A Hoover Institution Essay on contemporary american politics

Hoover Institution
The 2016 Presidential ElectionAn
Abundance of Controversies
As the polls universally predicted, Hillary Clinton won the popular vote. But contrary to
universally held expectations, Donald Trump shocked the political world by breaching the
Democrats blue wall and winning a majority of the Electoral College. This was only
one of myriad controversies associated with the 2016 voting, several of which continue
months after the election. Disappointed commentators pronounced that Trumps win was
a victory for racism, sexism, ethnocentrism, and other social pathologies. As we discuss in
the next essay, preliminary analyses indicate that such charges are exaggerations at best. In
this essay, we note that the voting in 2016 was little different from that in 2012. But in a
majoritarian system like that in the United States, small shifts in votes can have enormous
consequences for political control and public policy. The data indicate that millions of
Americans faced a choice between two candidates they regarded as extremely unattractive.
Far from an endorsement of Trumps more incendiary positions, Americans voted for him in
spite of those positions, believing that the alternative was worse.

MorrisP. Fiorina Series No. 10

Even by the colorful standards of presidential primaries, the 2016 election cycle has been filled with
jaw-dropping, head-scratching moments.Eric Bradner

While the worldcelebratesand commiseratesaDonald Trump presidency,one thing is clear: this will go
down as the most acrimonious presidential campaign of all.Rachel Revesz

Controversial presidential elections are nothing new in American electoral

history, 2016 being the latest, but certainly not the first. Despite much apocalyptic
commentary, however, the implications of the 2016 election seem less dire than those
of some elections held in earlier eras. The four-candidate 1860 election started the
country on the path to civil war and the disputed election of 1876 threatened to reignite
that conflict. In more recent times, the strong showing of a racist third party in 1968
coupled with political assassinations and civil disorders on a scale not seen since

Quotations are from Eric Bradner, 13 jaw-dropping moments of the 2016 campaign, CNN, August10, 2015,
www.cnn.com/2015/08/09/politics/2016-campaign-surprise-moments-donald-trump/; and Rachael Revesz,
How the 2016 Presidential Election Was Won: The Timeline, Controversies and Seats that Led to the White
House, The Independent, November9,2016, www.independent.co.uk/news/world/americas/us-elections
/presidential-election-2016-results-timeline-controversies-quotes-seats-maps-polls-quotes-a7398606.html.
2

the labor violence of the early twentieth century led some contemporary observers
to believe that the country was coming apart.1 The 2000 Florida electoral vote
contest raged for more than two months, threatening a constitutional crisis and
deeply dividing partisan activists on both sides. Still, even allowing for the fact that
secession and revolution are not seriously on the table, for the sheer number and
breadth of the controversies that accompanied it, the 2016 election does seem out of
the ordinary.

Parties have nominated flawed candidates beforeRepublican Barry Goldwater in

1964 and Democrat George McGovern in 1972, for examplebut at least since the
advent of scientific survey research, no major party has nominated a candidate so
wanting in the eyes of the electorate, let alone both doing so simultaneously. Charges
of ethnocentrism and racism are as American as apple pie, but in their prevalence
and virulence in 2016 (with misogyny added to the toxic mix) they were reminiscent
of 1928, if not the late nineteenth century.2 Biased media is a complaint common
to all elections, but the retreat from objectivity by the mainstream media in 2016
struck many observers as a significant break with modern journalistic practices.3 The
increasingly visible role of social media like Twitter threatened to further diminish
the importance of the legacy media. Swing voters, largely missing in action in recent
elections, suddenly reappeared in 2016.4 Possible foreign intervention in the election
was a new development (at least insofar as the United States was the intervenee
rather than the intervener), as was FBI involvement (but possibly only because earlier
instances did not become public). Meanwhile journalists scrambled to read up on
populism, which had not played such a significant role in American elections since
the 1960s. Class, long ago displaced by discussions of race, ethnicity, gender, and
sexual orientation in college course syllabi, enjoyed an academic as well as political
revival (so did authoritarianism, another oldie but goodie).5 All of this was overlaid

1WilliamL. ONeill, Coming Apart: An Informal History of America in the 1960s (New York: Quadrangle, 1973).

2In 1928, the Democratic candidate was Catholic Al Smith of New York. Religious, ethnic, and urban-rural
divisions dominated the election. As noted in essay 4, much of the politics of the late nineteenth century
revolved around ethno-cultural divisions.

3See the debate sparked by Jim Rutenberg, Trump Is Testing the Norms of Objectivity in Journalism,
NewYork Times, August7, 2016, www.nytimes.com/2016/08/08/business/balance-fairness-and-a-proudly
-provocative-presidential-candidate.html?_r=0.

4Nate Silver, The Invisible Undecided Voter, FiveThirtyEight, January23, 2017, fivethirtyeight.com/
features/the-invisible-undecided-voter/.

5Amanda Taub, The Rise of American Authoritarianism, Vox, March1, 2016, www.vox.com/2016/3/1/11127424
/trump-authoritarianism; cf. Wendy Rahn and Eric Oliver, Trumps Voters Arent Authoritarians, New Research

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 3

on a split decision where Hillary Clinton won a clear popular vote plurality and
Donald Trump a clear electoral vote majority. The impact of this troubling outcome
was probably exacerbated by the sheer shock of a Trump victory when the various polls
and models assured the political universe that Clinton was a surefire winner: you
could take it to the bank.

This essay and the next review some of the aforementioned developments. Such
a review is necessarily modest and incomplete, given that in these immediate
postelection months the ratio of opinion to research is highly skewed toward the
former.

Some Perspective
In the aftermath of every election, commentators vie to explain its meaning.
Winners rejoice and losers lament, both often arguing that the key to the outcome
was some specific factor supporting their point of view. In the aftermath of the
2016 election, sentiments like these were common among disappointed Clinton
supporters:

For anyone who voted for Donald Trump, bald-faced racism and sexism were not the
deal-breakers they should have been. Hatred of women was on the ballot in November,
and it won. (emphasis in original) 6

Donald Trump has won the presidency, despite an unprecedented level of unfitness and
in defiance of nearly every prediction and poll. And hes done this not despite but [sic]
because he expressed unfiltered disdain toward racial and religious minorities in the
country.7

As I will discuss in greater detail in the next essay, many disappointed Clinton
supporters made such claims and no doubt many more agreed with them. But imagine
an alternative universe in which the Clinton campaign followed Bill Clintons advice

Says. So What Are They? Washington Post, March9, 2016, https://www.washingtonpost.com/news/monkey

-cage/wp/2016/03/09/trumps-voters-arent-authoritarians-new-research-says-so-what-are-they/.

