19 Things We Learned from the 2016 Election

ABSTRACT We can all agree that the presidential election result was a shocker. According to news reports, even the Trump campaign team was stunned to come up a winner. So now seems like a good time to go over various theories floating around in political science and political reporting and see where they stand, now that this turbulent political year has drawn to a close. In the present article, we go through several things that we as political observers and political scientists have learned from the election, and then discuss implications for the future.


The Shock
Immediately following the election there was much talk about the failure of the polls: Hillary Clinton was seen as the clear favorite for several months straight, and then she lost. After all the votes were counted, though, the view is slightly different: by election eve, the national polls were giving Clinton 52% or 53% of the two-party vote, and she ended up receiving 51%. An error of 2 percentage points is no great embarrassment.
The errors in the polls were, however, not uniform. As Figures 1 and 2 show, the Republican candidate outperformed by about 5% in highly Republican states, 2% in swing states, and not at all, on average, in highly Democratic states. This was unexpected in part because, in other recent elections, the errors in poll-based forecasts did not have this sort of structure. In 2016, though, Donald Trump won from his better-than-expected performance in Wisconsin, Michigan, North Carolina, Pennsylvania, and several other swing states.
Trump's win in the general election, and the corresponding success of Republican candidates for the U.S. Senate, then raises two questions: (1) What did the polls get wrong in these key states? (2) How did Trump and his fellow Republicans do so well? The first is a question about survey respondents, the second a question about voters.
Going backward in time from the election-day shocker, there is the question of how Trump, as a widely unpopular candidate without the full backing of his party, managed to stay so close during the general election campaign. Given the weaknesses of the Trump candidacy as traditionally measured, it was a surprise to many that he was polling at 48% of the two-party vote rather than, say, 40%. And, even before that, the surprise was that Trump won the nomination against so many in the Republican party.

The Party did not Decide
We can start with the primaries, which provided a counterexample to the Party Decides theory of Cohen et al. (2008), who wrote that "unelected insiders in both major parties have effectively selected candidates long before citizens reached the ballot box. " You cannot blame authors of a book on political history-its subtitle is "Presidential Nominations Before and After Reform"-for failing to predict the future. But it does seem that the prestige of the Party Decides model was one reason that Silver (2016), Cohn (2015; see also Palko 2015), and a bunch of other pundits not named Nate were so quick to dismiss Donald Trump's chances of winning in the Republican primaries, even while the leaders of the Democratic party seemed to be deciding according to plan (see Figure 3).
Most social scientists, ourselves included, were tempted to dismiss Trump's chances during primary season. But as we discussed several years ago (Gelman 2011), primary elections are inherently difficult to predict, with multiple candidates, no party cues or major ideological distinctions between them, unequal resources, unique contests, and rapidly changing circumstances. More recently, we entertained the idea that Trump's candidacy represented divergent preferences between Republican primary voters and elites (Azari 2015).
Where can we turn for insights about how the two parties function? The recent work of Blum (2016) on the Tea Party as  a "party within a party" illuminates some of the dynamics we observed in the 2016 nomination contest. While Donald Trump is not precisely an outgrowth of the Tea Party, this research offers insights about how an insurgent movement operates within a party while simultaneously challenging its leaders and methods of politics. Blum likens the post-2012 Tea Party to a third party movement, an important blueprint for the Trump candidacy.
While the Tea Party research builds on the parties-asnetworks theory that undergirds The Party Decides, there have been other strands of work within political science. Brown (2010) questioned the group-centric model, theorizing instead that nominations are shaped by candidates' ability to take advantage of party structures and rules. It is possible we could cast Trump as the kind of political opportunist that Brown describes: his ability to draw crowds and media attention worked well in the modern primary system, and he was able to navigate a crowded field of candidates to win the party nod. There is one area where Brown's theory gets it wrong, though: she argues for the importance of broad political experience as a big influence on whether candidates are successful at winning the nomination. Arguably, Trump succeeded by differentiating himself from the Republican field on issues, staking out distinctive positions on immigration and trade.
A growing group of scholars, including one of the authors of this piece, have begun to assess the impact of weak parties combined with strong partisanship (Azari 2016a). Schlozman and Rosenfeld (2017) identified the phenomenon as "hollow parties, " noting that "parties feel weak. They seem inadequate to the tasks before them, of aggregating and integrating preferences and actors into ordered conflict in American politics, of mobilizing participation and linking government to the governed. " The ideas of Milkis (1993), on executive-centered party politics, give perspective on the way the Republican Party became Trump's party over the course of the election, and explaining some of the governing developments since Trump took office, including high levels of approval from Republicans in the electorate and support from Congressional Republicans. Azari maintains that weak parties and strong partisanship tells much of the story of the 2016 race, in which party coordination efforts proved inadequate to stop Trump, but once he was the nominee, partisan loyalties kicked in. On the Democratic side, the situation can be summarized as too much branding and not enough building, in the sense of Galvin (2009).

