political environment. There is no use sugar-coating it. A batch of fresh polling that has come out after the Labor Day holiday seems to confirm deep trouble for the GOP.
But in 2016, didn’t Republicans hear the same thing? For those trying to gauge what may come to pass in November, how much of today’s polling is a valuable guide to how things might go? And are there other data points to consider?
While the end of 2017 saw President Trump reach record low job approval and showed Republicans with massive deficits on the generic ballot, the passage of tax reform and the start of a new year kicked off seven months of good news for the GOP. The president’s job approval went above 40 percent and stayed there by a reasonable margin. Controversy after controversy failed to “move numbers” in the national environment. In fact, the generic ballot lead for Democrats shrank well into the single digits. The economy was growing at a fast clip, Republicans were hanging on in special elections, and there was a sense that maybe, just maybe, Republicans would get a break and avoid losing control of one or both chambers of Congress. Things were looking up.
But that run of good news has ended. The generic ballot is usually the gold-standard metric when it comes to midterms, and Tuesday, the ABC News/Washington Post poll emerged with grim tidings for Republicans, showing Democrats with a 14-point lead. Wrapped up deeper in the data are shockingly slim margins favoring Republicans among key constituencies like non-college educated white men, along with widening deficits among college-educated women. (The ABC poll has historically been one of the better of the media polls, and has an A+ rating from FiveThirtyEight based on its past performance.*)
One should never take any individual poll and consider it gospel. What should be more worrying for Republicans are the many other indicators that are all pointing in the same direction. Multiple polls fielded in the last month of August, and most showed a double-digit lead for Democrats on the generic ballot. In 2006, the last midterm to lead a serious blow to the Republicans, the polling averages showed Democrats with a single-digit advantage in September that widened to a double-digit lead during October.
Republicans would quite like to avoid a 2006 rerun, but the polls are suggesting that the comparisons go beyond the generic ballot. That fall, unemployment was approaching 4 percent, just a bit higher than it is right now in the U.S. During September 2006, George W. Bush’s job approval hovered around 40 percent, which is right where you’ll find Trump’s today.
This is why all sorts of forecasters, after ingesting copious polling, conclude that Democrats are will probably take over the House of Representatives. The folks at 0ptimus are showing Democrats with a 75 percent chance to take the House, with FiveThirtyEight pegging it around 80 percent in their “classic” model that blends polls with fundraising figures and data like historical trends. (Their “polls only” approach is more favorable to Republicans, currently giving Democrats only a 70 percent chance of taking the House.)
But let’s set aside the polls and the models for a moment. Plenty of people still look at the predictions and the forecasts with deep skepticism post-2016. What data points are relevant for them, then? If you’re uncomfortable with trying to predict the future, then let’s look at the not-too-distant past, where we find perhaps the best indicator that Republicans are going to face a surge of energized Democratic voters: the turnout patterns seen across primaries and special elections.
Take my home state of Florida: Republicans are more reliable primary voters and tend to outnumber Democrats in primaries, even in years where there isn’t much interesting going on on the Republican side. In 2010, a year where Republicans and Democrats alike had contested high-profile races, around 860,000 Democrats showed up to vote in the primary, compared to nearly 1.3 million Republicans. In 2014, with almost nothing of interest statewide on the ballot for Republicans, Republicans outpaced Democrats with around 950,000 primary votes to about 840,000.
But fast forward to this year, where both Republicans and Democrats had contests of note. Republicans grew to 1.6 million primary votes cast; clearly, Republicans weren’t staying home. But the growth of turnout on the Democratic size was massive, with around 1.5 million votes cast in their primary. Comparing that to their usual turnout in primaries, and you see that suddenly Democrats have gotten quite good at mobilizing voters who do not often turn out in these kinds of contests.
This isn’t just a pattern in Florida — we’ve seen it in voter file analyses in places like Virginia and more — where you don’t need a poll to tell you a blue wave is coming if the turnout pattern holds. You can look at the votes already cast and begin to see the wave forming on the horizon.
Republicans may want to hold on to the comforting idea that last time, the polls and the models all got it wrong, and therefore they are safe. A look at turnout patterns suggest they should not allow themselves to feel too comfortable.
*Kristen Soltis Anderson is an ABC News political contributor.
