If the midterms were a referendum, Trump won
By Musa al-Gharbi
November 9 at 6:15 PM
Musa al-Gharbi is a Paul F. Lazarsfeld Fellow in sociology at Columbia University.
After flipping dozens of seats in the midterm elections, Democrats are set to take control of the House of Representatives. Many pundits and analysts have attempted to frame the results as a referendum on President Trump. Among these, there seems to be a consensus that the president has somehow been “repudiated.”
Not so fast.
To be sure, there are reasons for Democrats to cheer: Despite
significant structural disadvantages and a difficult Senate map, some great ballot initiatives were passed, state legislatures got bluer in many instances, and Democrats won governorships in some key states. These are worth celebrating (in contrast with claims to have “won the popular vote,”
which are spurious). Yet, on balance, Democrats should be more disturbed than comforted by how the elections shook out.
For instance, turnout was
much higher than in 2014. However, the increased engagement
proved to be bipartisan: Trump’s supporters
also showed up in force, significantly undercutting the expected “blue wave.”
Yes, Republicans ultimately lost control of the House — but even here, the Democrats’ continued weakness shines through:
It was expected that the Republicans would lose a significant number of seats, irrespective of public opinions about Trump. Republicans had many more difficult House seats to defend than Democrats overall. There were
twice as many Republican incumbents defending House seats in states Hillary Clinton won in 2016 than there were Democrats defending seats in states Trump won.
Republicans also had
more than twice as many “open” House seats to hold on to as their Democratic rivals had: 36 Republican representatives chose not to stand for reelection this year because they were retiring or seeking another office. Seven others either resigned or otherwise left office before the election. As a result, Republicans had 43 House seats to defend without the
benefit of a true incumbent candidate. On top of this, Republicans had three “open” Senate seats, and one more with a pseudo-incumbent (interim Mississippi Sen. Cindy Hyde-Smith took office in April).
Yet Democrats managed to win surprisingly few of these “open” contests. In the vast majority of cases, a new Republican was elected instead, and they tended to be even
closer to Trump than their predecessors. So Trump actually cemented his hold over the Republican Party: Most of his staunchest Republican critics have either stepped down, been removed through a primary challenge or otherwise failed to win reelection. On top of this, virtually all of the Senate Democrats who
voted against Supreme Court Justice Brett M. Kavanaugh from the states that Trump won in 2016 were voted out of office and replaced by Republicans.
Historically speaking, Democrats delivered a thoroughly average result in their first round as Trump’s opposition.
Going all the way back to the Civil War, there were only two instances when a new party seized the presidency but
didn’t lose seats in the House during their first midterm elections: Under President Franklin D. Roosevelt in 1934 (during the Great Depression), and President George W. Bush in 2002 (in the shadow of the 9/11 terrorist attacks). Even
including these outliers, the average attrition during a party’s inaugural midterms is 35 House seats; excluding these two exceptions, the average loss is 41. Regardless of which number we run with, Trump could end up performing better than average in preserving his party’s influence in the House. He performed
much better than his last two Democratic predecessors: Bill Clinton
lost control of both chambers in the 1994 midterm elections. Barack Obama saw historic losses in the House in 2010, and lost seats in the Senate as well — the
most sweeping congressional reversal in 62 years.
Yet, not only did Trump suffer far less attrition than Obama or Clinton in the House, his party will
gain in the Senate. This may not be surprising given the slanted map against Democrats. It is also somewhat typical overall: Between 1862 and 2014, the president’s party picked up seats in the Senate during their first midterms 56 percent of the time, lost seats 37 percent of the time and broke even once. In other words, there did not seem to be a thorough rebuke of Trump. In fact, there was little exceptional in the results at all, beyond the fact that they were so very normal.
Virtually everything Trump says or does seems so beyond the pale that it becomes difficult for most to imagine that historical patterns may apply. Given the extraordinary context leading into the 2018 midterms, it may seem inconceivable that they yielded perfectly
ordinary results. Consider Trump’s
historic unpopularity, his passionate opposition, our unprecedented levels of political polarization, the approaching migrant caravan, the mass shooting at a synagogue just before the midterms, the ongoing
Mueller investigation and myriad other scandals. Surely these must
matter , right?
Truth told, elections are complex social events, and it is difficult to determine (let alone predict) what matters, how much it matters and in what sense it matters. We’re
still arguing over what happened in 2016! Yet one thing we
do know is that the 2018 election results were consistent with the norm for a ruling party’s initial midterms. This reality should make Democrats deeply anxious because, as I’ve demonstrated elsewhere, if the 2020 presidential election similarly conforms to historical tendencies, the odds are roughly 8 to 1 that
Trump wins reelection.
Indeed, the president’s inaugural midterm results are eerily similar to those of another entertainment-star turned political game-changer: Ronald Reagan. In 1982, his party lost 26 seats in the House — but picked up one seat in the Senate. He, too, faced a split Congress. His approval rating going into those midterms was also in the
low 40s. He went on to win reelection
by a landslide in 1984.