The Ground Game was Overrated
The Democrats were supposed to be able to win a close election using their ability to target individual voters and get them out to the polls. McKenna and Han (2014) discussed the innovative use of volunteers in Obama campaigns, and Masket, Sides, and Vavreck (2015) and Enos and Fowler (2016) estimated that ground campaigning did not determine the election outcome in 2012 but that it did increase turnout in the most targeted states by several percentage points. After 2016, the new consensus is that some ground game is necessary, but it is hard to get people to turn out and vote, if they were not already planning to. During the past few decades, campaigns have moved from the "air war" to the "ground game" to some new world of social media. Fowler, Ridout, and Franz (2016) argued that the 2016 presidential election was a significant departure from past patterns in terms of advertising, and, of course, the presidential candidate who advertised less was the winner. However, they also suggest that Clinton's surprise losses "came in states in which she failed to air ads until the last week" and that Clinton's ads also differed from previous patterns in their lack of policy content.

Overconfident Pundits get Attention
From one direction, neuroscientist Wang (2016) gave Hillary Clinton a 99% chance of winning the election; from the other, cartoonist and political provocateur Adams (2015) gave 98% odds in favor of Trump. Looking at it one way, both Wang and Adams were correct: Clinton indisputably won the popular vote while Trump was the uncontested electoral vote winner. After the election, Wang blamed the polls, which was wrong. The polls were off by 2%, which from a statistical standpoint was not bad. Indeed this magnitude of error was expected from a historical perspective , even if it did happen to be consequential this time. The mistake was not in the polls but in Wang's naive interpretation of the polls, which did not account for the possibility of systematic nonsampling errors shared by the mass of pollsters, even though evidence for such errors was in the historical record. Meanwhile, Adams explains Trump's victory as being the result of powers of persuasion, which might be so but does not explain why Trump received less than half the vote, rather than the landslide that Adams had predicted.
We continue to think that polling uncertainty could best be expressed not by speculative win probabilities but rather by using the traditional estimate and margin of error. Much confusion could have been avoided during the campaign had Clinton's share in the polls simply been reported as 52% of the two-party vote, plus or minus 2 percentage points. That said, when the general presidential election is close, the national horse race becomes less relevant, and we need to focus more on the contests within swing states, which can be assessed using some combination of state polls and state-level results from national polls. An additional problem is the difficulty that people have in understanding probabilistic forecasts: if a prediction that Clinton has a 70% chance of winning is going to be misunderstood anyway, why not just call it 98% and get more attention?
There is a theory that academics such as ourselves are petrified of making a mistake, hence we are overcautious in our predictions; in contrast, the media (traditional news media and modern social media) reward boldness and are forgiving of failure. This theory is supported by the experiences of Sam Wang (who showed up in the New York Times explaining the polls after the election he had so completely biffed) and Scott Adams (who triumphantly reported that his Twitter following had reached 100,000).

That Trick of Forecasting Elections using Voter Predictions rather than Voter Intentions? Does not Work
Economists Rothschild and Wolfers (2013) had argued that the best way to predict the election is not to ask people whom they will vote for, but rather ask whom they think will win. Their claim was that when you ask people whom they think will win, survey respondents will be informally tallying their social networks, hence their responses will contain valuable information for forecasting. When this idea was hyped a few years ago, Gelman (2012) was skeptical, taking the position that respondents will be doing little more than processing what they had seen in the news media, and we remain skeptical, following a 2016 election that was a surprise to most. That said, we recognize the value of research into social networks and voting, especially in a fractured news media environment and declining trust in civilian institutions (as documented, e.g., by Weakliem 2016). In future studies, we recommend studying information about networks more directly: instead of asking voters who they think will win the election, ask them about the political attitudes of their family, friends, and neighbors.

