Sabatos Crystal Ball

Extended term for Trump? No way, most Americans say

Large majority of Americans support respecting the peaceful transfer of power, according to new UVA Center for Politics-Ipsos poll

UVA Center for Politics May 16th, 2019


The lion’s share of Americans believes that respecting the results of elections and the peaceful transfer of power are essential elements of American democracy, according to a new national poll conducted by Ipsos in conjunction with the University of Virginia Center for Politics.

By a 77%-16% margin, respondents did not think that the 2020 election should be delayed and President Donald Trump given an extra two years in office. This question was based on a recent tweet by Liberty University President Jerry Falwell suggesting that because of Robert Mueller’s investigation into Russian interference in the 2016 election, the president’s term should be extended two years (Trump retweeted Falwell). There were partisan differences on this question: Democrats said no overwhelmingly, 89%-9%, while Republicans said no by a smaller 62%-31% margin.

Just 7% of respondents said that if Trump loses the 2020 election, he should ignore the results and stay in office.

Overall, about nine in 10 respondents said that the “peaceful handover of power after elections” and “both parties respecting the results of elections” were essential for the “healthy functioning of American democracy.”

A majority of respondents agreed with the statement that “Donald Trump does not respect the customs or traditions of the Presidency” (61% overall, and 86% of Democrats) and that he “does not respect the laws of the United States” (56% overall, and 83% of Democrats).

About two-fifths of all respondents (41%) — and over three-quarters of Republicans (77%) — agreed that a “deep state” is “trying to undermine Donald Trump’s presidency.” Also, 30% of all respondents and 60% of Republicans agreed that the “Mueller investigation prevented President Trump from achieving significant portions of his agenda.”

Therefore, as we would expect, Democrats were much likelier to agree that Trump does not respect the traditions of the presidency and the laws of the United States; Republicans were much likelier to agree that the “deep state” and the Mueller investigation have undermined Trump’s presidency.

Respondents were split, 41%-39%, as to whether Trump has achieved many of his goals in his first two years in office. Three-quarters of Republicans strongly or somewhat agreed that Trump has achieved many of his goals, while only about one-fifth (22%) of Democrats said the same.

The poll also asked some broader questions about respondents’ views of the power of the federal government and whether or not some of the rules and laws that govern the three branches are essential:

— Respondents had mixed opinions about which of the three branches of the federal government was the most powerful. About a third (35%) said that all three branches are equal, while about a fifth of the respondents apiece picked the Supreme Court (22%), Congress (20%), or the presidency (17%) as the most powerful branch. There were not any significant partisan distinctions among responses to this question.

— Majorities said the presidential veto (62%) and the two-term limit (79%) were essential to the “healthy functioning of American democracy.”

— By a 55%-33% margin, Americans did not agree that lifetime appointments to the Supreme Court were essential to the functioning of democracy.

— Respondents also did not deem the filibuster in the U.S. Senate as essential, with 39% indicating it was not essential, 30% indicating it was, and the remaining 30% not expressing an opinion.

— Just 33% of respondents agreed with the somewhat authoritarian proposal that to fix the United States, “we need a strong leader willing to break the rules.” Less than a quarter of Democrats (23%) agreed with that statement, but about half of Republicans did (53%).

Finally, the poll found support for beliefs that could broadly be described as anti-establishment:

— 71% (79% of Democrats, 63% of Republicans) agreed that “America needs a strong leader to take the country back from the rich and powerful.”

— 70% (84% Democrats, 49% Republicans) agreed that “the American economy is rigged to advantage the rich and powerful.”

— 69% (64% Democrats, 74% Republicans) agreed that “traditional parties and politicians don’t care about people like me.”

— 66% (49% Democrats, 87% Republicans) agreed that “the mainstream media is more interested in making money than telling the truth.”

— 65% (62% Democrats, 67% Republicans) agreed that “experts in this country don’t understand the lives of people like me.”

— 59% (52% Democrats, 67% Republicans) agreed that “politicians should be able to say what’s on their minds regardless of what anyone else thinks about their views.”

These are findings from an Ipsos poll conducted May 14-15 on behalf of the University of Virginia Center for Politics. For the survey, a sample of 1,005 adults 18+ from the continental United States, Alaska, and Hawaii were interviewed online in English. The sample includes 376 Democrats, 345 Republicans, and 205 Independents.

For more information about how the survey was conducted, and for the full results and data tables, please visit:

This is the latest collaboration between Ipsos and the University of Virginia Center for Politics, which have worked together on several state-level and national polls and partnered to create the UVA Center for Politics/Ipsos Political Atlas during last year’s midterm elections.

