GOP midterm strategist asks: What does it mean to 'approve' of Trump?

Donald Trump is riding a wave of popularity, at least by Trump standards.

The president's job approval rating hit 44.6 percent this week in the RealClearPolitics average of polls. That is the highest it has been since March 2017. And while that is not high by any measure, it is good for Trump, who has never been higher than 46 percent in the poll average. (That was in early February 2017, his third week in office.)

The question is what effect Trump's improved ratings -- should they last until November -- will have on the midterm elections.

There is a traditional relationship between presidential job approval and midterm results. But it's not clear whether that relationship will hold up when Trump is involved. "When a president has job approval ratings of 50 percent or higher, his party tends to keep its losses fairly low," political analyst Charlie Cook wrote last year. "But in six of the seven midterm elections since 1966, when presidential approval ratings hovered below 50 percent, his party has lost two dozen or more seats in the House, giving the opposition party a majority the next year."

If that holds, and Trump remains below 50 percent, it's a reasonable guess the GOP will lose enough seats to give Democrats control of the House.

The last dozen years have seen extremes in midterms. In 2006, with George W. Bush at 39 percent job approval, Republicans got clobbered, losing 30 seats.

In 2010, with Barack Obama at 45.4 percent job approval, the damage was even worse: Democrats lost a devastating 63 seats in the House.

But then, in 2014, with Obama at 42 percent job approval -- below where Trump is today -- House Democrats, already in the minority, lost a modest 13 seats.

Does any of that experience help predict what will happen under Trump? On the one hand, it's easy just to say a president at his level of popularity will lose a bunch of seats. On the other hand, remember that Trump's personal approval rating was 37.5 percent, with a disapproval rating near 60, on the day he won the presidency.

"These are certainly different times," says Curt Anderson, a GOP strategist whose firm is involved in a lot of House races this year. One reason for Republican caution, Anderson explains, is that this year it will be easier for moderates to cast an anti-Trump vote than it was in 2016.

"In 2016, people who for whatever reason didn't like Trump had to swallow hard and vote for Hillary to show their displeasure," Anderson explains. "That's some nasty castor oil right there, and many refused to take it. The fear in 2018 for Republicans is that voters who don't like Trump can send him a message -- by voting against his party -- and this time they don't have to vote for Hillary in order to punish Trump."

Given that, Anderson says, "I do think Trump's approval numbers will matter this fall ... [and] already this year we have seen small shifts that have pretty dramatic consequences."

On the other hand, another GOP strategist working on multiple races, who asked to remain anonymous, takes a more nuanced, and ultimately more optimistic, position.

"Of course there's a correlation," he says. "The more popular [Trump] is, the easier it is to keep the House." Right now, things are going reasonably well, because there has been "an undeniable positive change" since December in Trump's standing in many key districts.

But the strategist notes that with Trump, not everything can be measured by job approval. "What does it mean to approve of Trump?" he asks. "You can love everything that is happening and not approve of him."

For the midterms, the strategist suggests, a better predictor might be the traditional polling question of whether the country is on the right track or the wrong track. "In our data, we have right track at 40 percent," the strategist notes, meaning that 40 percent of those surveyed say the country is going in the right direction. "If you look back, there's never been a wave election with right track at 40."

Indeed, back in November 2010 the right track number was 31 percent. In November 2006, it was around 30.

None of that tells what will happen in November 2018. But it's simply not enough to say that Trump is unpopular, and therefore Republicans will lose. Given the nation's experience in 2016, the presence of Trump, even though he is not on the ballot, makes the coming midterms more complex than midterms in the past, and extremely hard to predict.