6Christina Cauterucci, In 2016 American Was Forced to Face the Reality of Sexual Assault, Slate, Decem-
ber28, 2016, www.slate.com/blogs/xx_factor/2016/12/28/_2016_was_the_year_america_learned_what
_sexual_assault_looks_like.html.

7Jenee Desmond-Harris, Trumps Win Is a Reminder of the Incredible, Unbeatable Power of Racism, Vox,
November9, 2016, www.vox.com/policy-and-politics/2016/11/9/13571676/trump-win-racism-power.

Hoover Institution Stanford University

4

to devote more attention and resources to the Rust Belt states, with the result that
Hillary Clinton gained 39,000 more votes distributed in such a way that she carried
Pennsylvania, Michigan, and Wisconsin (which Trump won by a bit less than 78,000).
This would have given her a comfortable Electoral College majority along with a clear
popular vote plurality. Then, in all likelihood, the day-after story line in the media
would have been, Americans reject racism and sexism!

The larger meaning of a presidential election should not hinge on the distribution
of .0006 of the vote in three states. Failing to appreciate that fact led to a
widespreadloss of perspective among election commentators in the aftermath
of the 2016 elections. In a majoritarian electoral system like ours, small changes
in the vote can have enormous consequences for party control of our governing
institutions and the policies they produce.8 Turning the sentence around,
consequential changes in control of our institutions produced by an election do
not imply that the electorate underwent any kind of sea change, which seems to
be the assumption made by many of those disappointed by the 2016 outcome. If
the US electoral system was a variation of a proportional representation system as
in most parliamentary democracies, then ceteris paribus, the verdict among the
commentaries would have been that 2016 was a status quo election that produced
no significant change from 2012.9

So, although there is no discounting the potential consequences of Trumps

victory,in itself it provides little basis for concluding that the election reflected
some sort of massive shift in the values and beliefs of the American public. As
in all elections, the vote reflected a combination of long-term conditions in the
country and short-term factors associated with the candidates and the campaigns.
In the case of 2016, a substantial portion of the electorate had become increasingly
dissatisfied with long-term developments. One candidate (Trump) was positioned
to capitalize on this dissatisfaction and the other one (Clinton) was not. Overlaid
on these long-term considerations were short-term factors, most important, the two
candidates.

8Conversely, large changes in the vote can have minimal consequences for institutional control and policy
change. Ronald Reagan gained 8percentage points in the popular vote between 1980 and 1984, but the large
Democratic majority in the House diminished by only sixteen seats and the narrow Republican majority in the
Senate fell by one.

9Parliamentary systems have no equivalent to our midterm elections.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 5

The Flight 93 Election10

As emphasized in essays 1 and 7, voters can choose only between the alternatives the
political parties offer them. If both parties nominate unacceptable candidates, voters
will elect an unacceptable candidate.11

The media tend to emphasize candidate personality characteristics. Is the candidate

authentic, warm, modest, sincere, trustworthy, and moral or their opposites? In
general, research indicates that candidate personality characteristics are overrated
as influences on the vote. In 1980, for example, voters thought that Jimmy Carter
was a peach of a guy personally and Ronald Reagan a somewhat scary prospect,
but that did not stop them from replacing what many viewed as a failed president
with a risky alternative.12 To a greater degree than usual, campaign coverage in 2016
revolved around the personas of the two candidates, especially Trumps. Indeed, some
Democratic critics of the Clinton campaign complained that it had too little substance
and focused too much on driving home the notion that Trump was a horrible human
being.13 Although I recognize the deep admiration for Clinton among her ardent
supporters, the data clearly indicate that a substantial portion of the American
electorate viewed the election as something akin to Alien vs. Predator.14 Rightfully or
wrongfully, the simple fact is that the American public saw both candidates as deeply
flawed. According to Gallup, Donald Trump and Hillary Clinton head into the final
hours of the 2016 presidential campaign with the worst election-eve images of any
major-party presidential candidates Gallup has measured back to 1956.15 Figure10.1

10Publius Decius Mus, The Flight 93 Election, CRB, September5, 2016, hwww.claremont.org/crb/basicpage
/the-flight-93-election/.

11Political scientists of a certain vintage will recognize the allusion to V. O. KeyJr., the great mid-twentieth-
century political scientist who wrote, If the people can choose only from among rascals, they are certain to
choose a rascal. The Responsible Electorate (New York: Vintage, 1966), 3.

12Morris Fiorina, Youre Likeable Enough, Mitt, New York Times, June7, 2012, campaignstops.blogs.nytimes
.com/2012/06/07/youre-likable-enough-mitt/. More generally, research finds that voters prioritize candidate
qualities that are relevant to governing rather than characteristics that determine whom theyd rather have a
beer with. See DavidB. Holian and CharlesL. Prysby, Candidate Character Traits in Presidential Elections (New
York: Routledge, 2015).

13A Wesleyan University study of campaign ads reported that the Clinton campaign was imbalanced in
just this way. Kyle Olson, STUDY: Hillary Campaign Most Negative, Least Substantive, The American Mirror,
March9, 2017, www.theamericanmirror.com/study-hillary-campaign-negative-least-substantive/

14https://www.youtube.com/watch?v=Xh1TwRilcLo.

15Lydia Saad, Trump and Clinton Finish with Historically Poor Images, Gallup, November8, 2016,
www.gallup.com/poll/197231/trump-clinton-finish-historically-poor-images.aspx.

Hoover Institution Stanford University

6

Figure 10.1. Unfavorable Ratings of Major Party Presidential Nominees:

19562016

70
60

% Unfavorable 50
40
30
20
10
0
Stevenson

Kennedy
Nixon
Johnson

Humphrey
Nixon
McGovern
Nixon

Ford

Reagan
Mondale
Reagan
Clinton
G.H.W. Bush
Kerry
G.W. Bush
Obama
McCain
Obama
Romney
Clinton
Trump
Eisenhower

Goldwater

Carter

Carter
graphs Gallups candidate Scalometer,which asks voters to rate the candidates
positively or negatively on a 1 to 5 scale.

