A Trump Effect on the EU's Popularity? The U.S. Presidential Election as a Natural Experiment

Abstract

Did the election of Donald Trump affect the popularity of the European Union (EU) in Europe? Theoretically, both a positive rally effect (due to a perceived external threat) and a negative domino effect (due to resignation among Europhiles and/or reinforcement among europhobe nationalists) are thinkable. We treat Trump's unexpected victory as an external shock and use a Eurobarometer survey that was conducted in all EU-28 member states four days prior to (control group) and six days after the election (treatment group) as source material for a natural experiment. The analysis reveals that the election of Trump caused a significant increase in the EU's popularity in Europe immediately after the election. This "Trump effect" is considerable in size, roughly equivalent to three years of education. Gains in popularity were particularly high among respondents who perceived their country as economically struggling and, surprisingly, among the political right, suggesting that Trump's victory broadened and ideologically diversified the EU's base of support.

Full text

A Trump Effect on the EU’s Popularity?
The U.S. Presidential Election as a Natural Experiment
Lara Minkus1*, Emanuel Deutschmann2, Jan Delhey3
1 Bremen International Graduate School of Social Sciences, Germany, 2 European University Institute, Italy, 3 Otto
von Guericke University Magdeburg, Germany, *Corresponding author: lminkus@uni-bremen.de
Forthcoming at Perspectives on Politics. Please check publication status and cite the journal
version, if already available.
Abstract: Did the election of Donald Trump affect the popularity of the European Union (EU)
in Europe? Theoretically, both a positive rally effect (due to a perceived external threat) and a
negative domino effect (due to resignation among Europhiles and/or reinforcement among
europhobe nationalists) are thinkable. We treat Trump’s unexpected victory as an external
shock and use a Eurobarometer survey that was conducted in all EU-28 member states four
days prior to (control group) and six days after the election (treatment group) as source material
for a natural experiment. The analysis reveals that the election of Trump caused a significant
increase in the EU’s popularity in Europe immediately after the election. This “Trump effect”
is considerable in size, roughly equivalent to three years of education. Gains in popularity were
particularly high among respondents who perceived their country as economically struggling
and, surprisingly, among the political right, suggesting that Trump’s victory broadened and
ideologically diversified the EU’s base of support.
Key words: U.S. election, Donald Trump, European integration, EU support, public opinion,
natural experiment
Introduction
In the night between November 8 and 9, 2016, the election of Donald Trump as the 45th
president of the United States took pollsters by surprise. The New York Times’s prediction, for
instance—followed by a global audience on the newspaper’s website during election night—
made an exorbitant swing: in a matter of hours, Trump’s “chance of winning the presidency”
rose from a mere 15 percent at 1:10 UCT to 95 percent at 3:56 UCT, while Hillary Clinton’s
chance dropped accordingly (Figure 1A). And the New York Times’s miscalculation was no
exception: 14 out of 15 national polls conducted in the United States during the first week of
November predicted a Clinton victory.1 The public, too, appeared surprised by the Democratic
candidate’s defeat. Immediately after the unexpected election outcome, Internet searches for
“Donald Trump” skyrocketed, as Google Trends data reveal (Figure 1B).

2
Figure 1. The election of Donald Trump as an “external shock.”
Figure 2. Relative frequency of observations (Eurobarometer 86.2)
Note to Figure 1 and 2: Figure 1A combines scraped data from two New York Times websites (nyti.ms/2miH2qZ
and nyti.ms/2gOIYTY, last accessed 8/8/2017); Figure 1B is based on Google Trends (trends.google.com, last
accessed 10/10/2017). Dashed lines in Figures 1B denote trends for the 27 EU countries in our sample, while the
solid lines denote the average across these countries. Dashed line in Figure 2 denote all countries available in
Eurobarometer 86.2, solid line denotes their average. Deviations to the actual final sample exist (see Data and
Sample section below).