Survey Nonresponse Bias is a Thing
It is harder and harder to reach a representative sample of voters, and it has been argued that much of the swing in the polls is attributable not to people changing their vote intention, but to changes in who responds or does not respond. In short, when there is good news about a candidate, his or her supporters are more likely to respond to polls.  floated this theory following some analysis of opinion polls from 2012, and it seems to have held up well during the recent campaign season .
The only hitch here is that the differential nonresponse story explains variation in the polls but not the level or average shift. The final polls were off by about 2 percentage points, suggesting that, even at the end, Trump supporters were responding at a lower rate than Clinton supporters, most notably in certain key swing states.
Shortly after the election, some people attributed these differences to "shy Trump voters" who were not willing to admit their unpopular views to pollsters. We are skeptical of this explanation, given that, as shown in Figures 1 and 2, Trump outperformed the polls the most in states such as North Dakota and West Virginia where we assume respondents would have had little embarrassment in declaring their support for him, while he did no better than the polls' predictions in solidly Democratic states. Also, Republican candidates outperformed expectations in the Senate races, which casts doubt on the model in which respondents would not admit they supported Trump; rather, the Senate results are consistent with differential nonresponse or unexpected turnout or opposition to Hillary Clinton. It is possible that the anti-media, anti-elite, and even anti-pollster sentiment stoked by the Trump campaign has been a part of the reason for the low response of Trump supporters in states with large rural populations.

News is Siloed
For years we have been hearing that liberals hear one set of news, conservatives hear another, and moderates are being exposed to an incoherent mix, so that it is difficult for anyone to make sense of what everyone else is hearing. There have always been dramatic differences of opinion (consider, e.g., attitudes toward civil rights in the 1950s and the Vietnam war in the 1970s) but research on public opinion has shown an increase in partisan polarization in recent decades. The 2016 election, with its sharp divide between traditional news organizations on one side and fake news spread by Twitter and Facebook on the other, seems like the next step in this polarization.
It is the political version of Moore's Law, which says that every time the semiconductor manufacturers have run out of ways to squeeze more computing power on a chip, they come up with something new. Whenever it starts to seem like there is no more room for Americans to polarize, something new comes up-in this case saturation of social media by fake news, along with a decline of the traditional TV networks and continuing distrust of the press. Studying the effect of news media more generally, is challenging. In a recent study using passive monitoring data, Guess (2016) found that "most people across the political spectrum have centrist media diets composed largely of mainstream portals. " Barbera (2015) pointed out that "social media platforms facilitate exposure to messages from those with whom individuals have weak ties, which are more likely to provide novel information to which individuals would not be exposed otherwise through offline interactions. " Barbera continues: "Because weak ties tend to be with people who are more politically heterogeneous than citizens immediate personal networks, this exposure reduces political extremism. " This last claim may need to be reassessed in the light of aggressive social media campaigning in 2016. In particular, it may be that the largest effect of fake news is in how it has affected voters' reception of mainstream media coverage. This is related to the work of Mutz (1998) on the indirect connections between voters and the news media.

A Working-Class Pundit is Something to Be
Filmmaker and political activist Michael Moore (2016) gets lots of credit for writing, over a month before the election, an article entitled "5 Reasons Why Trump Will Win, " specifically pointing to the Rust Belt, angry white men, voter turnout, and other factors that everybody else was writing about after the election was over. Moore even mentioned the Electoral College. And unlike the overconfident pundits mentioned above, Moore clearly stated this as a scenario ("As of today, as things stand now, I believe this is going to happen ... ") without slapping a 98% or 99% on to it. From the other side, conservative Vance (2016) has been mined for insight into the evident appeal to rural whites of the anti-establishment message sent by brash New Yorker Donald Trump.
What if Hillary Clinton had won 52% of the two-party vote and a solid Electoral College victory? Would we now be hearing from pundits with a special insight into white suburban moms? Maybe so. Or maybe we would still be hearing about the angry white male, since 48% of the two-party vote would still be a lot more Trump support than most were expecting when the campaign began. When considering explanations of the vote, we should distinguish between predictions of who will win (which is difficult in an era of partisan polarization and in an election year in which the economy was neither crashing nor booming) and insights into the attitudes of particular subgroups.

