Tuesday’s special election in Ohio’s 12th Congressional District is likely to be close, but several things are entirely predictable. There will be plenty of hyperventilating pundits on Tuesday night, and then on Wednesday, the winning side will be confidently saying that this outcome is a bellwether for what will happen in the November midterms. Meanwhile, the losing side will be trashing its candidate and campaign, explaining why this race was an anomaly and why not too much can be gleaned from it.
The fact is, we knew a year ago that the Republican majority in the House was in danger, we knew it a month ago, we know it today, and we will know it tomorrow, next week, and all the way up to Election Day. Special elections in odd years and early in election years can give us an indication of which way the wind is blowing and a rough idea of velocity, but once you get within 100 days or so of the general election, electoral dynamics are pretty much set. In modern history, we’ve never seen a directional change in the last three months of a midterm-election campaign. Waves can stay the same or increase in the closing months, but they don’t reverse direction or dissipate.
If Republicans hold onto the Ohio seat, it won’t likely be by much, and if they lose it, it probably won’t be by much. The fact that such a strongly Republican district—one that hasn’t voted for a Democrat for Congress since Moby Dick was a guppy—is in play is another sign of this being a tough year for the GOP, as midterm elections usually are for the party in the White House. But this is not a new revelation.
Nationally speaking, it’s perfectly normal for presidents to have job-approval problems going into their first-term midterm elections. Trump’s numbers are in the same ballpark today as many of his predecessors were at this point. The July 30-Aug. 5 Gallup poll put Trump’s approval rating at 41 percent (54 percent disapproved). At this point, Presidents Jimmy Carter and Bill Clinton had 41 percent approval ratings, Ronald Reagan’s was 42 percent, and Barack Obama’s was 44 percent. (George W. Bush’s approval was 67 percent, still in the post-9/11 stratosphere). Though Trump’s overall numbers are similar to others, the trend line is very different. Most presidents start high with their honeymoon period, then gradually decline over their first two years in office, a phenomenon that New York University political scientist Paul Light has called a “cycle of decreasing influence.” The theory is that as presidents lose popularity, they lose clout. The other half of Light’s dual cycles is that as presidents continue in office and master the office and the issues, they increase their effectiveness, though it is unclear that this is happening either.
Instead, Trump’s numbers started low and have remained “remarkably stable," as noted in a recent Pew Research Center report, but with “a wider gap between Republicans’ and Democrats’ views of Trump than for any other U.S. president in the modern era of polling.” The Pew report noted that the partisan polarization of presidential approval numbers is a continuation of a trend seen with Obama and George W. Bush. Interestingly, Trump, Obama, and Bush had roughly comparable approval ratings among their own party members; it is the levels of opposition among members of the opposition party that have changed. Pew Research reports that an average of just 7 percent of Democrats have approved of Trump's job performance over the course of his presidency so far; Obama had 14 percent approval among Republicans during his tenure, and under George W. Bush 23 percent of Democrats approved.
Three weeks ago, Trump’s approval rating among Republicans in the Gallup Poll was 85 percent. It bumped up to 87 percent the week before last and 89 percent this past week. Among Democrats, during that same period, Trump’s approval dropped from 11 to 8 to 7 percent, while his rating among independents dropped from 37 percent three weeks ago to 34 percent two weeks ago, to 33 percent this past week.
Given that midterm elections are referenda on the incumbent president, one worrisome sign for Republicans in the Pew study is that while 84 percent of voters who identify as Republicans approve the job that Trump is doing, the share among independents who lean toward the Republican Party is 14 points lower, at 70 percent. Conversely, there is little difference between Democratic-identifiers and independents who lean toward Democrats: Trump’s approvals are 10 and 12 percent, respectively. Historically, roughly 90 percent of partisan identifiers vote for their own party’s candidates but about 80 percent of independents who lean to a party vote for that party’s candidates. This year, the independents who lean toward Democrats seem more likely to behave like Democrats, and those independents who lean Republican might be a little more problematic for the GOP, a turnout concern.
What would be the difference between a tight contest for control of the House and a big Democratic year is the makeup of the electorate this fall. On the generic-congressional-ballot test, the Pew study showed that the fight between Democrats and Republicans is pretty evenly matched among voters 35 years of age and older: Democrats have a 3-point lead among the 35-49-year-old cohort, Republicans have a 1-point lead among those 50-64, and the race is even among those 65 and older. But among those between the ages of 18 and 34, Democrats have a whopping 20-point lead, 57 to 37 percent. Drilling down deeper, Republicans are actually ahead by 3 points among men age 18-34 but are 44 points behind among women in that age group. While young people are making a lot of noises about voting in big numbers and Trump’s numbers are worst among young people, historically you could go broke betting on younger voters showing up at the polls, particularly in midterm elections. In Obama’s two presidential elections young turnout was up a bit, but it was dismal in both 2010 and 2014, as well as in previous midterm elections.