3
To many European observers, Trump’s victory came not only as a surprise2 but also as
a shock. Apart from a general dismay vis-à-vis Trump’s apparent misogynism3, anapirism4,
xenophobia5, and Islamophobia6, many Europeans suspected that his election would affect
transatlantic relations and prospects of European integration. Such concerns were fueled by
Trump’s nationalist credo “America first,” condemnatory statements about NATO and TTIP,
and consentient comments regarding the looming partial breakup of the EU in the wake of
Brexit7. A need to reinvent Europe’s role in the world was one fear; a reinforcement of right-
wing populist parties in Europe and according consequences for upcoming national elections
was another. It has been argued that “[d]ramatic and extraordinary real-world events have the
power to impact on public opinion and to cause shift in public attitudes.”8 Trump’s election
certainly constitutes such an event, raising the question of whether this political earthquake in
North America has led to tectonic shifts in public opinion on another continent. Specifically:
did Trump’s surprise victory affect the EU’s popularity in Europe?
In search of an answer, this study takes the “shock” element described above literally
and treats the U.S. election as a natural experiment in which the victory of Trump constitutes
an external shock in the sense of experimental research. We are not the first to have this idea.
Silver argued that “the May and December [2016] elections in Austria made for an interesting
controlled experiment. The same two candidates were on the ballot, but in the intervening
period Trump had won the American election.”9 However, months and months in which
countless potentially significant events happened over and above Trump’s victory lie between
these two measurement points, making Silver’s claim contestable. His “setup” is not actually
“controlled,” and causal claims are accordingly hard to make. By contrast, our paper exploits
the felicitous circumstance that a Eurobarometer survey was conducted during a gapless time
period running from precisely six days prior to six days after the U.S. election (Figure 2). This
exceptionally fortunate setup makes it possible to test causally whether the EU became more—
or less—popular as a response to Trump’s victory. To do so, this research compares the group
of respondents who were surveyed prior to the election (control group) to those interviewed
after Trump’s victory (treatment group).
Such natural experiments are a novel methodological approach in the social sciences
that has become increasingly popular in recent years. Past work has, for instance, used terrorist
attacks10, football match victories11, celebrity suicides12, exposure to refugee arrivals13,
selective passage of right-to-carry concealed handgun laws14, political devolution15, and
electoral quotas16 as external shocks. By taking this innovative method to a new context, this
study makes two important contributions over and above the immediate relevance of knowledge
about the relation between the incumbent U.S. president and the EU. First, it adds to the fields
of regional integration research and EU studies by showing how an exceptional historical event
can affect public support for integration. While most past research on public support for
European integration has focused on monitoring long-term trends17 and exploring the
underlying social stratification18, a growing number of studies has looked at the effect of
particular events, from EU summits19, corruption scandals20, and the EU’s receipt of the Nobel
Peace Prize21 to the Euro crisis22, the refugee crisis23, Brexit24, and a media boycott25. This study
adds to this growing corpus by revealing the instant impact of a singular and particularly salient
event from one day to the next on these ostensibly inert public opinion structures. Furthermore,
by examining how post-election opinion dynamics varied between societal subgroups, we

4
reveal event-driven sociopolitical shifts in the EU’s base of support. Our analysis may thus also
aid in understanding current changes in the position of the European right toward the EU.
Second, by combining two hitherto disconnected theories on political dynamics that
predict opposing effects—the rally theory and the domino theory—and testing them on a new
empirical case, this study contributes, more generally, to knowledge regarding the complex,
unintended, and partially unpredictable and counterintuitive dynamics that political events in
one part of the world can have in another. As this article will show, a reinforced spread of
nationalism to Europe (domino effect) appeared just as plausible at the outset as its opposite, a
positive rally effect. Yet, empirically, the rally effect prevailed.
This paper proceeds as follows: first, it lays out the competing theories of how Trump’s
unexpected victory could have affected the EU’s popularity. Thereafter, the paper introduces
the research design in more detail. Next, the results are presented, focusing consecutively on
the overall impact Trump’s election has had on the EU’s popularity, a couple of subgroup
analyses, and a summary of robustness checks that were run (which are available in full as
Supplementary Material). The paper concludes with a summary and discussion of the findings.
The puzzle: three plausible, yet mutually exclusive potential outcomes
Three different effects of the election of Donald Trump on the EU’s popularity in Europe are
theoretically thinkable: (a) an increase, (b) a decrease, and (c) a non-effect. In the following,
arguments for and mechanisms behind each of these potential outcomes are discussed and
competing hypotheses are formulated. All of them are credible, making it difficult to formulate
assumptions about the adequacy of one of them—and the falsity of the others—ex ante. Instead,
this study subjects these competing hypotheses of rival theories to a fair test.26
Arguments for a positive rally effect
The first plausible effect of Trump’s surprise victory is a higher popularity of the EU in Europe.
The central mechanism behind such a positive impact could be a rally effect. The term, in its
full notation—“rally-’round-the-flag effect”—was originally used to describe rises in the U.S.
president’s popularity in the wake of international crises.27 By now, scholars have already
uncoupled the “rally effect” from this initially strict focus and replaced the popularity of the
U.S. president with trust in government, the ruling party, other leaders, or general public
opinion in the United States and other countries as the dependent variable.28 It is possible to
dissociate this idea further from its narrow original context to make the general argument that
a perceived external threat can bring members of a social entity to unite. A similar proposition
can be found in Karl Deutsch’s transactionalist theory of integration.29 Admittedly, Deutsch
and colleagues speak of unifying effects of an external military threat, and in present days
Europeans certainly do not fear a direct attack by the United States against Europe. However,
many Europeans are indeed worried about the unpredictability of Donald Trump as the
commander-in-chief of the world’s largest military force.30 Yet more importantly, no reason is
immediately apparent why this mechanism should not work similarly for sociopolitical threats
more generally. In the present case, the sudden election of Trump as an American nationalist