Before Trump, the most negatively rated Republican candidate was Goldwater in 1964
with a 47-point unfavorable rating. Trump obliterated this long-standing record by
16points. Before Clinton, the most negatively rated Democratic candidate was McGovern
in 1972, with a 41-point unfavorable rating. Not to be outdone, Clinton nearly matched
Trumps record-shattering performance by topping McGoverns negatives by 14 points.16

Of course, voters can harbor positive or negative feelings about a candidate for
reasons other than their personasnamely, the candidates records, the positions
they advocate, the groups who endorse them, and other considerations. Various polls
provide more specific measures of candidates personal characteristics, although they
do not provide the longtime series of Gallup and ANES measures. Table10.1 compares
personal ratings of the 2016 candidates. The top panel reports ratings for all adults
from the Economist/YouGov survey and the bottom panel only for voters from the exit
polls, but the figures are very similar. For many Clinton supporters, her long record of
public service was a major reason to support her, but only half the electorate believed
that she was qualified to serve as president and had the right temperament to serve. A
possible reason for discounting her record is that nearly two-thirds of those who voted
considered her dishonest and untrustworthy. Not even one-third of voters believed
she was sincere in what she said: Trumps strongest point. Whether the reason was a

16On average candidates appear to be evaluated more negatively beginning in the 1980s. A possible explanation
is that because the Gallup measure captures policy and performance evaluations as well as personal qualities,
it would trend downward after the process of party sorting begins, as partisans expressed increasingly
negative evaluations of the other party. See essays 2 and 3.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 7

quarter-century-long Republican campaign Table 10.1. Americans Voted for Trump In

of character assassination and trumped-up Spite Of
scandals (as Democrats claimed) or the just
Trump Clinton
desserts from a quarter century of skating
on the boundary between the ethical and Economist/YouGov
unethical (as Republicans claimed) are not Qualified 34 49
the issue. Whatever the causes, the negative Honest and Trustworthy 31 24
numbers were the issue. Says what s/he believes 53 29
Like a lot 16 17

The figures for Trump are striking in their Dislike 47 53

implications. Only about a third of the Exit polls (voters only)

electorate considered him qualified to serve, Qualified 38 52

to have the right temperament to serve, and Honest and Trustworthy 33 36

to be honest and trustworthy. Only 16percent Right Temperament 35 55

Bring Needed Change 83 14
of the voters liked him a lot and nearly half
disliked him (Clintons figures were slightly
worse). Yet Trump received 46percent of the popular vote. Rather than an enthusiastic
endorsement of Trumps controversial comments and positions, the conclusion must be
that a significant number of Americans cast their vote for him in spite of their negative
views of him. As noted repeatedly in previous essays, a vote for a candidate does not
imply enthusiastic support, only that the voter thinks that candidate is preferable to
the alternative. Why was Trump preferable for voters who considered him unqualified
to serve? A strong hint from the exit polls comes from the reasons voters gave for their
decisions. By a significant margin they chose change over empathy, experience, and
judgment, and on that dimension Trump led Clintonthe candidate of continuityby
a margin of 6 to 1. Numerous liberal commentators embraced this interpretation, often
expressing it in vivid prose. For Thomas Frank, She was exactly the wrong candidate for
this angry, populist moment. An insider when the country was screaming for an outsider.
A technocrat who offered fine-tuning when the country wanted to take a sledge-hammer
to the machine.17 Former MSNBC commentator Krystal Ball wrote, Voters were offered
a choice between a possibility of catastrophe in Trump and a guarantee of mediocrity
in Clinton. Clearly, they picked the high-risk bet that they felt at least gave them some
chance to escape the certain economic doom that they feel in their current lives.18

17Thomas Frank, Donald Trump Is Moving to the White House, and Liberals Put Him There, The Guardian,
November9, 2016, https://www.theguardian.com/commentisfree/2016/nov/09/donald-trump-white-house
-hillary-clinton-liberals.

18Krystal Ball, The Democratic Party Deserved to Die, Huffington Post, November10, 2016, www.huffing
tonpost.com/entry/the-democratic-party-deserves-to-die_us_58236ad5e4b0aac62488cde5.

Hoover Institution Stanford University

8

The implication that many if not most votes for Trump did not reflect enthusiasm for
him as much as negative judgments about Clinton received clear support in various
polls. Throughout the campaign, only about 40percent of Trump supporters said that
they were voting for Trump rather than against Clinton.19 According to Harry Enten,
No candidate since 1980 has had a lower percentage of voters say they plan to cast
a votefortheir candidate. That includes candidates whose campaigns were viewed as
disastrous, including Jimmy Carter in 1980, Michael Dukakis in 1988, and Bob Dole
in 1996.20 Clinton did not fare much better. In the exit polls, 57percent of Americans
said they would have negative feelings if Trump won, but 53percent said the same
about Clinton.

Other things being equal, one would have expected an average Democrat to crush
a historically flawed candidate like Trump. Instead, Clintons margin over Trump
was lower than Obamas margin over Romney in thirty-seven states. In my view,
explanations of why the Democrats lost the presidency in 2016 focus too much on
Trump and not enough on Clinton. Earlier in the summer of 2016, when Republican
acquaintances expressed the hope that the FBI would recommend an indictment
of Hillary Clinton, I cautioned them to be careful what they wished for, suggesting
that this would be the worst possible outcome for Republicans. The likely last-minute
replacement on the ticket would be Joe Biden, and in all likelihood he would win the
election and with it the Senate. Biden is largely scandal-free, a pauper by senatorial
standards, and his background and record appeal to precisely the segment of the
electorate that defected from Obama to Trump. Some believe the same is true for
Bernie Sanders as well, although that is a harder case to make given some of his
economic views.

The Split between the Electoral College and the Popular Vote
For the second time in sixteen years, the popular vote leader did not win an Electoral
College majority. When this happened in 2000, many political analysts expected
much more of a negative popular reaction than the limited one that ensued. Most
Americans who were not locked into the two partisan camps seemed to accept the
sports analogies that were offered: in the World Series and the NBA playoffs the winner
is determined by games won, not the most runs or points. In the 1960 World Series,
for example, the New York Yankees outscored the Pittsburgh Pirates 55 to 27; but

19Diminished Enthusiasm Dogs Trump: Clinton Gains in Affirmative Support, ABC News, October24, 2016,
www.langerresearch.com/wp-content/uploads/1184a22016ElectionTrackingNo.2.pdf.

