Donald Trump’s weekly approval rating in Gallup polling last week sunk to his lowest average yet, with just 35% of adults giving him positive marks for his performance as President. He hasn’t averaged 40% approval over a full week in the Gallup daily surveys since May – and hasn’t even cracked 40% on any single day since July 11. In polling dating back to Harry Truman’s administration, every president except one has averaged at least 57% approval in Gallup surveys over their first year; the sole exception was Bill Clinton, who averaged 49% in 1993. Trump’s average for 2017, Gallup recently reported, has dipped to 39%, and the trend line is pointing down. In other words, maybe Trump has his opponents just where he wants them. At least, that’s the contention of some political observers, not just on the right, who are cautioning that Trump’s historically weak approval numbers don’t necessarily mean he’s doomed to drag down Republican candidates in 2018 or face long odds for re-election should he seek it in 2020. For these analysts, the fact that Trump won in 2016, even though most voters said in Election Day exit polls that they viewed him unfavorably and doubted he was qualified for the job, demonstrated a unique capacity to garner support even from Americans deeply ambivalent about him. “By any historical standard, these were also politically catastrophic numbers, and yet, well, Trump is in the White House,” one Trump critic, Steve Kornacki, wrote recently. “In 2016, the numbers didn’t mean quite what we thought they did. As narratives of collapse take shape around Trump’s presidency now, the campaign should at least serve as a cautionary tale. It may look like his base is crumbling — and maybe it is — or maybe we’re living through a new version of what happened last year.” Another Trump critic, Damon Linker, recently wrote that while his presidency could implode in many different ways, “he might also survive, stick around, run, and win another term in 2020 — even with historically low approval numbers.” The importance of presidential approval ratings After Trump’s upset victory in 2016, it’s surely good advice to never say never when it comes to his political future. But it’s also important to understand just how much it would rewrite the political trends of the past three decades for Trump to avoid hurting Republicans in 2018 or to capture a second term in 2020 if he cannot improve his anemic approval ratings. Presidential approval ratings have become arguably the single most important predictor of election outcomes, both in midterm and re-election campaigns. In the increasingly parliamentary environment of modern politics, more voters are treating elections as a referendum on the party in power, which they associate mostly with the side holding the White House. In exit polls dating back to the 1980s, overwhelming majorities of voters who approve of the president’s performance have voted for his re-election and his party’s congressional candidates, and comparable majorities of those who disapprove have voted against him and his party. “There is a very close relationship between an incumbent president’s approval rating and his vote share,” said Emory University political scientist Alan Abramowitz. “No incumbent since World War II has been re-elected with an approval rating below 50%.” Two polling sources track the relationship between attitudes toward the president’s performance and voters’ behavior. The national exit polls conducted by Edison Research for a consortium of media organizations (including CNN) have asked voters whether they approve of the president’s job performance in each of the past two presidential re-election campaigns: George W. Bush in 2004 and Barack Obama in 2012. The University of Michigan’s American National Election Studies, a post- election survey, has asked the question for many years before that. Both of these polls point to the same conclusions. First, not many people approve of a president’s performance but vote against him because they like the alternative more. And second, even fewer dislike a president’s performance and vote for him anyway because they like the alternative less. Presidents don’t normally get votes from those who disapprove of them Since Ronald Reagan’s re-election in 1984, no incumbent has won the vote of more than 9% of the voters who disapproved of his performance, the NES surveys have found. Put another way, in every election since 1984, at least 91% of those who disapproved of an incumbent president’s job performance voted against his re-election. In the three re-election campaigns of the 1980s and 1990s – Reagan in 1984, George H.W. Bush in 1992 and Clinton in 1996 – the incumbents could not match that level of support from voters who approved of their performance, particularly when Ross Perot offered a third party alternative. Reagan carried 87% of his approvers, George H.W. Bush just 69%, and Clinton 79%, the NES found. But in both 2004 and 2012, the NES studies found, at least 90% of the voters who approved of the job performance of W. Bush and Obama voted for his re-election – while at least 90% of those who disapproved voted for his opponent. (The exit polls produced almost identical results for those years.) These results suggest that, if anything, the relationship is tightening between attitudes toward a president’s performance and his re-election vote. That