Beware of Stories that Explain too Much
After the election, which shocked the news media, the pollsters, and even the Clinton and Trump campaigns, our colleague Thomas Basboll (2016) wrote that "social science and democracy are incompatible. The social sciences conduct an undemocratic inquiry into society. Democracy is an unscientific way of governing it. " Maybe so. But Basboll could have written this a few days before the election. Had the election gone as predicted, with Clinton getting the expected 52% of the two-party vote rather than the awkwardly distributed 51% that was not enough for her to win in the Electoral College, it still would have been true that half of American voters had refused to vote for her. So there is something off about these sweeping election reviews: even when you agree with the sentiments, it is not clear why it makes sense to tie it to any particular election outcome.
The Republicans have done well in political strategy, tactics, timing, and have had a bit of luck too. One party right now controls the presidency, both houses of Congress, most of the governorships, and soon the Supreme Court. But when it comes to opinions and votes, we are a 50/50 nation. So we have to be wary of explanations of Trump's tactical victory that explain too much. It is possible to attribute the outcome of a close election to any number of possible factors, and thus we prefer to think of explanations as additive.

The Election Outcome was Consistent with "The Fundamentals"
Various models predict the election outcome not using the polls, instead using the national economy (as measured, e.g., in inflation-adjusted personal income growth during the year or two preceding the election) and various political factors. In 2016, the economy was growing slowly but not booming (a mixed signal for the voters), the incumbent party was going for a third term in office (traditionally a minus, as voters tend to support alternation), and the Republicans controlled both houses of Congress (a slight benefit for the Democrats in presidential voting, for that minority of voters who prefer party balancing), and, on the left-right scale, both candidates were political centrists relative to other candidates from their parties. This information can be combined in different ways: Running a version of the model updated by the political scientist Hibbs (2013), we gave Hillary Clinton a forecast of 52% of the two-party vote (Gelman 2016b). Fitting a similar model but with slightly different parameters, political scientist Drew Linzer gave Clinton 49% (see Kremp 2016). In October the political science journal PS published several articles on forecasting the election, including one from Erikson and Wlezien (2016) who concluded, "the possibility of greater campaign effects than we typically observe should constrain our confidence in the predictions presented here. " All these fundamentals-based models have uncertainties on the order of 3 percentage points, so what they really predicted is that the election would not be a landslide. The actual outcome was consistent with these predictions. That said, a wide range of outcomes-anything from 55-45 to 45-55-would have jibed with some of these forecasts. And the nonblowout can also be explained by countervailing factors: Perhaps Trump was so unpopular that anyone but Clinton would have destroyed him in the general election, and vice versa. That seems doubtful. But who knows?

Polarization is Real
Democrats vote for Democrats, Republicans vote for Republicans. It is always been thus-what would the party labels mean, otherwise?-and party identification is a longstanding theme in political science, as in the classic The American Voter by Campbell et al. (1960). But cross-party voting keeps declining, and members of the out-party hold the president in lower and lower esteem. Consider, for example, Donald Trump's criticism of Barack Obama during the presidential debates. Obama was popular so this might seem to have been a mistake to stand against him-but Obama was deeply unpopular among Republicans, especially those Republicans who are likely to vote. Another sign of polarization is a decrease in the proportion of voters who split their ticket by voting for different parties in the senatorial and presidential races.
While we know a great deal about the extent and impact of polarization, scholars have been trying to figure out what really fuels polarization and what lies at the root of voters' attachments to their party labels. Huddy, Mason, and Aarøe (2015) found, through a series of experiments, that citizens are more powerfully motivated by messages about threats to their partisan identity ("this election will be a bad one for Democrats") than about issues (messages like, "this election is about healthcare"). Partisanship is also growing more negative over time, with Abramowitz and Webster (2016) finding that partisans are both increasingly loyal and increasingly likely to express negative feelings about the other party. Studies like these help illustrate how Republicans who did not initially like Trump were unwilling, in the end, to abandon their party's nominee and risk a Clinton presidency. Huddy, Mason, and Aarøe's finding that the team identity aspect of partisanship serves as a stronger motivating force than concern about issues also helps explain how Trump's breaks with Republican Party orthodoxy on issues like trade were not a great hindrance to his candidacy.
Many voters are bothered by the increasing polarization of American political life. Klar and Krupnikov (2016a) examined how independents differ from partisans in their political expression, often voting like partisans, but not wanting to be identified as such. Applying their findings to the recent campaign, Klar and Krupnikov (2016b) wrote that Trump and Sanders allowed their parties to reject their parties-while still staying with them. This might turn out to be an important point about the general election, too.
A corollary of polarization is that, if there are not many people in the middle to be persuaded, it can make sense for candidates in the primary and general election campaigns to focus on firing up their base and forget about moderation, and this is a key part of the story of the success of the Trump campaign. You can bet that activists of both parties will have learned this lesson when 2020 comes along.
The success of Donald Trump demonstrates both the strength of party identification and its internal contradictions.
On one hand, a candidate who was massively unpopular even among his own party regained almost all the Republican votes by the time the election came on. On the other hand, if activists feel they can rely on partisan voting in the general election, this motivates the continuing choice of less traditionally "electable" candidates, so that "party identification" is less and less about the parties themselves.