5
could be subjectively experienced to pose an external threat to the stability and prosperity of
Europe. As mentioned above, observers suspected that his election would affect transatlantic
relations and prospects of European integration. Such concerns were fueled prior to the election
by Trump’s nationalist credo “America first,” condemnatory statements about NATO and
TTIP, and consentient comments regarding the looming partial breakup of the EU in the wake
of Brexit.31 One fear was thus the need to reinvent Europe’s role in the world upon realizing
that the United States is no longer a reliable partner. As stated by Angela Merkel in May 2017
after meeting with Trump, Europe “really must take our fate into our own hands.”32 Thus, it is
possible that already in immediate response to Trump’s unexpected victory, Europeans rallied
around the European “flag”—creating a feeling of unity that could be measurable through an
increased popularity of the EU.
Notably, Hannah Arendt made a similar argument in the first half of the twentieth
century, stating, “If it is true that each nationalism … begins with a real or fabricated common
enemy, then the current image of America in Europe may well become the beginning of a new
pan-European nationalism.”33 She evaluated this “anti-American Europeanism”34 negatively,
as nationalistic with ties to fascism and in opposition to a liberal federalism. Hence, a unifying
effect through an external threat may work for liberal cosmopolitans just as for conservative
nationalists. The former may move toward increased EU support, seeing it as a stronghold of
an open, post-nationalist world, while the latter may embrace Europe as a bastion in a world of
strong and nationalistic regional powers, such as the United States, China, and Russia.
More recent research has also looked deeper into the question of who is most susceptible
to rally effects. Baum argues that while different social groups may have varying propensities
to engage in rallying, individuals who are closest to the point of ambivalence between approval
and disapproval on the issue in question are most likely to change their opinion.35 Colaresi
suggests that rally effects do not require an emotional or irrational public but can be modeled
as a rational response to international crises.36 Baker and Oneal additionally find that the size
of a rally effect is influenced by how the media covers the event in question, potentially making
differences in media consumption a significant factor.37
Recently, it has been argued that Brexit—a victory for right-wing populism in many
ways comparable to Trump’s election—may be responsible for a rise in Eurobarometer
respondents who say they “feel like citizens of the EU.”38 The adequacy of this claim has not
yet been corroborated through a rigorous causal analysis. But if the assessment is correct, it
could suggest that a similar positive effect, in which Europeans rally around “their flag” in
defiance of nationalist sentiment, may be at work in the case of Trump. We can thus formulate
as a first hypothesis that
H1: Trump’s election increased the EU’s popularity in Europe (rally effect).
At the same time, however, arguments for the opposite case, a negative domino effect,
have been made.
Arguments for a negative domino effect
Another plausible outcome is a decrease in the EU’s popularity in the wake of Trump’s surprise
victory. The underlying mechanism could be a domino effect in which the United States’ “fall”