link has been especially powerful among voters who strongly approve or disapprove of the president’s work. That trend is particularly relevant for Trump because an unprecedented share of the voters who disapprove of his performance say they do so strongly: in a late August Quinnipiac University national survey, for instance, 54% of adults said they strongly disapproved of Trump’s performance, while just 4% somewhat disapproved. (Only half as many, 27%, said they strongly approved of his performance, while 8% somewhat approved.) Voters who express such strong emotions about a president’s performance have broken even more decisively during his re-election. In the NES studies, 99% of the people who strongly approved of Obama’s performance voted for him in 2012, while 96% of those who strongly approved of Bush’s backed him in 2004. (The exit poll results again were virtually identical). Of more relevance for Trump may be the behavior of voters who strongly disapprove of a president’s record. According to the NES results, in 1992, H.W. Bush won just 5% of those who strongly disapproved of his job performance; in 1996, Clinton carried 2%; in 2004 and 2012, Bush and Obama both carried 4%. (The exit poll gave Obama only 1% of his strong disapprovers and Bush 2%.) These are daunting precedents for Trump given how many of his critics say they strongly disapprove of him. “We heard a lot of people say in 2016 ‘he doesn’t mean all those things he is saying – he’ll be better than that, he won’t really want to build a wall, he won’t really want to deport them,’” said Joel Benenson, Hillary Clinton’s chief strategist in 2016. “The problem is a lot of voters did not take him literally but now they are taking him seriously because he’s president. That’s why the strong disapproval is so high, because the people in that election said ‘he doesn’t mean all that.’ Now what they are seeing in the last seven months, it’s real. His behavior, his character are coming through and it is turning off a lot of voters.” A test in 2018 The clouds over Trump also loom over 2018. In every midterm election since 1994, at least 82% of voters who disapproved of the incumbent president voted against his party’s candidates in elections for the House of Representatives, according to the exit polls. Over that same period, at least 82% of those who approve of a president’s performance voted for his party’s candidates, except in 1998, when 77 percent did, the exit polls found. (No exit poll results were released in 2002). Despite some conspicuous exceptions, even Senate elections have largely followed similar tracks. In 2014, for instance, Republicans won 14 of the 15 Senate races in states where exit polls placed Obama’s approval rating at 42% or below; in 2006, Democrats won 19 of 20 Senate races in states where Bush’s approval rating stood at 45% or below. There are reasons this past might not be prologue. Veteran Republican pollster Bill McInturff notes that despite Trump’s historically meager overall approval rating, his numbers among Republicans remain competitive with other presidents in their own party at this point in their administrations. That support, he said, should help Republican House and Senate candidates running this fall on conservative leaning terrain. “Because of redistricting, Americans live in two incredibly separate worlds, Republican [Congressional Districts] versus Democratic CDs,” McInturff said. “To sustain a majority in the House, tracking Trump’s approval in GOP CDs may be more important than … Trump’s low approval continuing to erode in Democratic CDs.” Abramowitz said that historically enough voters have been willing to give an incumbent the benefit of the doubt to provide a small “boost” for his reelection. Combined with a strong economy, he said, that might make it “possible for Trump to win a second term even with relatively weak approval numbers.” These are all important considerations. And yet the president’s party has consistently faced significant losses in midterm elections when his approval rating has lagged. And even the incumbency boost hasn’t been enough to deliver a second term to presidents whose approval ratings fell too low. While Trump still has much time to recover, his disapproval rating in the daily Gallup poll has already spiked to peaks at times that exceed the highest levels ever confronted by Jimmy Carter or George H.W. Bush, the only first-term presidents defeated for re-election since World War II. For Trump, his opponent is key From all points on the spectrum, observers agree one reason Trump won despite his own poor numbers is that he faced in Hillary Clinton an opponent who was also uniquely unpopular. That points toward what may be the most likely implication of Trump’s lagging approval ratings: if the White House and Republicans can’t build up the president, it will increase the pressure on them to bring down his Democratic opponents, starting in 2018 and then in 2020 if he runs again. “If he can’t dramatically turn these numbers around, which I don’t think he has the capability to do,” said Benenson, “their only strategy is going to be to completely try to tear down and destroy their opponent.” Given the intensity of feeling Trump has already provoked, seeding more doubts about his rivals in both parties may be a more realistic mission for him than easing the doubts about his own priorities and performance.