20Clinton Voters Arent Just Voting against Trump, FiveThirtyEight, October25, 2016, fivethirtyeight.com
/features/clinton-voters-arent-just-voting-against-trump/.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 9

the Pirates won the series four games to three, and no one questioned the outcome.
The same analogies were offered in 2016 after Clinton supporters railed that she had
legitimately won the election but lost in an undemocratic vestige of an eighteenth-
century political compromise.

I am of two minds about such analogies. When the topic is the legitimacy of elections
and the governments they determine, such analogies are inapt. Elections are the way
democracies determine a legitimate government. The political equality embodied
in majority rule is the most fundamental component of a democratic form of
government. So if the rules of the electoral game crown a candidate who gets fewer
votes than an opponent, it violates political equality and undercuts the legitimacy of
the winner. I see no way around that conclusion.

As a description of the way to play the electoral game, however, sports analogies are
apt. Many Clinton supporters claim that she would now be president had the election
been based on the popular vote. Probably they are correct, but we can never know for
sure because one cannot assume the popular vote would have been the same absent the
Electoral College: the candidates would have conducted different campaigns. From the
standpoint of the Trump campaign it mattered not at all whether he lost California by
270 votes or by the 4,270,000 that he actually did. The 4-million-plus vote margin that
Clinton racked up in California was irrelevant under the Electoral College rules. Reports
indicate that the Trump campaign set aside the states certain to go either Republican
or Democratic and concentrated on the sub-election occurring in the thirteen
battleground states.21 Trump won that sub-election by about 800,000 votes.22 Had the
outcome been determined by popular vote, however, Trump would have made more of
an effort in friendly areas of states like California and New York, and Clinton in friendly
areas of deep red states. How it all would have netted out is the kind of counterfactual
that sparks interesting discussions but is probably beyond resolution with data.

The Polls and Models

In the immediate aftermath of the election, a number of columnists concluded that

the polls blew it.23 But as Clintons popular vote plurality mounted, analysts realized

21John Judis, On the Eve of Disruption: Final Thoughts on the 2016 Election, TPM, December18, 2016,
talkingpointsmemo.com/edblog/100887.

222016 National Popular Voter Tracker, undated, https://docs.google.com/spreadsheets/d/133Eb4qQmOxN

vtesw2hdVns073R68EZx4SfCnP4IGQf8/edit#gid=19.

23Michelle Jamrisko and Terrence Dopp, Failed Polls in 2016 Call into Question a Professions Precepts,
Bloomberg, November8, 2016, https://www.bloomberg.com/politics/articles/2016-11-09/failed-polls-in

Hoover Institution Stanford University

10

that the polls had performed pretty well. Most of the major national polls pegged
Clintons lead at 24percentage points, and she ultimately won the popular vote by
about 2 points. So, the polls slightly overestimated Clintons lead, but that overestimate
is actually a bit smaller than the underestimate of Obamas 2012 vote.24 A bigger
problem was the exaggerated certainty that various polls and models gave to a Clinton
win. Although criticized earlier for giving Trump as high a probability of winning
as 30percent, Nate Silver at 538 turned out to be closer to the mark than Huff Posts
98percent and Sam Wangs 99percent predicted probabilities of a Clinton victory.25

The state polls, on the other hand, revealed some problems. On average they tended
to underestimate Trumps support, particularly in the states he ultimately carried
including, critically, Pennsylvania, Michigan, and Wisconsin, where 46 point poll
advantages for Clinton the week before the election evaporated on Election Day.26 The
polling disparities were correlated with the states proportion of voters without college
degrees.27 The polls may have under-sampled white noncollege voters, underestimated
their likelihood of voting, or both.

A great deal of ink and airtime were devoted to predictive models during the
campaign. There are two kinds of election models, although they overlap a bit. 28

-2016-call-into-question-a-profession-s-precepts; Andrew Mercer, Claude Deane, and Kyley McGenney, Why

2016 Election Polls Missed Their Mark, Pew Research Center, November9, 2016, www.pewresearch.org/fact
-tank/2016/11/09/why-2016-election-polls-missed-their-mark/.

24Sean Trende, It Wasnt the Polls that Missed, It Was the Pundits, Real Clear Politics, November12, 2016,
www.realclearpolitics.com/articles/2016/11/12/it_wasnt_the_polls_that_missed_it_was_the_pundits_132333
.html.

25Who Will Win the Presidency? FiveThirtyEight, November8, 2016, https://projects.fivethirtyeight

.com/2016-election-forecast/#plus; Forecast, Huffington Post, November8, 2016, elections.huffingtonpost
.com/2016/forecast/president; Is 99% a Reasonable Probability? Princeton Election Consortium, November6,
2016, election.princeton.edu/2016/11/06/is-99-a-reasonable-probability/. I have not delved into the guts of
these models, but any methodology that generates an election prediction with confidence levels in the range
of 9899percent strikes me as prima facie problematic.

26Dhrumil Metta, How Much the Polls Missed by in Every State, FiveThirtyEight, December2, 2016,
fivethirtyeight.com/features/how-much-the-polls-missed-by-in-every-state/; David Weigel, State Pollsters,
Pummeled by 2016, Analyze What Went Wrong, Washington Post, December30, 2016, https://www
.washingtonpost.com/news/post-politics/wp/2016/12/30/state-pollsters-pummeled-by-2016-analyze-what
-went-wrong/?utm_term=.e4e8d652d728.

27Nate Silver, Pollsters Probably Didnt Talk to Enough White Voters without College Degrees, FiveThirtyEight,
December1, 2016, fivethirtyeight.com/features/pollsters-probably-didnt-talk-to-enough-white-voters-without
-college-degrees/.

