A “Referendum” on Trump?

UPDATED 11/20/18

I was wrong when I predicted (here and here) that the Dems wouldn’t retake the House. They did, and in rather convincing fashion. Including three undecided races (two GOP leaders, one Dem leader), the Dems will have gained 39 seats. That’s a pretty good showing by historical standards:

So, was 2018 a “wave election” for Democrats, as many Democrats and pundits expected it to be? I don’t know, and neither does anyone else. “Wave” is vague. (Pun intended.)

But I will say this: If there had been something like a “wave”, the Dems would have retaken the Senate, too. They certainly were in a position to do so, needing only a net gain of only two seats to go from a 49-51 minority to a 51-49 majority. (I’m counting two so-called independents as Dems because that’s how they vote.) But, instead, the GOP added two seats (assuming that Sen. Hyde-Smith of Mississippi retains hers in next week’s runoff), and will open the 116th Congress with a 53-47 majority.

Here’s how 2018 stacks up against previous mid-term results for the Senate:

Why the disparity between the House and Senate in this year’s mid-terms?

If the mid-terms had been a “referendum” on Trump, as often suggested (even by Trump), the House and Senate would have gone in the same direction. They didn’t because the election wasn’t entirely about Trump.

One story is that the Dems gained in the House because of a focus on health care. It is the Dems’ own creation — Obamacare — that has pushed health-care costs and premiums ever higher in recent years. But that matters not to ignorant voters, who are sucked in by promises to “get it right”. Dems are good at making such promises.

The Senate results tell a different story. It was possible for Trump to lend visible and vocal support to GOP Senate candidates in a way that he couldn’t for the vastly greater number of GOP House candidates. So, if anything, the “referendum” on Trump occurred in Senate races, and Trump won.

Trump and Election 2018

This was a final update before election day. The Dems won a majority in the House, though  a narrow one. Meanwhile, the GOP has increased its majority in the Senate. That is the better half of the loaf because control of the Senate means that Trump can continue to remake the judiciary in a conservative image. Further, the House will be perceived as the obstructionist body for the next two years, setting the stage for a GOP restoration there. Barring the unforeseeable, a largely successful Trump presidency will set the stage for Republican dominance in 2020.

How is Trump’s popularity these days? And how will his standing with voters affect the outcome of tomorrow’s elections?

Trump’s approval ratings have been fairly steady since early in the year, with a recent uptick that bodes well for GOP candidates:

FIGURE 1
Derived from Rasmussen Reports approval ratings for Trump.

Lest you believe that those numbers are weak, consider this comparison with Obama’s numbers:

FIGURE 2
Derived from Rasmussen Reports approval ratings for Obama and Trump.

In this age of polarization, it’s hard for any president to routinely attain high marks:

FIGURE 3
Source: Same as figure 2.

The good news, again, is that Trump’s strong approval rating has been significantly higher than Obama’s for the past several months.

Ratios of the ratios in figure 2 yield enthusiasm ratios: the strength of strong approval ratings relative to overall approval ratings:

FIGURE 4
Source: Same as figure 2.

Since the spike associated with the Singapore summit, Trump”s enthusiasm ratio has settled into a range that is comfortably higher than Obama’s.

There is a different poll that is more revealing of Trump’s popularity. Every week since the first inauguration of Obama, Rasmussen Reports has asked 2,500 likely voters whether they see the country as going in the “right direction” or being on the “wrong track”. The following graph shows the ratios of “right direction”/”wrong track” for Trump and Obama:

FIGURE 5
Source: Rasmussen Reports, “Right Direction or Wrong Track“.

The ratio for Trump, after a quick honeymoon start, fell into the same range as Obama’s. But it jumped with the passage of the tax cut in December 2017, and has remained high since then, despite the faux scandals concocted by the leftist media and their concerted attack on Trump.

Figure 5 suggests that the squishy center of the electorate is lining up behind Trump, despite the incessant flow of negative “reporting” about him and his policies. (See “related reading” at the end of this post.) His base is with him all the way.

Trump’s coattails may be be decisive in November. Based on an analysis of the relationships between Obama’s popularity (or unpopularity) and the outcome of House elections, it looks like the GOP will hold the House while losing about 10 seats. (This is a very rough estimate with a wide margin of error.)

Rasmussen’s generic congressional ballot affords a similar view. The polling data, which are behind a paywall, span April 2007 to May 2015 (when the poll was discontinued), and January 2018 (when the poll was resumed) to the present.

