By Jim Ellis
Professor Larry Sabato, University
of Virginia
Dr. Sabato supports his claim for several basic reasons. First, he sites the historic trends that a new president’s party loses seats in the first midterm election, and traces this electoral pattern all the way back to the Civil War era. Second, he turns to the typical polling regularly released that places President Trump’s approval ratings in what he terms “the low 40s”, and includes the generic House ratings, along with the “enthusiasm” analysis. Third, is the Democrats’ record in the current cycle’s federal and state special elections, and fourth is their second quarter fundraising “advantage.”
There are counter arguments that need mentioning for each of these points.
It is questionable to compare electoral trends developed during the 1800s to the elections of today because the world has changed so much. Bringing the analysis to at least the 20th Century and looking just at the post-World War II patterns (from President Harry Truman, inclusive, to today), we find that the average seat loss in the House during a new president’s first midterm is 26 seats. But, this average combines the six Democratic presidents losing 32 seats, and the five Republicans’ dropping 15 districts. Just three elections, 1966 (Johnson; -47 seats), 1994 (Clinton; -54), and 2010 (Obama; -63) have substantially upped the overall average.
Additionally, President Trump’s approval is actually trending up, in many polls beyond 45 percent, which are the best ratings of his tenure. It is important to remember that his approval rating on the day he won the 2016 election was between 38 and 40 percent in national polling. Therefore, it is worthy asking the question about whether applying the Trump numbers to election projection is wise because there is good reason to believe to believe his performance is an anomaly and not a standard.
As we have discussed in other daily Updates, the congressional generic rating attempts to extrapolate a polling sample of usually 1,100 respondents or so, and apply it to 435 individual House campaigns. Even when Republicans have won in wave years: 1994, 2010, 2014, they were behind on the generic vote.
In terms of the Democrats’ strong record in special elections, it is important to remember that we have seen eight US House special elections decided during this election cycle and Republicans have won six. The two Democratic victories come in California’s 34th District in early 2017 (Rep. Jimmy Gomez defeating a fellow Democrat in the run-off), a highly Democratic Los Angeles County district, and in Pennsylvania.
In the latter case, Rep. Conor Lamb (D-Pittsburgh) won a western state district by a 755-vote victory (less than one percentage point when polling predicted a Democratic victory of between two and six points) in a conservative union district against a Republican candidate who voted for the state’s Right to Work law. Furthermore, Lamb said he would not support Democratic Leader Nancy Pelosi (D-CA) for Speaker, and refused to campaign as a national Democrat.
The argument that Democrats over-performed in these districts even when losing largely comes from comparing the Democratic congressional result to the Trump-Clinton campaign result and not historical data. As mentioned above, it is more likely that the Trump-Clinton results will prove to be an anomaly rather than a standard.
The Democrats have performed well in special elections in various states for the state legislature. But as Dr. Sabato points out, while certainly a good sign for them, it doesn’t necessarily serve as a future predictor. He does not mention a polling factoid from around the country that shows strong Democratic performance in the Independent category, and is likely a stronger indicator that the party is on the uptick for the coming general election.
In the polling analysis realm, Dr. Sabato again refers to the Democratic “enthusiasm advantage” as detected in most surveys. But the primary turnouts so far recorded in 31 states, which actually records how people vote as opposed to poll responses, shows a normal turnout pattern. Democrats have turned out better in Democratic states, but more Republicans are voting in the states that usually vote their way. In many key targeted districts, six of seven in California, for example, and the three key Texas targets where both parties had primaries, more Republicans than Democrats voted. In the 13 New York contested Democratic primary congressional contests, average party turnout was just 11.4 percent.
In the coming OH-12 special election where Dr. Sabato details Democratic chances of winning what should be a reliable Republican seat, he fails to mention, while citing a Democratic enthusiasm advantage, that Republicans turned out 67,847 voters in the May special primary election compared to the Democrats’ 43,945.
Finally, he makes mention of the Democrats’ “extraordinary” aggregate second quarter fundraising totals, which are impressive. And, he questions whether the Republican incumbent in his 24 races that he lists as either “Lean Democratic” or “toss-up” will even have a financial advantage in the fall. But, he fails to mention that while Democrats may have raised more money in the second quarter, in almost every one of these instances the Republican incumbent still has more cash-on-hand.
Finally, 36 seats reside in his toss-up category, while Sabato has an additional net six districts moving clearly into the Democratic column. But in order for the Democrats to take an even one-seat edge in the new House, they would have to win a clear majority in 38 of his toss-up or lean D districts where any Democratic candidate has won only four times in the aggregate 114 elections since these seats were created in their current configurations from the 2011 redistricting process.