6
for right-wing populism constitutes the start of a chain reaction in which other countries—e.g.,
in Europe—successively “tilt over,” resulting in rising levels of nationalism and anti-
supranationalism. Although domino theory originated in the cold war context of countries
supposedly acceding consecutively to communism, it has been applied to other contexts, such
as democratization39, regionalism40, and, most recently, populism after Brexit.41 Thus, it does
not appear far-fetched to apply domino theory to the spread of right-wing populism (and an
according decline of supranationalism) after Trump’s success. In the present case, the domino
effect could mean lower popularity levels for the EU due to a combination of resignation among
cosmopolitans and reinforcement of nationalists in their beliefs.
In Europe, fears of such a domino effect were visible in public discourses prior to the
national elections in Austria (December 2016), the Netherlands (March 2017), France (April-
May 2017), and Germany (September 2017). Many observers anticipated significant political
shifts to the right, often naming Trump’s victory as one ground for their expectance. Mudde,
for instance, argued, writing immediately after the U.S. election, that “the surprise win of
Donald Trump is a gift from heaven for the far right around the globe.” Their victory in the
United States, he suspected, “gives them a narrative of hope and success.”42 New impetus to
right-wing populism should go hand in hand with increased Euroscepticism43 and thus a less
popular EU. Research shows that this link has become even stronger during recent years.44
Thus, a competing hypothesis to H1 would be that
H2: Trump’s election decreased the EU’s popularity in Europe (domino effect).
To some extent, however, converse trends became observable in the months after the
U.S. presidential election, making concerns over a domino effect appear somewhat exaggerated
in hindsight.45 Yet, it is still unclear which form the Trump effect on the EU’s popularity took
immediately after the election. Was it a positive rally effect, a negative domino effect—or rather
no effect at all?
Arguments for a non-effect
A non-effect could simply be the outcome of no effect, if
H0: Trump’s election did not affect the EU’s popularity in Europe.
A more complex “non-effect” could result from positive effects and negative effects (partially)
offsetting each other. To arrive at an understanding of how such a canceling out could come
about, it is necessary to compare different social subgroups. This article looks at two46 specific
potentially relevant societal divides: first, subgroups defined by the (perceived) economic
situation of the respondents’ country. Here, the way Europeans rate the economic situation of
their country serves as a proxy for the degree to which they feel their country has been affected
by the economic crises in the EU. The underlying idea is that in economically struggling
countries, the EU will likely have low initial popularity levels but potentially high upward
momentum, whereas in economically thriving countries that are hardly affected by the crises,
the EU likely has comparatively high initial popularity but low potential for upward momentum
(ceiling effect). We thus expect that

7
H3a: Trump’s election has had different effects on the EU’s popularity among
respondents who perceive their country as economically struggling as opposed to those
who perceive it as economically well-off.
Second, this paper looks at Europeans’ political orientation. As discussed in the
preceding sections, people with different political orientations could react differently to
Trump’s surprise victory. It is plausible that the EU has high initial popularity levels among the
political center and lower ones among the left and the right.47 This could again be connected to
little upward momentum among the center (which likely tends to be very much in favor of the
EU already prior to the election) whereas the right and the left may be more susceptible to
changing their views on the EU in response to the unexpected coming to power of a right-wing
nationalist in the U.S. (cf. the argument by Baum presented above). Dissecting the Trump effect
on the EU’s popularity by political orientation will allow to shed light on these potentially
diverging effects and to test whether they (partially) cancel each other out. We thus assume that
H3b: Trump’s election has had different effects in the political subgroups of the
European population.
Research design
Our research design exploits the fact that the 2016 U.S. presidential election took place amid
the fieldwork of a Eurobarometer survey. An explanation follows of how this fortunate
coincidence makes it possible to solve the fundamental problem of causal inference and thus to
estimate causally the effect of Trump’s victory on the EU’s popularity in Europe.
Solving the fundamental problem of causal inference
In order to estimate the causal effect of a treatment T, a study setup is required that allows the
same (or very similar) observations to be exposed to treatment and control simultaneously.48
However, observing treatment and control outcomes for the same observation at the same time,
that is, the counterfactuals of each state, is impossible. To be able to distinguish correlation
from causation despite this “fundamental problem of causal inference,”49 a few conditions must
be fulfilled. First, the assumption of independence must be satisfied, ensuring that differences
in outcome between treatment and control group are only due to the treatment. One possibility
to meet this assumption is randomization. If the treatment is assigned randomly to participants,
no observable or unobservable factors can bias potential outcomes. While randomization is
easily achieved in laboratory experiments given that assignment to treatment can be
manipulated by the researcher, studies based on observational data rarely meet this strict
assumption. Sometimes, however, randomization-like conditions occur naturally, allowing
researchers to draw causal inferences based on observational data. In such natural experiments,
the treatment is randomized not through manipulation by the researcher, but by an event that is
exogenous to the outcome in question. Thus, in natural experiments, individuals are exposed to
the treatment as-if random.50