Demography is not Destiny
We had been hearing a lot about how the Republican party, tied to a declining base of elderly white supporters, needs to reassess, as in this headline from Slate: "It Lost Black Voters. Now It's Losing Latinos. What's Left Is a Broken, White GOP" (Bouie 2016). Hillary Clinton won most of the minority vote but the electorate is whiter than observers had thought based on exit polls (Cohn 2016). Longer term, it may well be that the Republican party needs to change with the times-after all, their presidential candidate did lose by three million votes-but destiny has not kept them from maintaining their control of both houses of Congress and most state legislatures.
At the same time, a growing body of social science suggests that race, gender, ethnicity, and religion shape the electorate. Based on available data and extrapolations, a debate rages about how many Latino voters chose Trump. Exit polls assigned Trump nearly 30% of the Latino vote, defying expectations and logic in light of Trump's comments on the campaign trail, but Segura and Barreto (2016) argued that exit polls are not designed to accurately reflect the votes of minority populations, and suggest that Clinton did extremely well in areas that are predominantly Latino.
One of the more confounding demographic stories of 2016 was that of gender. Gender solidarity did not drive votes for the first female major-party nominee, and it did not for Trump, either. Men and women alike voted along party lines, and a majority of white women voted for Trump, but the gender gap in 2016 was about twice as high as in recent years.
Finally, there is an argument to be made that the election was decided by a demographic bloc not usually identified as such: white voters. Nate Cohn summed this up with an election-night tweet: "How to think about this election: white working class voters just decided to vote like a minority group. They are >40% of the electorate. " Abrajano and Hajnal (2017) similarly documented a "transformation" in the white electorate, in which white voters have shifted from the Democratic to the Republican Party, and attitudes about Latinos and immigration partly drive partisanship. Tesler (2016a, b) had studied how racial attitudes have taken on new importance in the Obama era, and after the election he wrote that "views about race mattered more in electing Trump than electing Obama. " Racial appeal also interacted with the rules of the Electoral College: in 2016 whites were overrepresented in swing states, and calculations based on forecast probabilities showed average voting power to be higher among whites than other ethnic groups (Gelman and Kremp 2016).

Public Opinion does not Follow Elite Opinion
Perhaps the most disturbing theoretical failure of political science is the general idea that voters simply follow elite opinion. This worked in 1964 to destroy Goldwater, for instance.
Or so the story goes. The implication is that voters had to be told Goldwater was scary. They could not figure it out for themselves.
In 2016, Trump was opposed vigorously as dangerous, incompetent, xenophobic, tyrannical, and unhinged, by almost everybody in elite circles: most of his Republican primary opponents at one time or another, a large number of conservative intellectuals, former Republican candidates Romney and McCain, the various Bushes, the media, almost all newspaper editorialists including those that were reliable Republican supporters, all Democrats, about 10 Republican senators, and even some pundits on Fox News. Further, Trump's breaking of all the standard niceties of politics was there for all to see for themselves. But half the voters said, we go with this guy anyway. "The falcon no longer hears the falconer, " as W. B. Yeats put it.
To put it another way, the elites in the Republican party had a coordination problem, which allowed one of the most disliked choices to win the nomination in a multi-candidate primary campaign. At this point, one might well ask whether elites are now following public opinion: are elected officials who would like to challenge Trump afraid to alienate their voters? These sorts of questions demonstrate the connections between public opinion and legislative politics: Congressional Republicans are reliant on Trump's support within their party but fearful of his unpopularity among Democrats and independence; meanwhile, Trump relies on the forbearance of a Republican-led Congress to avoid being engulfed by investigations of scandals.