Notes on the State of the House

The Democrats' generic ballot edge endures, at least for now, but they shouldn’t get their hopes up on redistricting

Kyle Kondik, Managing Editor, Sabato's Crystal Ball May 9th, 2019



The Center for Politics is hiring a new coordinator of media relations who will help with the Crystal Ball and broader communications efforts at the Center. See here for a description of the position, information on how to apply, and who to contact if you have questions.

— The Editors



— While it’s very early in the cycle and these polls are not predictive so far in advance, the House generic ballot polling right now looks very similar to what we saw this time two years ago.

— Republicans almost certainly will need to lead on the generic ballot to retake the House, but perhaps they won’t need as big of a lead as we’ve seen in the past because of the nature of partisan voting in a presidential year and their abundance of targets in districts President Trump can or will carry.

— If new House maps are created in Maryland, Michigan, North Carolina, and Ohio because of various court orders, Democrats would benefit on balance. But it may very well be that no maps end up being changed.

The 2019 generic ballot polls: They’re a lot like the ones we saw in 2017

Let’s start here with a caveat: House generic ballot polling from a year and a half before a general election should not be treated as predictive. The 2020 House election depends on a lot of factors that remain unknowable, most importantly the tone-setting effect of the presidential race, which at this point appears to be something of a 50-50 proposition.

That said, if you’re looking for signs of a growing backlash against the new Democratic-majority U.S. House of Representatives, you won’t find it in national generic ballot polling so far.

Democrats led the House generic ballot the whole way in the 2017-2018 RealClearPolitics average. From Jan. 1, 2017 through April 30, 2017, there were 12 generic ballot polls included. The Democrats led every single one, and by an average of six points.

This year, from Jan. 1 through April 30, there are 11 polls in the average, and the Democrats lead in every one… by an average of six points.  

It is true that Democrats led the House generic ballot early in both the 2010 and 2014 cycles, too, both of which ended up being great years for Republicans in the House. That’s why we say this polling is not necessarily all that predictive. It may be that a pro-Democratic sentiment on the House generic ballot right now is essentially just a proxy for a preference for a generic Democratic presidential candidate against President Donald Trump. Former Vice President Joe Biden, typically the strongest-polling Democrat against Trump, leads the incumbent by about seven points, on average, in polling so far, pretty similar to the Democratic House ballot lead. If and when polling in the presidential race tightens, the generic may tighten, too. After all, we are in a partisan era where, as congressional scholar Gary Jacobson has documented, the incumbency advantage for members of Congress has diminished to a significant degree. A nationalized election could benefit House Republicans, most notably if Trump wins.

Republicans need to flip at least 18 House seats to win the chamber. With 31 Democrats in Trump-won seats, Republicans could win the House back just by sweeping all the Democratic House districts where Trump won by three points or more in 2018. That’s easier said than done. For one thing, we could imagine Trump’s performance weakening in some of these districts because they cover some affluent, highly-educated suburban areas that are trending away from Republicans. Also, and to the extent incumbency matters, Democrats have strong officeholders in some of these districts. Still, the districts are there for the Republicans to retake the House, at least on paper.

Typically one would expect a party flipping the House to have a considerable edge on the House generic ballot on Election Day. When the Democrats won the House from the Republicans in 2006 and 2018, their final generic ballot poll average lead in RealClearPolitics was 11.5 and 7.3 points, respectively. When Republicans took control of the House in 2010, their lead was 9.4 points. So the Republicans will need to be leading on the House generic ballot in order to win the House; they just might not need to lead by as much as we’ve seen in past years in order to win it given the characteristics of their best targets and the decline of ticket-splitting and the power of incumbency.

As it stands now, we continue to consider the Democrats favorites to hold their House majority, with a Trump reelection victory a necessary but perhaps not sufficient condition for a restoration of unified GOP rule in Washington.

Hold your horses on redistricting

Democrats’ chances to hold the House would be bolstered if court decisions mandating new districts in several states do in fact produce new maps for 2020. But don’t hold your breath.

Despite the 2018 House blue wave that flipped the lower chamber, Democrats came up empty in two big states: Ohio and North Carolina. Republicans maintained a 12-4 edge in the Buckeye State’s House delegation, and in the Tar Heel State, Republicans seemed to hold their 10-3 advantage until credible accusations of fraud prompted a do-over election in NC-9. The GOP primary for that race is next week (veteran Dan McCready, the 2018 nominee, is unopposed for the Democratic nomination).