8
Table 1. Descriptive statistics.
Control group (N=6,395)
Treatment group (N=9,890)
Δ means
Mean
SD
Min
Max
Mean
SD
Min
Max
EU popularity index
-0.074
1.06
-3.02
2.62
0.045
1.05
-3.02
2.62
-0.120***
Education (in years)
17.869
6.18
0
26
17.841
6.10
0
26
0.028
Female
0.506
0.50
0
1
0.515
0.50
0
1
-0.008
Occupation
Self-employed
0.071
0.26
0
1
0.080
0.27
0
1
-0.009*
White collar
0.243
0.43
0
1
0.259
0.44
0
1
-0.016*
Manual worker
0.186
0.39
0
1
0.203
0.40
0
1
-0.175**
Homemaker
0.044
0.20
0
1
0.044
0.21
0
1
-0.000
Unemployed
0.066
0.25
0
1
0.067
0.25
0
1
-0.001
Retired
0.323
0.47
0
1
0.280
0.45
0
1
0.043***
Student
0.066
0.25
0
1
0.066
0.25
0
1
-0.000
Age
50.838
18.0
15
96
49.246
17.5
15
98
1.591***
Urban (=1, 0=rural)
0.689
0.46
0
1
0.696
0.46
0
1
-0.007
Note: Based on Eurobarometer 86.2, own calculations, not weighted; SD=Standard deviation, education was top-coded at 26 years to reduce the influence of outliers. * p < 0.05,
** p < 0.01, *** p < 0.001.

9
Our research design meets the as-if random criterion of a natural experiment and thus
satisfies the independence assumption: The U.S. presidential election took place between the
sixth and the seventh day of the Eurobarometer’s 12-day fieldwork (see Figure 2). Being
assigned to an interview date before or after the presidential election was random, since it did
not depend on the respondents’ political preferences, socio-economic characteristics, or other
observable or unobservable confounders. Thereby, the outcome (change in the popularity of the
EU among respondents) should be related to nothing but the assignment to treatment, namely,
being interviewed before or after the election.
Table 1 shows that this as-if random criterion actually holds empirically. It provides a
summary of descriptive statistics on a range of widely used socio-economic and demographic
variables in our sample, including age, gender, education in years (top-coded at 26), occupation,
and a dichotomous variable indicating whether the individual lives in the countryside or in a
town. As can be seen from the Δ means column, the sample is mostly well-balanced, with no
significant differences in education, gender, or place of living between the treatment and the
control group. For age and some occupational classes, minor differences can be found. For
instance, the mean age in the treatment group is 49.2 years, compared to 50.8 in the control
group. To correct for these small divergences, one of our models controls for the variables
depicted in Table 1 (see equation 2 below). Furthermore, robustness checks were run in which
older respondents were excluded (leading to an entirely balanced sample), and in which the
potential influence of between-county differences in the distribution of fieldwork across time
(see Figure 2) were tested. All robustness checks confirm the validity of our conclusions.51
Additionally, it is necessary to have plausible evidence that people “complied” with the
treatment, that is, that they were aware that Donald Trump was elected president. The Google
Trends data depicted in Figure 1B reveal that in each and every European country contained in
the Eurobarometer, the relative amount of online searches targeting “Donald Trump” spiked to
its maximum on the day following the election (dashed lines). The universality and extreme
nature of this pattern strongly suggest that most people became aware of Trump’s election as
president very quickly. While, of course, not everybody uses the Internet, there are good reasons
to assume that the general public soon took notice of the surprise victory of the Republican
candidate, making it a legitimate treatment in a natural experiment.
Furthermore, the exclusion restriction needs to be fulfilled.52 That is, it is necessary to
be certain that no other events over and above the election influenced the outcome. For this,
Google Trends data can again help. Exploring trending topics on Google during the examined
time frame reveals that “election” (relative rank in Google searches: 4), “Trump” (6), “election
results” (16), “Donald Trump” (17), “election 2016” (18), “Clinton” (23), and “polls” (24) were
the dominating date-specific keywords. All other high-ranking keywords relate to unspecific
everyday interests, including “Facebook” (1), “YouTube” (2), “Google” (3), “you” (5), “news”
(7), “Gmail” (8), “fb” (9), and “Hotmail” (10). This suggests that during the time-span under
study, there were no other relevant events that could distort the results. While not salient in the
Google Trends data, it should be noted that several European countries commemorate the end
of World War I, the Holocaust, and the end of communism in Eastern Europe during the time
frame under study. To exclude the possibility that these holidays are responsible for a potential
change in the EU’s popularity (due to the EU becoming salient as a peace project in people’s
minds during these days), we ran a robustness check in which we excluded countries that have

