Skip to main content
Menu
Trump Won Despite Being Unpopular, So Can He Govern That Way?

President-elect Donald Trump was the least-liked major-party presidential nominee since at least 1980. He won the Electoral College even though most Americans held an unfavorable view of him — the first time that’s happened since pollsters first began consistently asking the favorable/unfavorable question. That raises a reasonable set of questions: Are favorability ratings meaningless? And if Trump’s favorability rating didn’t matter, will his job-approval rating — the benchmark by which modern presidents are judged — have any effect on his ability to govern or enact his agenda?

These are complicated questions with complicated answers that we’re just beginning to untangle. But it would be wrong to look at the 2016 election results and conclude that favorability ratings are irrelevant. Trump actually did about as well nationally as you’d expect, given his and Hillary Clinton’s favorability ratings. She was a little more popular than he was and she will probably win the national popular vote by a couple of percentage points. In state after state, people who had a favorable view of Clinton generally voted for her, and people who had a favorable view of Trump generally voted for him.

But here’s the deciding factor: The group that made the difference turned out to be people who disliked both candidates. They swung toward Trump, giving him the White House.

The CBS News/New York Times poll has been asking about presidential candidates’ favorability ratings in the fall of election years since the 1980 campaign. From September to November, Trump averaged a minus-24 percentage point net favorability rating (the percentage of people who view him favorably minus the percent who view him unfavorably). That’s bad. The lowest net favorability rating before Trump’s was George H.W. Bush’s minus-16 percentage points in the fall of 1992, and Bush lost re-election. No winning candidate before Trump ever had a negative net favorability rating, according to the CBS News/New York Times poll.

YEAR PARTY NOMINEE NET FAVORABILITY
1980 Republican Ronald Reagan +7
1980 Democratic Jimmy Carter -7
1984 Republican Ronald Reagan +24
1984 Democratic Walter Mondale -6
1988 Republican George H.W. Bush +13
1988 Democratic Michael Dukakis -7
1992 Democratic Bill Clinton +1
1992 Republican George H.W. Bush -16
1996 Democratic Bill Clinton +13
1996 Republican Bob Dole -12
2000 Republican George W. Bush +12
2000 Democratic Al Gore +10
2004 Republican George W. Bush +5
2004 Democratic John Kerry -7
2008 Democratic Barack Obama +14
2008 Republican John McCain -2
2012 Democratic Barack Obama +3
2012 Republican Mitt Romney -1
2016 Democratic Hillary Clinton -16
2016 Republican Donald Trump -24
Clinton and Trump were historically unpopular

Average poll from September to Election Day in the year of the election.

Source: CBS News/The New York Times

But what you’ll notice in the table is that Clinton was the second-most-disliked candidate, and elections are choices. In the end, Clinton was a smidge more popular than Trump. It was close enough that it was not determinative. Historically, a candidate who has an 8-point edge on net favorability in the CBS News/New York Times poll would be forecast to win by 3 points nationally. That’s about the size of Clinton’s “win.”

enten-trumpfavorability-1

This generally matches what I wrote in September. At that time, and even on Election Day, Clinton was seen as somewhat less unpopular than Trump. (Her edge on net favorability back then forecast a 4-point victory, which is pretty close to how it ended up in the popular vote.)

Of course, American elections are fought for state electoral votes. And it’s on the state level that you can see how Trump overcame his national favorability deficit.

In many states, the favorability gap was close and the final election result was close. Clinton and Trump had identical favorable ratings in Pennsylvania, for example, so Trump’s small win there matches up well. Trump also won pivotal states such as Florida, Michigan, North Carolina and Wisconsin, even though Clinton was better liked there. But Clinton’s advantage in favorability in most of those states was 4 percentage points or less. Given that the final margin in all these states was relatively close, and polls are inherently imprecise, Trump’s wins didn’t really break some iron-clad favorability rule.

Wisconsin, though, is a different story. Clinton’s favorable rating there was 42 percent compared to Trump’s 35 percent. That 7-point gap in favorable ratings (and 15-point gap in net favorability) “should” have resulted in an easy win for Clinton. Instead, she lost the state by 1 point. What happened? It’s not the case that voters who liked Clinton cast a ballot for Trump, or vice versa. In Wisconsin, Clinton won 97 percent of voters who viewed her favorably. Trump won 98 percent of voters who viewed him favorably. The answer: Trump won the 22 percent of voters who held an unfavorable view of both candidates. In fact, he won them by nearly 3 to 1. And that pattern was true nationally and in almost every state.

TRUMP’S MARGIN AMONG VOTERS WHO RATED …
STATE ONLY CLINTON FAVORABLE ONLY TRUMP FAVORABLE BOTH UNFAVORABLE
Kentucky -98 +100 +46
Georgia -98 +97 +39
Florida -98 +98 +37
Texas -99 +97 +37
Wisconsin -98 +99 +37
N. Carolina -97 +97 +36
Missouri -96 +98 +35
Indiana -97 +92 +32
Ohio -98 +97 +30
S. Carolina -100 +99 +30
Pennsylvania -98 +98 +25
Nevada -97 +95 +23
Iowa -96 +98 +21
Michigan -98 +97 +21
Minnesota -98 +98 +20
Utah -92 +95 +19
Arizona -97 +96 +17
Colorado -97 +96 +13
Virginia -97 +98 +12
New Mexico -94 +96 +10
N. Hampshire -97 +94 +9
Washington -98 +97 +5
New Jersey -93 +96 +3
Maine -97 +99 +2
Oregon -97 +98 0
Average -97 +97 +22
Trump won big among voters who disliked him and Clinton

Sample sizes were not large enough to compute for California, Illinois or New York.

