Will 2018 Be a Wave Election?

Will 2018 Be a Wave Election?
AP Photo/Wilfredo Lee
Will 2018 Be a Wave Election?
AP Photo/Wilfredo Lee
Story Stream
recent articles

If we define off-year “wave” elections as a loss of at least 30 seats for the incumbent presidents’ political party, there have been nine such elections since 1932 – though in five of them control of Congress did not change hands. But Democrats don’t need 30 seats this year; they only need 23. 

Democrats controlled the House for 60 of the 64 years between 1930 and 1994. In the current era of party parity, there have been three wave elections, two Republican (1994, 2010) and one Democrat (2006). In all three, the House did change hands. 

In the two GOP wave elections, Republican voters of all stripes -- strong, weak and independent-leaning -- turned out at higher levels than comparable Democrats, which mitigated the Democratic Party’s advantage in party registration. 

Table 1 shows the results of turnout in wave elections. The numbers indicate the percentage of voters who identified as strong or weak in party identity, or as independent-leaning, who cast votes. Although in 1994 and 2010, Republicans across all categories turned out in higher numbers than Democrats, this was especially true in 2010. The higher turnout is crucial for Republicans since in both elections they had to overcome the Democratic advantage in numbers (46 percent to 42 percent of registered voters in 1994 and 46 percent to 39 percent in 2010). In the Democratic wave of 2006, the percentage of Democratic and Republican turnout was about even but there were again more registered Democrats than Republicans (47 percent to 41 percent), which gave Democrats the advantage.


David Brady/Brett Parker

Another factor aiding the GOP in trying to make up for its lower numbers relative to Democrats is that, on average, Republicans split their tickets less often than do Democrats. In 1994, about 30 percent of weak and leaning Democrats voted Republican while only a little over 20 percent of similar Republicans voted Democratic. In 2010, 20 percent of weak Democrats voted Republican, as did 10 percent of leaning Democrats, and these numbers doubled the number of similarly categorized Republicans voting Democratic. In the 2006 Democratic wave, weak and leaning Republicans voted Democratic at levels of 25 percent and 16 percent, respectively, while weak and leaning Democrats were more loyal to their party candidates, thus adding to the Democrats’ numerical advantage. 

Also in that 2006 election, there was little or no turnout advantage, making the Democrats’ registered-voter numbers advantage work for them as well as the fact that Democrats voted the party line more than did Republicans. 

This analysis leaves out independents, who turn out in lower numbers than partisans but are still sufficient in number to tilt an election. In the 1994 wave, independent turnout was 34 percent, and they split their votes 50/50. In the 2010 Republican wave, their turnout numbers fell slightly to 33 percent, but of those who turned out, slightly over 60 percent voted Republican. 

The 2006 midterms were a high point for independent turnout at 44 percent, and 62 percent voted Democratic. In sum, in wave elections the ascendant party turns out more of the vote -- and those turning out are loyal to the party -- and independents either don’t affect the result (1994) or add to the party wave. 

Democrats have gone into each post-1992 off-year election with a numerical advantage, which Republicans have to mitigate by enhanced turnout and greater straight party voting. Democrats, on the other hand, only have to approximate Republican turnout and party voting levels to win elections. Independents can add to the wave, as they did in 2006 and 2010, by voting for a blue or red wave, or they can split their vote as they did in 1994. In September of each of these wave elections, the president’s approval rating was below 50 percent and, in George W. Bush’s case, below 40 percent. Moreover, in two of the three (2006 and 2010), the economy was in decline. 

Given these scenarios, what does 2018 look like for the Democrats, since no one (other than Newt Gingrich) thinks that 2018 will be a red wave election? As of this writing, President Trump has a RealClearPolitics poll average approval rating of 44.2 percent, putting him lower than Bill Clinton in 1994 or Barack Obama in 2010 but above Bush in 2006. A look at the YouGov weekly polls shows that over the last two months, Democrats outnumber Republicans by between 5 percent and 6 percent, and from August 1 on, Democrats have led Republicans in the generic vote by about 45 percent to 40 percent, with the current breakdown by party strength shown in Table 2.

David Brady/Brett Parker

These numbers for Democrats are approximately equal to their 2006 figures, while the Republican numbers are below theirs in the 2010 and 1994 wave elections.  Across every strength category, Democrats are signaling that they will vote Democratic for Congress at higher rates than Republicans. The Democratic advantage in numbers, and their September vote intention, are good portents for the party. 

Time and events before November could change these indicators. However, the odds of the generic ballot and party identification shifting to the Republicans between now and Election Day are not good. One group that could affect these results is independents, who vote less than partisans and make up their minds later in the campaign. In the latest poll, independents lean 28 percent to 26 percent for Democrats, with 29 percent saying not sure; if these percentages carry over till November the Democrats’ numerical advantage will prevail. 

Looking at 2018 less than two months before the election gives the Democrats the advantage in numbers, vote intention and a small lead among independents, all of which point to a Democratic takeover of the House of Representatives. The missing variable is, of course, turnout. 

If more Republicans turn out in greater-than-expected numbers, the Democrats’ advantages are diminished. We asked two questions that attempt to discern the likely turnout levels. The first asks respondents whether, compared to previous congressional elections, they are more or less enthusiastic; the second asks flatly if they plan to vote in this year’s elections. So far, Republicans score a bit higher on the enthusiasm measure while more Democrats say they will definitely vote in 2018. 

One event which could affect the outcome is the contentious Brett Kavanaugh Supreme Court nomination, especially regarding turnout rates for women. Trump has problems with independents and women, as we noted in a March 22, 2018, RealClearPolitics piece

In the most recent YouGov poll of September 24, 2018, college-educated women favor Democrats in the generic ballot by slightly over 30 percentage points, 57.3 to 27.2, and over three quarters are committed to voting. In short, while the FBI continues to investigation this nominee, women and independents already have doubts about Trump and Republican candidates. Depending on what happens next, those numbers could get worse.

David Brady is a professor of political science at Stanford University and the Davies Family Senior Fellow at the Hoover Institution.

Brett Parker is a JD/PhD student at Stanford University.

Show comments Hide Comments