10
such public commemorations from the analysis. Results reconfirm the main findings.
Furthermore, we replicated the analysis with an earlier Eurobarometer from 2013 and did not
find a similar effect, confirming that the observed effect is due to the singular event of Trump’s
election.53
Finally, it is necessary to show that the premise of considering the presidential election
as an exogenous event is actually valid. The election would not have been exogenous if
Europeans could have anticipated Trump’s victory and started adapting attitudes on the EU
already before the election took place. However, Figure 1A illustrates that basically no one—
not even political insiders—expected Trump to win, and polls remained in Clinton’s favor until
the election night. European media outlets also speak of “one of the most improbable political
victories in modern US history”54 and argue that no one saw the victory of Donald Trump
coming.55 Thus, it is very unlikely that ordinary citizens had adapted their attitudes before the
election took place.56 Another argument in favor of exogeneity is that Europeans were, of
course, not the ones who went to the ballot boxes in the U.S. election. Thus, arguing that they
could anticipate Trump’s win because they were planning on voting Trump into office is
implausible. In the wake of the U.S. election, Europeans merely knew that Trump’s becoming
the next president was—according to polls and media reports—highly unlikely. All this
provides a high degree of certainty that Trump’s election took Europeans by surprise and thus
acted as an exogenous shock.
Data and sample
The Eurobarometer is a large-scale cross-sectional public opinion survey, conducted on behalf
of the European Commission. The survey program started in 1974 and is released biannually.
Here, we draw on the Eurobarometer 86.2, which contains data from 35 European countries,
collected between November 3 and 14, 2016. In the sample for this study, all non-EU member
states were excluded, since several central items were not included in these countries’
questionnaires. To increase consistency, Bulgaria was also excluded, because the data
collection started and ended two days earlier there than everywhere else (cf. Figure 2). All other
27 EU member states are contained in the analysis. Following best practice from a range of
previous research,57 respondents who were interviewed immediately after the external shock
(on November 9) were excluded to ensure that respondents in the treatment group actually had
time to become aware of the fact that Donald Trump was elected president. The number of
observations in the sample was additionally reduced by listwise deletion of missing cases in
creating the dependent variable (see below). The final sample contains 16,285 observations.
Dependent variable (DV)
The study used 12 items from the Eurobarometer to construct an aggregate measure of EU
popularity (see Table 2).58 All 12 items are based on a 4-option Likert scale and measure certain
positive attitudes toward the EU (e.g., agreement with statements such as “the EU is modern,”
“the EU is efficient,” or “the EU creates jobs”). Employing exploratory factor analysis, only
one factor with an Eigenvalue >1 (5.17) emerges, on which all items load with at least .55.
Thus, following commonly accepted standards59, all 12 items have good to excellent factor
loadings. The one-factor solution was confirmed by various adjustments and robustness
measures based on different sets of variables, as well as orthogonal and Varimax rotation