28Actually, three, but models of the third typeformal or mathematicalare too abstract to make empirical
predictions about an election.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 11

Table 10.2. 2016 Election Forecasting Models

Predicted Two-Party Certainty of Popular

Popular Vote for Clinton Vote Majority

Erikson & Wlezien (economic indicators/polls) 52 82%

Lockerbie (economic expectations/1stterm) 50.4 62%
Lewis-Beck & Tien (approval/growth) 51.1 83%
Campbell (convention/growth) 51.2 75%
Abramowitz (approval/growth/1stterm) 48.6 66%
Norpoth (primaries) 47.5 87%

Source:
James Campbell. Forecasting the 2016 American National Elections.PS: Political Science & Politics 49 (2016): 652.

The first type, discussed above, is poll-based, although it aggregates, evaluates,

simulates, standardizes, and in other ways tries to extract more accurate information
from the universe of polls. The second type consists of political science forecasting
models that get less media attention, probably in part because they omit the subjects
most dear to the hearts of journalists: the candidates and campaigns. The political
science models are generally based on the so-called fundamentals: chiefly peace
and war and the state of the economy, which are viewed as factors that set the
election context and determine the kinds of campaigns that can be conducted.29
For the most part these models do not include the candidates; the latter are implicitly
assumed to account for only a little variation on the margins.30 A few of these models
incorporate some poll data, just as the 538 polls plus model includes some aspects of the
fundamentals. But for the most part these models implicitly assume that the election is
often determined before the campaign formally begins or in some cases even before the
candidates are nominated. Table10.2 summarizes the principal forecast models for 2016.

Most of the models predicted a narrow Democratic edge in the two-party vote,
although two models went the other way.31 None of them calculated the same degree

29Lynn Vavreck, The Message Matters (Princeton, NJ: Princeton University Press, 2009).

30Its not that the campaigns and candidates are irrelevant, but the assumption is that both candidates
fall within a range of acceptability and that both campaigns will have access to roughly equal resources and
expertise. Thus, the candidates and campaigns generally offset, leaving the election to reflect mostly the
underlying fundamentals.

31Abramowitz predicted a lower Clinton vote because his time for a change model includes a variable for an
incumbent party that has held office for two terms. Despite the prediction of his model, Abramowitz expressed
doubt about the forecast because he viewed Trump as outside the bounds of acceptability. AlanI. Abramowitz,
Will Time for Change Mean Time for Trump? PS: Political Science and Politics 49 (October2016): 659660.

Hoover Institution Stanford University

12

of confidence (figures in parentheses) as the polling models did. In this situation, the
most reasonable interpretation is that the fundamentals indicated a fifty-fifty election
that would be determined by the marginal effects of the candidates and campaigns. Thus,
the verdict is that they performed decently, although Brady and Parker report that the
accuracy of some of the best known economic models has been declining since 1992.32

Swing Voters
Analysts define swing voters in various ways.33 At its broadest the concept excludes only
those who are sure to turn out and sure to vote for one of the parties and not the other,
leaving as swing voters everyone who is uncertain about whom they will vote for and/or
whether they will vote at all. The conventional wisdom is that the proportion of swing
voters in the American electorate has greatly declined, but in essay 1 and various earlier
writings I have argued that swing voters are made, not born.34 To reiterate, todays sorted
parties nominate candidates who look similar to the ones they nominated in previous
elections, so most voters will probably vote the same as they did in previous elections.
Rather than being less willing to move between parties than voters in earlier decades,
contemporary voters may simply have less reason to do so. The three members of the
Democratic sequence of Gore, Kerry, and Obama from 20002008 looked a lot more
similar to each other than did the members of the Humphrey, McGovern, and Carter
sequence of 196872. On the Republican side, Bush, McCain, and Romney in 200412
looked considerably more similar to each other than the sequence of Eisenhower, Nixon,
and Goldwater in 195664. Other things being equal, party sorting at the candidate
level should produce less voter swinging in contemporary elections than in the late
twentieth century, even if voters were just as willing to swing now as they were then.

As argued in essay 4, however, to the dismay of the Republican thought establishment,

Trump broke the mold of recent Republican nominees. He demonstrated the appeal
of a de-sorter in the primaries. By the time the general election campaign was under
way he was a full-scale disrupter, in Silicon Valley jargon.35 This qualitative impression

32David Brady and Brett Parker, Now Is the Winter of Our Discontent: The 2016 U.S. Presidential Election
(Basingstoke, UK: Palgrave Macmillan, forthcoming).

33WilliamG. Mayer, ed., The Swing Voter in American Politics (Washington, DC: Brookings, 2008).

34MorrisP. Fiorina, If I Could Hold a Seminar for Political Journalists... The Forum 10 (no4): 210,
https://www.degruyter.com/downloadpdf/j/for.2012.10.issue-4/forum-2013-0011/forum-2013-0011.xml.

35Peggy Noonan, Trump Tries to Build a Different Party, Wall Street Journal, January26, 2017, www.wsj.com
/articles/trump-tries-to-build-a-different-party-1485478386; EdwardG. Carmines, MichaelJ. Ensley, and
MichaelW. Wagner, Ideological Heterogeneity and the Rise of Donald Trump, The Forum 14 (no.4), https://
www.degruyter.com/view/j/for.2016.14.issue-4/for-2016-0036/for-2016-0036.xml?format=INT.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 13

is consistent with the quantitative data about swing voters in 2016. Numerous
commentators noted the large increases in the number of voters who indicated
they were undecided, intended to vote for third-party candidates, or claimed that
they would vote but not for president. An Economist/YouGov survey in July and
August2016 found 31percent of voters distributed roughly evenly across those three
categories. In contrast, most commentators put the numbers of such voters in the lower
single digits in 2012.36 Some analysts correctly cautioned that the large number of
voters potentially in play meant that the 2016 election had more underlying volatility
than other recent elections. Preliminary analyses suggest that late deciders were
slightly more likely to go to Trump than to Clinton and that those who changed their
minds late in the campaign also did so.37 The numbers were not large, but they may
have eroded Clintons margin in key areas. Polls also found Republicans registering
gains in the generic congressional vote in the last week of the campaign, suggesting a
general movement in the Republican direction.38

The Media
On several subjects, the weight of social science research conflicts with what is widely
believed in popular, journalistic, and political circles. Two prominentand related
examples are the ability of the mass media to shape public opinion and the impact of
campaign events. Surveys show that Americans believe the media are very powerful
hence the concern about bias in the mediaand that people less sophisticated than
they are (that is, most of those on the other side) are easily manipulated by slanted
news and sophisticated ads. Naturally, people who work for the media or make their
livings producing political ads believe that the media are very important. But as
communications scholar Diana Mutz observed, scholarly research does not support such
claims of major media influence; there is an enormous chasm between the beliefs
held by journalists (and the typical voter) about the effects of campaign media and the
findings of political communications scholars. Public perceptions of the power of media
in elections, and the academic evidence of its influence, could not be further apart.39

362016 figures cited in Brady and Parker, The Winter of our Discontent 29. On 2012 see ReidJ. Epstein, The
Disappearing Undecided Voter, Politico, undated, www.politico.com/news/stories/0812/79504_Page3.html.