A big part of the reason why Democrats failed to win any new seats in Ohio, and may not in North Carolina, is because both states have House maps gerrymandered by Republicans. The GOP actually gerrymandered North Carolina twice in the 2010s, both in an initial map for the 2012 election and then a re-do in response to a court order in advance of 2016. Still, the North Carolina maps have held for the Republicans throughout the decade, with the possible exception of the unusual NC-9 situation.

North Carolina redistricting, along with Maryland (which features a Democratic gerrymander), is at the center of a major Supreme Court decision in which the justices are once again considering whether to intervene to curb the excesses of partisan gerrymandering. Additionally, federal courts recently ruled that Michigan and Ohio’s House maps are partisan gerrymanders. Both maps were drawn by Republicans but, unlike in Ohio, Democrats picked up a couple of suburban Michigan seats last year despite the GOP gerrymander.

So it’s possible that all four of these states will have new House maps in 2020. That combination of outcomes would almost certainly benefit Democrats on balance. While Republicans would almost assuredly pick up a seat in Maryland — MD-6, held by Rep. David Trone (D), would become significantly more Republican in a fair remap — Democrats likely would pick up multiple seats from North Carolina and Ohio. Michigan is harder to figure because the delegation is currently 7-7 and a remap would not necessarily be guaranteed to help Democrats.

But it’s also possible, perhaps even likely, that none of the maps in these states actually will change at all this cycle. Republicans in Michigan and Ohio are hoping that the Supreme Court decides not to intervene in the Maryland and North Carolina cases, which might effectively overrule the lower courts’ decisions in Michigan and Ohio, too.

It took arguably the most liberal Supreme Court in American history, the Earl Warren court of the 1960s, to issue the landmark decisions that finally ended the malapportionment of congressional and state legislative districts, forcing districts to have equal populations and enshrining the principle of “one person, one vote” into legislative districting jurisprudence. Prior to that, the high court had stayed out of what Justice Felix Frankfurter called “the political thicket” of legislative map-drawing (Frankfurter dissented in Baker v. Carr, the 1962 ruling that in which the Supreme Court decided to enter said “thicket.”). Since those decisions, the Supreme Court has repeatedly declined to intervene in partisan redistricting cases, and the current John Roberts court is very much unlike the Warren Court in terms of ideology. The Supreme Court punted on a gerrymandering decision last year; since then, Brett Kavanaugh has replaced Anthony Kennedy on the court, arguably positioning the court further to the right.

In other words, and without knowing what the Supreme Court ultimately will say, we’d be surprised if this ends up being the court that intervenes against partisan redistricting.

So while we’d love to speculate in more details about how the maps in these four aforementioned states might specifically change in the event of remaps, such an exercise might very well be a waste of time.

Instead, let’s wait and see what the Supreme Court says.

Assessing electability: Like nailing Jell-O to a wall

Democrats are trying to figure out who is the best to beat Trump. It’s a difficult task

Kyle Kondik, Managing Editor, Sabato's Crystal Ball May 2nd, 2019



— Trump’s victory in 2016 presents a great counter-argument to the idea that campaign professionals and pundits can confidently determine in advance who is electable to the presidency and who is not.

— Many presidents beyond Trump have seemed unelectable at various points of their ultimately successful campaigns.

— As Democrats consider who has the best chance against Trump, they will have to sort through different kinds of electability arguments, any one of which may be right (or wrong), and only one of which will actually be tested.

The perils of determining who is electable, and who is not

“A very conservative Republican can’t win in a national election.”Gerald Ford, speaking about Ronald Reagan to the New York Times, March 2, 1980.

“I think he’s got a big electability problem.” Jerry Brown, speaking about Bill Clinton, during a March 15, 1992 Democratic presidential primary debate.

We’re not exactly sure when the awkward word “electability” really entered the national lexicon, but the concept — voters and party bigwigs making a pre-election assessment about who is likeliest to win — is surely as old as democracy itself.

Concerns about electability are dotted throughout American history, and have sometimes led party leaders to consider alternatives to nomination frontrunners — alternatives who would go on to become historically important presidents. A couple of examples: Sen. William H. Seward (R-NY) entered the race for the 1860 Republican presidential nomination as a favorite, but questions about his perceived weakness in crucial states “formed a growing cloud on the horizon of his anticipated nomination,” wrote Civil War-era historian James McPherson. “Pragmatists from all regions and politicians from the doubtful states combined in a stop-Seward movement,” McPherson wrote. Seward led on the first ballot at the Republican convention but by the third the nomination went to a man who emerged from a position as “the darkest of horses” to become the Republican nominee: Abraham Lincoln, arguably the greatest of all the presidents. Electability concerns about Sen. Robert A. Taft (R-OH) also informed the decision of some Republican leaders to draft Dwight D. Eisenhower into the 1952 presidential race.