There is an Authoritarian Dimension of Politics
Political scientists used to worry about authoritarianism within the electorate. Mainstream politicians, ranging from Republicans on the far right to lefties such as Sanders, tend not to go there. Trump did. In doing so he broke the rules of politics with extreme comments about his opponents, etc., that are hard to forget. But a significant segment of the electorate, maybe 20%, have always been waiting for its authoritarian champion on what we now call the alt-right dimension. There had not been one in the modern era. Trump's absolute dominance of the political news for over a year signifies this uniqueness. There had been others with this sort of appeal, notably Joe McCarthy (see, e.g., Gelman 2016a) or George Wallace, but they never came close to becoming our national leader.

Swings are National
When you look at changes from one election to the next, the country moves together. If you plot vote swings by county, or by state, you see much more uniformity in the swing in recent years than in previous decades (see, e.g., Gelman et al. 2009). There has been lots of talk of Pennsylvania, Michigan, and Wisconsin, and these three states did make the difference in the electoral college, but these swings did not greatly alter the general national picture of Republican support in the center of the country and Democratic strength on the coasts; see Figures 4 and 5. To put it another way, nonuniform swings were essential to Trump's win, but looking at public opinion more broadly, the departures from a national swing were small, and consistent with the increasing nationalization of elections in recent decades. That said, within states there were clear differentials, with Trump consistently outperforming Romney in poorer counties and doing worse in richer counties, in all regions of the country.

The Election was not Decided by Shark Attacks
Political scientists Achen andBartels (2002, 2016) had argued that voters are emotional and that elections can be swayed by events that should logically be irrelevant to voting decisions. Most memorably, they analyzed data from the 1916 presidential election and identified a pattern in seaside counties in New Jersey that they attributed to voters irrationally responding to shark attacks. Others have analyzed recent county and state-level voting data and claimed to find that close elections can be decided by the outcomes of college football games (with happy voters being more likely to pull the lever for the incumbent party's candidate). However, reanalysis of both these studies revealed no clear effect in either case (Fowler and Montagnes 2015;Fowler and Hall 2016).
What does 2016 say about all this? Not much. You cannot prove a negative so it is possible that irrelevant stimuli could have made all the difference. But the big stories about this election were that (a) lots of bad information about Donald Trump did not sway much of the electorate, and (b) Clinton's narrow Electoral College loss may well be attributed to FBI leaks, which were relevant to the voting decision in reminding voters (perhaps inappropriately) of concerns about her governing style. The 2016 election was not about shark attacks or football games but rather about big stories that did not matter much, or canceled each other out.
That said, we should not let the silliness of the shark attack reasoning to distract us from Achen and Bartels's larger point that many voters are massively uninformed about politics, policy, and governing, which is relevant even if it is not true, as they claimed, that voters are easily swung by irrelevant stimuli.
Some emotional stimuli are more obviously connected to politics than others, and the outcome of 2016 suggests that emotions matter for political decision-making. Gadarian and Albertson (2015) found, in a series of experiments, that inducing anxiety in voters can lead to changes in preferences-namely, "support for protective policies" that will shield them from the effects of terrorism, pandemic outbreaks, and climate change. However, their research showed that voters anxious about terrorism would tend toward trusted figures-something that they argued seemed to give the advantage to Clinton (Albertson, Busby, and Gadarian 2016), which just goes to show the difficulty in applying general theories to particular elections.