11
techniques. Furthermore, given that the variables in question were not discrete but categorical,
the results were reconfirmed by using polychoric factor analysis, a procedure that makes it
possible to perform factor analysis with categorical variables. Since the results resembled those
from the conventional factor analysis, it was decided to adhere to the standard procedure.
Employing the respective factor, the DV was extracted by using the Bartlett method.
Table 2. Elements of the EU popularity index.
Variable
Factor loadings
Uniqueness
1. EU is modern
0.6575
0.5677
2. EU is democratic
0.7171
0.4858
3. EU is protective
0.7019
0.5074
4. EU is efficient
0.6914
0.5220
5. EU is forward looking
0.7304
0.4665
6. Optimistic for the EU’s future
0.6807
0.5366
7. EU creates jobs
0.6415
0.5885
8. EU makes business easier
0.5514
0.6960
9. Satisfaction with democracy in EU
0.6918
0.5214
10. Feeling attached to the EU
0.5902
0.6516
11. Feel like being a EU citizen
0.5835
0.6595
12. EU respects own country’s interests
0.6125
0.6248
Eigenvalue
5.1721
0.9676
0.8929
% Variance
Cronbach’s alpha
Note: Based on Eurobarometer 86.2
Treatment variable and subgroup splitting
To measure the Trump effect, a treatment variable was created that captures, in line with
equation 2, whether the respondent was interviewed prior (T = 0, control group, November 5-
8, N = 6,395) or after (T = 1, treatment group, November 10-14, N = 9,890) the U.S. presidential
election. For the subgroup analyses, the sample was first divided into four subgroups based on
how respondents perceived the economic situation of their country (“very good,” “rather good,”
“rather bad,” or “very bad”). Individual-level political orientation, the second variable used to
split the sample into subgroups, was inquired in Eurobarometer 86.2 through the question “In
political matters people talk of ‘the left’ and ‘the right.’ How would you place your views on
this scale?” The scale, which was shown to respondents, had ten categories ranging from 1
(“left”) to 10 (“right”), with the middle categories remaining unspecified. In line with common
practice60, respondents were grouped into the three meta-categories “left” (1-4), “center” (5-6),
and “right” (7-10) for the subgroup analyses. For several additional subgroup analyses and
corresponding variable descriptions, see Supplementary Material, sections 5, 7, and 8.
Control variables
Operating within the framework of a natural experiment, it is not necessary to adjust for
covariates, as the treatment is randomly assigned and individual socio-economic and
unobserved characteristics should thus not vary before and after the treatment. However,

12
covariates can be included to estimate the treatment effect more precisely and to control for
smaller imbalances between treatment and control group.61 Those variables, however, should
be strictly exogenous. As discussed above, the sample is mostly well balanced, but some minor
differences between treatment and control group do exist (cf. Table 1). To correct for these
small imbalances and to increase precision, the variables listed in Table 1 were included as
controls in one of the models (see equation 2 below).
Method
We use a regression discontinuity (RD) design. We start with a parsimonious model that
employs OLS regression with the EU popularity index as DV and the treatment dummy (=
Trump effect) as independent variable. Robust standard errors were estimated as recommended
in the literature.62 In formal terms, the base model can thus be described as:
       ,
(1)
where  stands for the effect of the treatment T, namely, being interviewed on the EU’s
popularity after Trump’s election;  represents the intercept; and  is an error term. Next, the
above-mentioned control variables X were included to estimate the treatment effect  more
precisely:
        
(2)
In a last step, a more conventional form of the RD design was implemented by testing for
varying slopes in the EU’s popularity before and after the U.S. presidential election. This was
accomplished by including the interaction between time to election and the treatment variable:
             ,
(3)
where    represents the effect of a mean-centered time variable based on the interview
date, which is set to zero at November 10. The term    stands for the interaction effect
between the treatment variable and the mean-centered interview date variable. Using this
interaction effect, the slope of the centered interview variable was allowed to vary before and
after the treatment. Additionally, in equation 3,  no longer depicts the treatment effect for the
period spanning November 10–14 as in equations 1 and 2. Rather,  now represents the
immediate causal effect of Trump’s election on the EU’s popularity on November 10, that is,
the day that now is specified to be the first day after the election (  for in equation 3).
Thus, it is now possible to test for an instant, “overnight” Trump effect.
It might be argued that, in contrast to equation 3, equations 1 and 2 are not conventional
implementations of the RD technique, as they solely employ a dummy and no additional linear
treatment estimator or its corresponding polynomials.63 However, given the Eurobarometer’s
short time frame—which may make pinpointing the discontinuity to exactly November 10
unnecessary—and a general interest in parsimonious models, we would still like to test the first
two specifications. Following the stepwise procedure of looking at equations 1 to 3