“Electability” is clearly on the minds of Democrats as they determine who gets the nod to challenge President Donald Trump, the likely Republican nominee seeking a second term in the White House.

A few weeks ago, the Huffington Post’s Kevin Robillard and Amanda Terkel explored the idea of electability in the Democratic Party primary. It’s something everyone seems to care about, but few can persuasively define or determine:

The perception of which candidates stand the best chance of toppling Trump will play a major role in deciding who ultimately wins the Democratic Party’s nomination, according to polling and interviews with campaigns, operatives and rank-and-file voters across the early primary states.

But many of those perceptions and theories — Joe Biden can win back the Rust Belt! Isn’t Elizabeth Warren a bit like Hillary Clinton? Bernie Sanders can win West Virginia! — are based on flimsy evidence. And unlike the simple question of whom voters like the most, the question of electability involves evaluating what other people might like. And that’s something voters — and even political operatives — aren’t great at.

Kellyanne Conway, counselor to President Trump and his campaign manager during the last months of the 2016 contest, spoke at our University of Virginia Center for Politics’ American Democracy Conference after Trump’s victory. She questioned the ability of campaign operatives and observers to truly figure out who is electable and who is not: “What happens early in the process is, people say ‘So and so’s electable — he can win.’ And you very patiently should ask, ‘OK great, how do you know that?’ ‘Well, a hundred other people just said it on TV, it must be true,’” she said.

Certainly Trump’s victory is a great argument against the idea that so-called electability can be discerned in advance. For much of the campaign cycle it didn’t seem like Trump was capable of winning, but then he did. This was a vindication for Conway, obviously, but also for many conservatives who believed in the wake of John McCain and Mitt Romney’s losses to Barack Obama that the rank and file had bowed to party leaders concerned about “electability” and nominated candidates many GOP base voters didn’t like only to see them lose anyway. As Laura Reston of the New Republic put it in a 2016 piece from the late stages of the GOP primary about Ted Cruz and John Kasich arguing they were more electable than Trump, “The way rank-and-file conservatives see it, the party acceded to its most ‘electable’ candidates in the last two cycles… For many, if not most, Republicans, ‘electability’ has come to look like a way to both compromise the party’s principles and lose general elections.”

There are some problems with this argument from conservatives. Politics is often about timing, and the political conditions in 2016 for Trump (an open-seat election coming off two terms of a Democratic president with relatively mediocre approval ratings) were better than those for McCain in 2008 and Romney in 2012. The former had to deal with the weight of George W. Bush’s unpopular presidency and the latter faced the unenviable task of running against an incumbent president. Obama, it now seems very fair to say, was also a superior campaigner than Hillary Clinton and had better favorability numbers.

One could argue that because of extenuating circumstances, McCain and Romney, who lost, were more “electable” than Trump, who won, but on its face it’s kind of a silly argument because of what actually happened. Yes, perhaps Romney and McCain both would have won in 2016, and Trump would have lost in both 2008 and 2012. But ultimately, how can one say for sure?

Other recent presidents have overcome electability concerns. As noted above, the question of Clinton’s ability to win in 1992 was such an issue that one of his competitors, then-former (and future) California Gov. Jerry Brown, was asked to opine on it during a debate. Clinton had dealt with a litany of controversies, including his rampant womanizing and questions about whether he had dodged the draft during the Vietnam war. Clinton, though consistently dogged by scandal, went on to win two terms in the White House.

One of Hillary Clinton’s top aides, Mark Penn, wrote a memo in the early stages of the 2008 Democratic presidential primary that included an observation that her chief rival, Obama, was “unelectable except perhaps against Attila the Hun.” Whether any African American could win a general election was a theme that hovered over 2008.

As late as March 1980, former President Gerald Ford was saying this of his 1976 primary rival, Ronald Reagan, as part of the same interview cited above: “Every place I go and everything I hear, there is the growing, growing sentiment that Governor Reagan cannot win this election.” Ford, who at that time was still considering running again, watched as Reagan won the nomination and the presidency, re-orienting American politics toward the very same conservatism that Ford (and many others) saw as electorally unviable.

The bottom line here is that there are a lot of candidates who seem unelectable until, that is, they are actually elected.

Polls almost universally show Joe Biden running better against Donald Trump in a hypothetical 2020 matchup than his rivals for the Democratic nomination. Similar polls during 2015 after Donald Trump took the lead in the GOP primary generally showed him performing poorly against Hillary Clinton compared to other Republican contenders.