Red State Blue State is Over
Republicans have done better among rich voters than among poor voters in every election since the dawn of polling, with the only exceptions being 1952, 1956, and 1960, which featured moderate Republican Dwight Eisenhower and then moderate Democrat John Kennedy. Typically the upper third of income votes 10 to 20 percentage points more Republican than the lower third ). But 2016 was different. For example, the exit polls reported that Clinton won 53% of the under-$30,000 vote and 47% of those making over $100,000, a difference of only 6 percentage points, much less than the usual income gap. And we found similar minimal income-voting gradients when looking at other surveys. And political scientist Cramer (2016) has been documenting the growing urban-rural divide within states (specifically, Wisconsin, ground zero of the demise of Hillary Clinton's "blue wall")-a division that is not just about income, but about education and culture as well. Will the partisan income divide return in future years? Will it disappear? It depends on where the two parties go. Next move is yours, Paul Ryan.

Third Parties are Still Treading Water
The conventional wisdom is that minor parties are doomed in the U.S. electoral system. The paradox is that the only way for a minor party to have real success is to start local, but all the press comes from presidential runs. Anyway, 2016 seems to have confirmed conventional wisdom. Both major-party candidates were highly unpopular, but all the minor parties combined got only 5.6% of the vote. On the other hand, 5.6% is a lot better than 1.7% (2012), 1.4% (2008), 1.0% (2004), or 3.7% (2000).
In light of the research on the topic, the real surprise is that third parties did not make a stronger showing in 2016. In a study of support for George Wallace, John Anderson, and Ross Perot, Abramson et al. (1995Abramson et al. ( , 2000 argued that the third party voting is rooted mostly in dissatisfaction with major party candidates, not the parties themselves. In other words, voters turn to third parties when they are not so happy with the party nominee, but do not want to cross all the way over and support the other side. It is possible the real story in 2016 is that more of this did not happen.
The glass half full for minor parties is that they are starting to get serious; the glass half empty is that not much bloomed even in such fertile soil.

Goldman Sachs Rules the World
This theory appears to still hold up. Hillary Clinton was criticized during the primary and general election campaigns for her close ties to Wall Street, in particular her paid speech at Goldman Sachs where she said, "There is such a bias against people who have led successful and/or complicated lives. " But Clinton's electoral vote loss did not lock the famed investment bank out of government, as Goldman Sachs-supported Senator Chuck Schumer may now be the most powerful Democrat in Washington, and former Goldman Sachs executives Stephen Bannon, Steven Mnuchin, and Gary Cohn are slated to decide political strategy inside the White House and economic strategy at the Treasury Department and the National Economic Council. So it looks like the banksters are doing just fine. They had things wired, no matter which way the election went. The work of Ferguson, Jorgensen, and Chen (2016) gives some insight into the links between party positions and their sources of funds that are necessary to compete in modern elections.

The Electoral College was a Ticking Time Bomb
Yup. Only three of the past five elections have been decided in favor of the popular vote winner, and presidential elections in recent decades have been closer than at any time since the 1880s (see sec. 8 of Gelman 2014). Much of the discussion of 2016 focused on Trump's unexpected performance in swing states. But, more generally, razor-thin margins at the national level create the conditions for Electoral College reversals.