37Dan Hopkins, Voters Really Did Switch to Trump at the Last Minute, FiveThirtyEight, December20, 2016,
fivethirtyeight.com/features/voters-really-did-switch-to-trump-at-the-last-minute/.

38The generic vote item reads, This November, do you plan to vote for a Democratic or a Republican candidate
in your congressional district? Harry Enten, Senate Update: The Generic Ballot Is Hurting Democrats Chances,
FiveThirtyEight, November7, 2016, fivethirtyeight.com/features/senate-update-the-generic-ballot-is-hurting
-democrats-chances/.

39Diana Mutz, The Great Divide: Campaign Media in the American Mind, Daedalus 141 (no.4): 8387.

Hoover Institution Stanford University

14

If ever there were an election designed to pit the popular belief of major media influence
against the scholarly consensus of minor influence, 2016 was it. The legacy media were
almost unanimous in opposing Trump. For many of them, their editorial positions
spilled over onto the news pages, with the apparent approval of their editors.40 The tone
of their op-ed and other editorial efforts seemed to grow more frantic as Clinton failed
to pull away from Trump in the polls. Some newspapers endorsed a Democrat for the
first time in memory or history.41 After every embarrassing revelation or outrageous
comment, the media chorus would pronounce the end of the Trump campaign, and still
he marched on to win. Thomas Frank asked, How did the journalists crusade fail? The
fourth estate came together in an unprecedented professional consensus.42

As the next essay will discuss, a plausible hypothesis is that a strong current of
antielitism was running in the 2016 election and that those associated with the national
media are viewed as card-carrying members of the elite. Along with other institutions,
trust in the media has declined precipitously in recent decades (figure10.2). Why should
voters be influenced by people they distrust? Did the medias recent record merit trust?
Columnist Matt Bai expressed this sentiment in an acerbic passage:

But Trump had figured out that no one really believed the elite media anymorethe
same media that said Iraq was an existential threat, that the banks had to be saved, that
Obama would transform our dysfunctional politics. The same media that nightly featured
a cavalcade of smug morons whose only qualification to opine on TV was an almost
pathological shamelessness.

Bai added,

Because this is what he [Trump] learned from his first-ever campaign experiencethat if
you pit yourself against powerful agencies or politicians or a corrupt media, people now
will believe almost anything. Or maybe they wont really care what youre saying, as long
as its infuriating to the so-called experts.43

40Rutenberg, Trump Is Testing the Norms.

41Hannah Levintova, This Newspaper Just Endorsed Its First Democrat for President in Almost a Century,
Mother Jones, September23, 2016, www.motherjones.com/politics/2016/09/donald-trump-just-forced-yet
-another-newspaper-endorse-democrat-first-time-almost-c; Rebecca Shapiro, For the First Time In its 126-Year
History, This Newspaper Endorsed a Democrat, Huffington Post, September28, 2016, www.huffingtonpost
.com/entry/arizona-republic-hillary-clinton-endorsement_us_57eb284fe4b024a52d2b7437.

42Frank, Donald Trump Is Moving to the White House.

43Matt Bai, The Moment That Made Trump Possible, Yahoo! News, December15, 2016, https://www
.yahoo.com/news/the-moment-that-made-trump-possible-100008601.html.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 15

Figure 10.2. Trust in Media Hits New Lows

80
70
60
50
40
30
20
10
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Great deal/Fair amount Not very much/Not at all
Source:
Jeff Jones and Lydia Saad. Gallup Poll Social Series: Governance File.

Thomas Franks answer to why the journalists crusade failed is similar: They [the
media] chose insulting the other side over trying to understand what motivated them.
They transformed opinion writing into a vehicle for high moral boasting. What could
possibly have gone wrong with such an approach?44

I suspect future research will find that there is considerable support for such
arguments. Media opposition to Trump may have backfired. The more unbalanced
the opposition to Trump became, the more some voters were tempted to strike back at
people and institutions they resented.45 The point here is not whether Trump deserved
unbalanced treatment but whether in the eyes of some voters the media had overdone
it. The next essay will return to this subject.

The obvious objection to the preceding line of argument is that the media were more
important than ever in 2016, but they were different media. The era of the New York
Times and Washington Post, the broadcast TV channels, and even the cable channels
has passed; the information world belongs now to Twitter, Facebook, and their ilk.
With his background in popular entertainment, Trump exploited the rapidly evolving

44Frank, Donald Trump Is Moving to the White House.

45After the election, I talked to an editor from a major national newspaper who related that his paper had
invited Trump voters to e-mail their reasons for supporting Trump. One of the most common responses was
some version of I wanted to see your heads explode.

Hoover Institution Stanford University

16

media environment while the Clinton campaign was spending millions of dollars
on adsso twentieth-century, as our grandchildren might say. Numerous analyses
document that Trump received billions of dollars in free media during the primary
and general election campaigns, considerably offsetting the huge funding advantage
enjoyed by the Clinton campaign.46 Trump was an inveterate Twitterer, and many of
his tweets were amplified by media coverage (even if much of the latter was largely
negative in tone).47 There are claims that Trumps son-in-law, Jared Kushner, ran a
highly sophisticated analytics and social media operation that flew under the radar of
the mainstream media and the Clinton campaign, an operation that helped to explain
the better-than-expected showing for Trump in critical areas.48 Journalists and politicos
focused on traditional metricsthe money spent, the field offices opened, and the
workers deployed by the Clinton campaignbut they had no way of evaluating the
impact of Trumps tweets, Facebook presence, Reddit groups, and so on, if they were
even aware of most of these activities.

Communications scholars are diligently examining these questions as I write.