Does that mean that any GOP nominee was destined to win in 2016 against Clinton? Perhaps. But then one looks at the granular election results, and the immense improvements Trump made on recent Republican performances in the electorally vital rural areas and small cities of Big Ten Country. It is reasonable to wonder: Could Jeb Bush have done that? Or Ted Cruz or Marco Rubio? Perhaps so, but perhaps not.

But maybe those other Republicans wouldn’t have hemorrhaged as much support in some affluent suburban areas in those same states, and they could have won in a different, more traditional way. We can assemble data to suggest that from polling and extrapolating from down-ballot performances by more traditional Republicans in 2016, but there’s no way of knowing for sure. Elections are not simulations one can run over and over again to test the performance of different candidates. They are one-time propositions with “what ifs” we can only speculate about using imperfect measures, like opinion polls conducted months or even years in advance of a general election that test different possibilities of candidates.

Following 2012, a report commissioned by the Republican National Committee laid out a roadmap back to national power that, among other things, emphasized the need for comprehensive immigration reform and better outreach to minorities. One couldn’t have designed in a lab a candidate who flew more in the face of that roadmap than Trump, who has demagogued immigrants, loudly insisted on building a wall along the southern border with Mexico, and broke into national GOP politics by suggesting, without proof, that the nation’s first nonwhite president wasn’t really a natural-born American. By the Republican Party’s own reckoning, Trump wasn’t electable in 2016, which helps explain why party leadership was so hostile to his candidacy (that leadership is much more supportive of Trump and his renomination now).

It would be tempting for Democrats to look at Trump and argue that if Trump could fly in the face of conventional wisdom and run hard to the right and still win, then they could beat him by doing the same thing themselves and running to the left.

That may end up being the case, but there are a few factors Democrats should consider before coming to that conclusion. For one thing, self-identified conservatives still outnumber liberals in the general public. It may be that the general electorate would be more open to a far-right candidate than a far-left one because of the nation’s ideological orientation.

Additionally, the reality of Trump’s 2016 candidacy may be more complicated than just saying he was and is a far-right conservative.

Certainly on immigration he is. But Trump also subverted usual GOP messaging on other issues. Even though he did not actually oppose the Iraq war when it started in 2003, he pitched himself in 2016 as an opponent of that war and was highly critical of Bush administration foreign policy. He also sounded like a labor-oriented Democrat on trade issues, assailing recent trade agreements. While Trump was and is a right-wing conservative on certain issues, such as immigration and other issues of patriotism and national identity (such as his loud denunciations of Colin Kaepernick and other NFL players who kneeled during the national anthem), but in other respects he positioned himself as a critic of Republican foreign policy and trade policy. His was a hybrid candidacy. Perhaps that helps explain why Trump performed so strongly among the segment of the population who, in response to a battery of questions from the Democracy Fund Voter Study Group, placed themselves in the economically moderate or liberal but culturally conservative “populist” portion of the electorate (we wrote about this in some depth in the Crystal Ball back in 2017).

Logically, the most “electable” Democrat would be able to claw back some Trump voters who reside in that populist ideological portion of the electorate. We already mentioned Biden’s strength in general election polls against Trump compared to other Democrats, although we don’t think such polls are all that predictive at this point. Still, a polling edge combined with Biden’s support from the more moderate elements of the Democratic Party and his coolness to some of the more progressive policy proposals that have been discussed during the Democratic primary so far, like Medicare for all, might lead one to believe Biden is the most electable.

That said, Bernie Sanders is already going after Biden for his votes to authorize the Iraq war and in support of NAFTA. Couldn’t Trump do the same, effectively getting to the left of Biden on international issues and holding his populist support in the key heartland states as a result?

From that standpoint, maybe Sanders actually would be the more electable candidate; in addition to his votes against NAFTA and the Iraq war, Sanders’ political arguments are couched more on class than on race, which has sometimes gotten him into trouble with nonwhite activists who want the Democratic nominee to focus more on racial injustice and issues of white privilege. But isn’t it possible that de-emphasizing racial disparities in favor of emphasizing more color-blind class distinctions is a more electorally plausible path to winning back some white working-class voters?

At the same time, could it also be possible that a Democrat who did emphasize racial justice issues could generate better turnout among nonwhite voters, thus improving on Clinton’s margins enough in big, diverse, and electorally important cities like Detroit, Milwaukee, and Philadelphia to overcome Trump’s huge margins in the countrysides of Michigan, Wisconsin, and Pennsylvania? Alternately, could such a Democratic candidate reshape the electoral map altogether by competing better in the diverse, growing Sun Belt than the whiter Midwest?

Different people will evaluate these different “electability” arguments in different ways. Our only point is to say that the uncertainty associated with each, and many other such arguments that will emerge, demonstrates how tricky assessing who is likeliest to win actually is.