Source: Edison Research

On average, Trump won the pox-on-both-their-houses vote by 22 percentage points. So while Clinton was moderately more popular than Trump overall, she was apparently a less-appealing option for voters who disliked both their choices. In other words, there’s no great favorability mystery here.

Once Trump becomes president, the main measure the press and American public will use to gauge his performance will be his job-approval rating. And this is where his popularity could matter. As John Sides of The Monkey Cage blog noted back in 2013, presidents have a better chance of getting Congress to pass favored legislation when they are more popular. Political science research has found a relationship between a president’s approval rating and whether the House approves high-profile bills that get a lot of news coverage (as opposed to run-of-the-mill, non-controversial legislation). A Trump immigration plan would fall into that category. Another study also found that Congress has been more willing to rubber-stamp a president’s agenda when his approval rating has been higher.

For now, though, Republican lawmakers seem united behind Trump despite his limited popularity. Although many Democrats have criticized Trump’s selections of Steve Bannon as chief strategist and Sen. Jeff Sessions as attorney general, Republicans are mostly standing by the president-elect. It could be that in a highly partisan era, job-approval ratings matter far less than party labels in terms of securing support from Congress. As Sides pointed out, there is some support for that thesis in a paper from political scientists Jon Bond, Richard Fleisher, and B. Dan Wood. If that’s the case, Trump may be able to ride partisanship to legislative success, at least initially. Republicans do, after all, have a large majority in the House and a small majority in the Senate.

Still, it’s probably not helpful for Trump if he remains unpopular. Every House member is up for re-election in 2018 and there are still some Republican senators from purple and blue states. These legislators may still be influenced by Trump’s standing with the American public. According to Gallup, Trump’s net favorability this past week was only minus-13 percentage points. That’s up from where it was before the election, but it’s still really low. Indeed, Trump’s net favorability is lower than any president-elect (or incumbent president after re-election) since at least 1980.1

YEAR PARTY PRESIDENT NET FAVORABILITY
1980 Republican Ronald Reagan +26
1984 Republican Ronald Reagan +40
1988 Republican George H.W. Bush +27
1992 Democratic Bill Clinton +20
1996 Democratic Bill Clinton +37
2000 Republican George W. Bush +18
2004 Republican George W. Bush +9
2008 Democratic Barack Obama +43
2012 Democratic Barack Obama +13
2016 Republican Donald Trump -13
Trump has an unusually low post-election favorability rating

First CBS News/New York Times poll taken after Election Day, except in 1996 and 2016 when no CBS News/New York Times poll was available

Sources: CBS News/The New York Times, Gallup, Time/CNN/Yankelovich Partners

No previous winner had a negative net favorability rating. The closest was George W. Bush in 2004, who had a net favorability rating of +9 percentage points. Bush went on to face major struggles in his second term. Trump’s lack of popularity stands in direct contrast to his soon-to-be predecessor, President Obama. Obama in 2008 was the most popular incoming president since at least 1980. He was even more popular than Ronald Reagan was after the 1984 election, when he won 49 states. (Obama’s occasional bursts of popularity, however, didn’t seem to help him win legislative victories once Republicans took control of the House in 2010.)

The good news for Trump is that his approval rating after his inauguration will probably be higher than his favorability rating is now. Although we’re working with a small sample size — just nine elections since 1980 — you get a pretty good idea of what a president’s first net approval rating from Gallup will be by knowing his net favorability rating after being elected, and whether he was an incumbent.

enten-trumpfavorability-2

For re-elected presidents, there hasn’t been a big difference between their favorability rating after election and approval rating after taking the second oath. People know what they are getting when they re-elect a president. There isn’t much room to be surprised. Presidents sometimes get a popularity boost after being re-elected, but they don’t get another one after inauguration.

New presidents, however, tend to get a big bump from their post-election net favorability rating — about 10 to 20 percentage points. That makes sense. Whether because of Americans rallying behind their leader or just giving the new guy a chance, the honeymoon period is a clearly defined phenomenon in American politics. A 10 to 20 percentage point bump would land Trump in positive territory, or close to it, in terms of his first net approval rating upon entering office.

But make no mistake: Trump’s in a deep hole, and his atypical personality may make it difficult to climb out. Even a 20-point bump from Trump’s current net favorability rating to his first net approval rating would leave him with an opening net approval rating of +7 percentage points. That’s not only lower than any of the presidents studied here, it would be the lowest first net job-approval rating for any president since at least 1941, when Franklin Roosevelt entered his third term.

Trump will probably be hampered at least a little bit by his lack of popularity at the beginning of his term. He didn’t really defy his favorability rating during the presidential election, so there’s no reason to think he’ll be able to escape the normal effects of approval ratings. The more popular he is, the more likely he’ll be to have legislative success. Without popular support, however, he’ll likely encounter more pushback.

Footnotes

  1. In all possible cases, I’ve taken the first CBS News/New York Times poll taken after the election. Right after Bill Clinton’s re-election and also in this cycle, there was no CBS News/New York Times poll conducted between the election and inauguration. In those cases, I took the first available live-interview poll asking about favorability whose question mirrored most closely the way CBS News/New York Times asks about favorability (most importantly, explicitly giving respondents the option of choosing “no opinion”). For 2000, I took the first poll conducted after Al Gore conceded to George W. Bush.

Harry Enten is a senior political writer and analyst for FiveThirtyEight.

Comments