This graph compares the polling results to date with the actual nationwide vote shares compiled by House candidates in the general elections of 2008, 2010, 2010, and 2014:

FIGURE 6

Taking a closer look:

FIGURE 7

Rasmussen advertises a 2-percentage-point margin of error, which is borne out by the results for the elections of 2008-2014. In fact, the generic congressional ballot was spot-on in 2010 and 2012, while the GOP under-performed slightly in 2008 (the year of the financial crisis) and over-performed slightly in 2014 (a mid-term referendum on Obama).

Let’s say, for the sake of argument, that this year’s polls are spot-on. The latest poll (figure 7) gives the GOP 51 percent of the two-party vote. How does that translate into House seats? Recent history is probably the best guide:

FIGURE 8

A 51-percent share of the vote would give the GOP about 52 percent of House seats; that is, the GOP would hold the House. In fact, two years ago the GOP won more than 55 percent of House seats with 50.5 percent of the two-party vote.

There’s more evidence against the loss of the House:

1. I correlated measures of Obama’s popularity (or lack thereof) with with the outcomes of House mid-terms during his presidency. I then applied those correlations to measures of Trump’s popularity (or lack thereof), which is markedly higher than Obama’s at this stage of their respective presidencies (according to Rasmussen, at least).

2. I correlated the outcomes of post-WWII mid-terms during GOP presidencies with the GOP presidents’ shares of the 2-party vote in the preceding elections. I then applied that result to Trump’s share of the 2-party vote in 2016.

Both methods yield the same result for 2018: a loss of 4 House seats by the GOP (yes, four seats, not 4 percent of seats). The estimates are surrounded by a wide margin of error. Given that, the results support the view that the GOP will hold the House.

In the end, the outcome will depend on turnout. Are Democrats more charged up than Republicans? I don’t think so.

Stay tuned.


Related reading:

Alex Castellanos, “How Trump Has Managed to Defy Gravity“, RealClearPolitics, July 31, 2018

Selwyn Duke, “Media Collusion: 100-Plus Papers Agree to Simultaneously Run Anti-Trump Editorials“, The New American, August 14, 2018

Dennis Prager, “The Greatest Hysteria in American History“, RealClearPolitics, July 24, 2018

Ned Ryun, “None Dared Call It Treason … When It Was a Democrat“, American Greatness, July 24, 2018

What Happens Next?

In what direction will leftists move after next month’s elections?

1. Will they be chastened by defeat and tone down their verbal and physical violence?

2. Will they be frustrated by defeat and keep it up?

3. Will they tone it down if broadly victorious?

4. Will they keep it up if broadly victorious, in the belief that more of the same will yield more of the same.

5. Or will they just keep it up because leftism and violence are intertwined?

The left isn’t monolithic, of course. The loony-left fringe — which is much more than a fringe — will choose #5.

Most left-wing politicians, on the other hand, will probably tone it down — win, lose, or draw. That is, they will choose #1 or #3.

I expect that it will come down to #1. This is from a recent article (behind a paywall) at Rasmussen Reports:

Republicans are madder about the Kavanaugh controversy than Democrats are and more determined to vote in the upcoming elections because of it.

A new Rasmussen Reports national telephone and online survey finds that 54% of all Likely U.S. Voters say they are more likely to vote in the upcoming midterm elections because of the controversy surrounding President Trump’s U.S. Supreme Court nominee. Only nine percent (9%) say they are less likely to vote. Thirty-four percent (34%) say the controversy will have no impact on their vote.

Sixty-two percent (62%) of Republicans are more likely to vote because of the Kavanaugh controversy, compared to 54% of Democrats and 46% of voters not affiliated with either major political party.

Sixty-two percent (62%) of all voters are angry about the U.S. Senate’s treatment of Kavanaugh, with 42% who are Very Angry. Fifty-six percent (56%) are angry about how the Senate treated Christine Blasey Ford, the woman who accused Kavanaugh of sexual assault, including 35% who are Very Angry.

Sixty-four percent (64%) of Republicans are Very Angry about the Senate’s treatment of Kavanaugh, a view shared by 30% of Democrats and 34% of unaffiliated voters. By comparison, fewer Democrats (48%) are Very Angry about the Senate’s treatment of Ford; 28% of GOP voters and 30% of unaffiliateds agree.

Democrats’ five-point lead on the weekly Rasmussen Reports Generic Congressional Ballot has vanished. The two parties are now tied with less than a month until Election Day. We’ll be watching to see if this is the beginning of a post-Kavanaugh trend.


Related posts:

Trump Defies Gravity
The House Hangs in the Balance