Lessons for the Future
How does the shock of 2016 affect how policymakers, political professionals, activists, and citizens should think about future elections?
Going into the off-year election 2018, a big question is the importance of balancing from the subset of voters who prefer divided government. George W. Bush's two terms laid the groundwork for Democrats to win both houses of Congress in 2006 and 2008, then Obama's presidency motivated enough voters to balance so that Republicans gained a bit more than 50% in congressional and local elections, allowing them as the opposition party to control the House of Representatives, the Senate, and most governors and state legislatures. 2016 also can be viewed as a balancing election: polls show that most voters expected Hillary Clinton to win the presidency, which partly explains the Republicans' retention of the Senate.
In the 2009-2010 cycle, the balancing trend was so strong that generic congressional ballot polling gave us the confidence to predict that "the Democrats are gonna get hammered, " over 8 months in advance of the 2010 election. We will see how the polls look this September, but research by Bafumi, Erikson, andWlezien (2006, 2010) shows that the generic congressional ballot, when suitably adjusted, has historically been an accurate prediction of the forthcoming House vote.
Balancing provides the conditions for an opposition electoral comeback but the party (in this case, the Democrats) still has to field the candidates and run the campaigns. Is a coherent message required? The experience of 1994, 2006, and 2010 suggests that opposition to the president can be enough, if tied to specific issues with broad appeal, most naturally health care and social security in the current political climate.
What about racial politics and the Republicans' strength among middle-aged and older whites, who remain the dominant voting group (especially in off-year elections)? The challenge for both parties is to craft a message with universal appeal while satisfying their core constituencies. The current political map puts the Democrats at a disadvantage because their votes are geographically concentrated; hence they might need 52% or more of the national congressional vote to win half the seats. This is possible-the Democrats received 55% of the vote in the wave elections of 2006 and 2008-but this higher threshold puts them under more of a burden to craft a broadly appealing message.
Moving from elections to policy, the elections of 2008-2016 and their immediate consequences have struck a strong blow to the millions of Americans who oppose corporate influence in politics. Barack Obama had a background as a community organizer and excited the liberal wing of the Democratic party, and he implemented significant liberal policies including a Keynesian fiscal stimulus and a national health care plan-but he also, notoriously, joined in the bipartisan bailout of Wall Street. In 2012, Obama was opposed by Wall Street candidate Mitt Romney, and in 2016, Hillary Clinton was attacked for her Wall Street connections and her newfound opposition to the Trans-Pacific Partnership, but then, as noted above, several of incoming president Trump's most important political and economic advisors have Goldman Sachs backgrounds. To the extent that corporate control of national politics is a real concern-rather than just another way for voters to express their partisan identities-it is not clear what will come next. Between Donald Trump on one side, and Paul Ryan and the Koch brothers on the other, there does not seem to be any space remaining for an anti-Wall Street bloc within the Republican Party. But Bernie Sanders's near-miss in 2016 suggests that core economic issues still have the potential to divide the Democratic leadership.
The continued strength of partisan polarization augurs close national elections and tough campaigning by both parties, as well as motivating activists on both sides to push for their favored candidates. If nearly everyone is voting on party lines, then "electability" is not such a concern. In the wake of the victory of Donald Trump and the loss of Hillary Clinton, it will be difficult for moderate leaders of either party to persuade primary election voters to set aside their hearts and choose the purportedly safe option.
Looking forward to 2020, the success of fundamentals-based models suggests that the Trump administration will, like its predecessors, try to time the business cycle for economic growth in years 3 and 4 of the presidential term. It is not clear how attainable this goal will be, but we expect it will be a key driver of policy debates. Polarization suggests the election will be close, which will put a premium on innovative campaign techniques as well as attempts to game the system via vote suppression. From the Democrats' end, the challenge will be to broaden their geographic base of support so they can win in an electoral college, which currently favors whites. Polling technology will continue to advance so that both parties will be aware of where their strengths and weaknesses are-even if they may find it difficult to do anything about them.
Finally as noted above, we are not so concerned that the election will be determined by irrelevant factors such as football games or shark attacks (see item 15 in our list), but close elections and partisan polarization can well lead to a crisis of legitimacy of the government. In this article, we have focused on campaigns and elections, but the other half of the story is what the parties do when in office and in opposition. Congressional Republicans' near-complete opposition to the Obama agenda seems in retrospect to have been a natural strategy but in 2008 few commentators were anticipating it. At this point the relation between Trump and congressional Republicans is unclear: so far the Senate has approved many of the most controversial presidential appointments but it remains open how strong is the Republican consensus for taking apart the U.S. government and replacing it with right-wing loyalists. Anything can happen, and indeed this uncertainty is, ironically, one of the more expected consequences of Trump's electoral victory.
The political science of public opinion and elections, which has informed this piece, does not in itself give much insight into the politics within the majority party. What the study of elections and attitudes can tell us about are constraints: politicians do not want to lose office, and for that reason among others they are motivated to be responsive to the opinions of voters and, as Jacobs and Shapiro (2000) explained, to do their best to manipulate these opinions. The 2016 elections have given the paradox of voters who do not follow their party leaders on key issues but remain strongly partisan and predictable in their generalelection voting. On one hand, this seems inherently unstablefor how long can partisan loyalty be sustained by little more than opposition to the other side-but the combination of winnertake-all elections and (until recently) political gridlock also can make it a hard pattern to break.