But until data prove otherwise, I remain skeptical of the Twitter effect at least.
Political journalists need to realize that nearly everyone they talk to is abnormal:
abnormally interested, abnormally well informed, and abnormally opinionated.
Relatively few normal people live in the Twitterverse that political journalists
and other members of the chattering class inhabit. According to a Pew Research
Center study conducted just as the 2016 primary season got under way, 16percent
of US adults claimed to be on Twitter, but only 9percent reported getting news
from Twitter.49 In another study, Allcott and Gentzkow reported that in 2016 only
14percent of the public relied on social media as the most important source of

46Nicholas Confessore and Karen Yourish, $2 Billion Worth of Free Media for Donald Trump, New York
Times, March15, 2016, www.nytimes.com/2016/03/16/upshot/measuring-donald-trumps-mammoth-
advantage-in-free-media.html?_r=0; Jason Le Miere, Did the Media Help Donald Trump Win? $5 Billion
inFree Advertising Given to President-Elect, International Business Times, November9, 2016, www.ibtimes.
com/did-media-help-donald-trump-win-5-billion-free-advertising-given-president-elect-2444115.

47Why arent such posts called twits rather than tweets? The former term seems more accurate.

48Stephen Bertoni, Exclusive Interview: How Jared Kushner Won Trump the White House, Forbes,
December20, 2016, www.forbes.com/sites/stevenbertoni/2016/11/22/exclusive-interview-how-jared-kushner
-won-trump-the-white-house/#23e54fea2f50.

49Jeffrey Gottfried and Elisa Shearer, News Use Across Social Media Platforms 2016, Pew Research Center,
May26, 2016, www.journalism.org/2016/05/26/news-use-across-social-media-platforms-2016/201, noting that
67percent of adults use Facebook, but only 44percent said it was a source of news. A study one year earlier
found that 55percent of adults said they get news from neither Facebook nor Twitter. Amy Mitchell and Dana Page,
The Evolving Role of News on Twitter and Facebook, Pew Research Center, July14, 2015, www.journalism.org
/files/2015/07/Twitter-and-News-Survey-Report-FINAL2.pdf.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 17

news. Moreover, as discussed in essay 2, Table 10.3

most Americans have an expansive concept
Millions of Twitter Followers
of what is news.50 In the Pew study noted
above, the news included sports, science Katy Perry 95.4
and technology, local weather and traffic, Justin Bieber 91.3

entertainment, crime, people and events in Barack Obama 80.7

Britney Spears 50
your community, and health and medicine.
Kim Kardashian 49.6
In sum, only a small fraction of the small
CNN Breaking News 45.4
minority of Americans with Twitter accounts
LeBron James 34.1
follow what we call hard news. New York Times 33
ESPN 30.5
The Trump campaign claimed 20 million Bruno Mars 27.7
Twitter followers, a claim difficult to fact- NBA 24.1
check because Twitter audiences are difficult Pitbull 23.3
to measure with any degree of accuracy. One Kourtney Kardashian 21.7

analyst calculates that after adjusting for NASA 21.2

Salman Kahn 20.5
accounts that are inactive, zombies, or held by
Zayn 20.7
foreigners, the number of (American) Trump
Donald Trump 16.9
followers is probably closer to 4 million.51
Moreover, most Twitter followers do not see
what is tweeted.52 But even taking highly inflated numbers as given, one early 2017
report put Trumps followers (17 million) as a small fraction of those who follow Katie
Perry (95 million) or Kim Kardashian (49 million), LeBron James (34 million), Pitbull
(who?) (23 million), and even Zayn (who?) (21 million).53

Comey, Russians, Hollywood Access, Fake News, Etc.

Scholarly research and popular beliefs also conflict on a second subject: the importance
of campaign events that receive so much coverage in the traditional media. Journalists
tend to view campaigns like an athletic contest where a single excellent or disastrous
play is a game changer that will shift momentum and ultimately be the turning point

50Hunt Allcott and Matthew Gentzkow, Social Media and Fake News in the 2016 Election, working paper,
January18, 2017, https://web.stanford.edu/~gentzkow/research/fakenews.pdf.

51Rob Salkowitz, Trumps 20 Million Twitter Followers Get Smaller under the Microscope, Forbes, January17,
2017, www.forbes.com/sites/robsalkowitz/2017/01/17/trumps-20-million-twitter-followers-get-smaller-under-the
-microscope/#6084f458675b.

52Danny Sullivan, Just Like Facebook, Twitters New Impression Stats Suggest Few Followers See Whats
Tweeted, Marketing Land, July1, 2014, marketingland.com/facebook-twitter-impressions-90878.

53Twitter: Most Followed, Friend or Follow, undated, friendorfollow.com/twitter/most-followers/.

Hoover Institution Stanford University

18

in determining the winner.54 Once again, political scientists tend to be skeptical. Careful
empirical research finds little evidence for such game changers.55 Campaign effects tend
to be small and transitory, in part because few people are paying much attention to them.

In 2016, this difference in popular and scholarly perspective emerged immediately

after it became apparent that Trump would be president. Clinton supporters blamed
the Comey letter, Wikileaks, Russian interference, and fake news, among other
things. In a very close election, almost everything matters, of course, but one cannot
pick and choose the events that went against your side and ignore those events and
developments that went against the other side.56 Determining whether campaign
events mattered is particularly difficult in this case. The real question is whether they
swung Michigan, Wisconsin, and Pennsylvania into Trumps camp. Moreover, there is
a tendency for partisans to come home during the course of the campaign. Thus,
Trumps numbers were expected to improve simply because his support among
Republicans had more room to grow than did Clintons among Democrats. Preliminary
evidence is conflicting. Some see evidence of a Comey effect in state and national
polls and early voting numbers.57 Others see no close correspondence between poll
trends and major campaign developments: Clintons numbers were rising when the
Wikileaks releases were at their peak, and Trumps numbers had begun to rise before
the Comey letter was released.58 Yglesias argued that the problem faced by the Clinton
campaign was real news, not fake news.59

Scholars are marshalling every bit of data they can get their hands on to study the
questions arising from the 2016 campaign. Whether they can provide definitive
answers remains to be seen. Consistent with past research, the first comprehensive
study reported minimal campaign effects. After an intensive analysis of three different

54John Heilemann and Mark Halperin, Game Change (New York: HarperCollins, 2010).

55John Sides and Lynn Vavreck, The Gamble (Princeton, NJ: Princeton University Press, 2013).

56After the election, journalist Carl Cannon identified thirty-one reasons why Trump won. How Donald
Trump Won, Real Clear Politics, November10, 2016, www.realclearpolitics.com/articles/2016/11/10/how
_donald_trump_won_132321.html

57Sean McElwee, Matt McDermott, and Will Jordan, 4 Pieces of Evidence Showing FBI Director James
Comey Cost Clinton the Election, Vox, January11, 2017, www.vox.com/the-big-idea/2017/1/11/14215930
/comey-email-election-clinton-campaign; Nate Silver, How Much Did Comey Hurt Clintons Chances?
FiveThirtyEight, https://fivethirtyeight.com/features/how-much-did-comey-hurt-clintons-chances/.

58Harry Enten, How Much Did Wikileaks Hurt Hillary Clinton? FiveThirtyEight, November6, 2016, fivethirtyeight
.com/features/wikileaks-hillary-clinton/.

59Matthew Yglesias, Fake News Is a Convenient Scapegoat, but the Big 2016 Problem Was the Real News,
Vox, December5, 2015, www.vox.com/policy-and-politics/2016/12/15/13955108/fake-news-2016.

Morris P. Fiorina The 2016 Presidential Electionan Abundance of Controversies

 19

databases, Allcott and Gentzkow concluded that even the most widely circulated
fake news stories were seen by only a small fraction of Americans. Considering the
counterfactual that fake news caused Hillary Clinton to lose Pennsylvania, Michigan,
and Wisconsin, they calculated that For fake news to have changed the election
outcome, one fake news article would need to be 36 times as persuasive as one
political ad.60 Based on previous scholarly research, my expectation would be that
campaign events in 2016 had little or no net effect on the outcome, subject to the
aforementioned caveat that in such a close election everything mattered if everything
else but that one thing is held constant. That is not to say that events like the Russian
connection may not have important consequencesinvestigations are ongoing as I
writebut only that the record of past research suggests that such campaign events
had little effect on the election itself.

60Allcott and Gentzkow, Social Media and Fake News in the 2016 Elections.

Essay Series

An Era of Tenuous Majorities: A Historical Context

Has the American Public Polarized?

The Political Parties Have Sorted

Party Sorting and Democratic Politics

The Temptation to Overreach

Independents: The Marginal Members of an Electoral Majority

The (Re)Nationalization of Congressional Elections

Is the US Experience Exceptional?

A Historical Perspective

The 2016 ElectionAn Abundance of Controversies

The 2016 Presidential ElectionIdentities, Class and Culture, Issues

The publisher has made this work available under a Creative Commons Attribution-NoDerivs license 3.0. To view a copy
of this license, visit http://creativecommons.org/licenses/by-nd/3.0.

Hoover Institution Press assumes no responsibility for the persistence or accuracy of URLs for external or third-party
Internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will
remain, accurate or appropriate.

Copyright 2017 by the Board of Trustees of the Leland Stanford Junior University

Hoover Institution Stanford University

 Series Overview
In contrast to most of modern American political history, partisan
control of our national elective institutions has been unusually
tenuous during the past several decades. This essay series argues that
the ideologically sorted parties that contest elections today face strong
internal pressures to overreach, by which I mean emphasizing issues and
advocating positions strongly supported by the party base but which
cause the marginal members of their electoral coalitions to defect.
About the Author
Thus, electoral losses predictably follow electoral victories. Institutional
control is fleeting.

The first group of essays describes the contemporary American

electorate. Despite myriad claims to the contrary, the data show that
the electorate is no more polarized now than it was in the later decades
of the twentieth century. What has happened is that the parties have
sorted so that each party is more homogeneous than in the twentieth
century; liberal Republicans and conservative Democrats have largely
passed from the political scene. The muddled middle is as large as ever
but has no home in either party. The growth in the proportion of self-
identified independents may be a reflection of the limited appeal of
MorrisP. Fiorina
todays sorted parties.
Morris Fiorina is the Wendt Family
Professor of Political Science at The second group of essays develops the overreach argument, discusses
Stanford University and a senior the role of independents as the marginal members of an electoral
fellow at the Hoover Institution. For majority, and explains how party sorting produces less split-ticket
more than four decades he has voting. Rather than most voters being more set in their partisan
written on American politics with allegiances than a generation ago, they may simply have less reason to
particular emphasis on elections split their tickets when almost all Democratic candidates are liberals and
and public opinion. Fiorina has all Republican candidates are conservatives.
written or edited twelve books and The third group of essays embeds contemporary American politics in
more than 100 articles, served as two other contexts. First, in a comparative context, developments in
chairman of the Board of the the European democracies are the mirror image of those in the United
American National Election Studies, States: the major European parties have depolarized or de-sorted or
and received the WarrenE. Miller both, whereas their national electorates show little change. The rise of
Career Achievement Award from anti-immigrant parties may have some as yet not well-understood role
the American Political Science in these developments. Second, in a historical context, the instability of
Association Section on Elections, American majorities today resembles that of the late nineteenth century,
Public Opinion, and Voting when similar significant social and economic changes were occurring.
Behavior. His widely noted book
Culture War? The Myth of a Polarized Two final postelection essays will wrap up the series.
America (with Samuel Abrams and
Jeremy Pope) is thought to have
influenced then Illinois state
senator Barack Obamas keynote
These essays naturally draw on the work of many people who have
speech to the 2004 Democratic
contributed to a very active research program. I thank colleagues John
National Convention (We coach
Aldrich, Douglas Ahler, Paul Beck, Bruce Cain, James Campbell, Shanto
Little League in the blue states, and,
yes, weve got some gay friends in
Iyengar, Matthew Levendusky, Sandy Maisel, Paul Sniderman, and
the red states). Guarav Sood, whose questions forced me to sharpen various arguments;
and David Brady in particular for almost daily conversations about the
matters covered in the posts that follow.

Hoover Institution, Stanford University Hoover Institution in Washington

434 Galvez Mall The Johnson Center
Stanford, CA 94305-6003 1399 New York Avenue NW, Suite 500
650-723-1754 Washington, DC 20005
202-760-3200