13
consecutively is therefore considered the most sensible approach. A range of previous research
has used similar strategies in estimating RD regressions.64
In the graphical representations of the findings (Figures 3-5), all independent variables
except the treatment effect are mean-centered. This mean-centering facilitates interpretation
since the intercept  now represents a meaningful estimate, that is, the EU’s popularity for the
average European citizen prior to Trump’s win, while    illustrates the average popularity
after the election.
Results
Overall effect
Table 3 shows three models that predict the overall Trump effect on the EU’s popularity across
all 27 EU member states under study. In line with equation 1, model 1 contains only the
treatment effect, which is positive and highly significant (RD = .120, p < .001). Thus, Trump’s
victory appears to have increased rather than decreased the EU’s popularity, lending support to
the rally- (H1) rather than the domino- (H2) or the no-effect hypothesis (H0).
Model 2 adds country dummies and a range of control variables as described in equation
2 to account for the fact that not all covariates are perfectly balanced between the treatment and
the control group. The effects of the control variables all go in expectable directions, in line
with the existing literature on EU support:65 the EU is more popular among the urban
population, the better educated, the younger, and those in better-paying (white-collar) positions.
The Trump effect becomes somewhat smaller in size but remains significant (RD = .045, p <
.01). Despite this decrease, it is still considerable: the effect of being interviewed after Trump’s
victory is roughly equivalent to the effect that three additional years of education have on a
person’s opinion of the EU (            ). Since education has
repeatedly been shown to be one of the key predictors of pro-EU attitudes,66 the Trump effect
seems anything but negligible. To draw one more comparison, the Trump effect is also about
two thirds the size of the rural-urban divide, another major cleavage in contemporary societies.
All this suggests that Trump’s election has had a substantial effect on the EU’s popularity in
Europe. The explained variance increases from .3 to 13.6 percent from model 1 to model 2.
Model 3 includes an additional check on whether allowing for slope variance (i.e., up-
or downward-pointing trends during the days prior and after the election) leads to a more
accurate description than the previous model. As discussed above, RD now represents the
immediate, overnight Trump effect (cf. equation 3). Model 3 illustrates that there was a
significant and prompt European reaction to Trump’s win, namely, an increase of the EU’s
popularity by RD = .077. The non-significant effects of days (βdays = –.013, p > .05) and the
interaction between treatment and days (RD × βdays = .013, p > .05) show that the Trump effect
was an overnight effect and that no further meaningful change occurred in the following days.
In line with this, the non-increase in explained variance (still R2 = .136) suggests that allowing
for slope variance does not improve the quality of the model. Accordingly, we proceed with the
specifications of the more parsimonious and equally powerful model 2 in our further analysis.

14
Table 3. OLS regression. DV: EU popularity index.
(1)
(2)
(3)
Treatment group (1=yes)
0.120***
(0.02)
0.045**
(0.02)
0.077*
(0.03)
Treatment*Days
0.013
(0.01)
Days (mean-centered)
-0.013
(0.01)
Female (1=yes)
0.029
(0.02)
0.029
(0.02)
Age
-0.004***
(0.00)
-0.004***
(0.00)
Occupation
White collar
(ref)
(ref)
Self-employed
-0.055
(0.03)
-0.054
(0.03)
Manual
-0.114***
(0.02)
-0.114***
(0.02)
Homemaker
-0.116**
(0.04)
-0.115**
(0.04)
Unemployed
-0.357***
(0.04)
-0.356***
(0.04)
Retired
-0.053
(0.03)
-0.052
(0.03)
Student
0.442***
(0.06)
0.441***
(0.06)
Urban (=1, 0=rural)
0.073***
(0.02)
0.074***
(0.02)
Education (in years)
0.014***
(0.00)
0.014***
(0.00)
Country dummies


Constant
-0.074***
(0.01)
0.155*
(0.07)
0.125
(0.08)
Observations
16,285
16,285
16,285
R2
0.003
0.136
0.136
Note: Treatment group: Interviewed after November 9. Robust standard errors in parentheses. Own calculations based on Eurobarometer 86.2, not weighted. * p < 0.05,
** p < 0.01, *** p < 0.001.

15
Figure 3 shows the overall effect as specified in model 2 graphically. It illustrates that
the EU’s popularity in Europe increased significantly (RD = .045, p < .01) after the election of
Donald Trump. Specifically, average popularity rates rose from a negative popularity rate of –
.029 () prior to Trump’s success to a positive one of .016    in the election’s aftermath
(results from mean-centered model, not shown). We can thus conclude that a positive rally
effect rather than a negative domino effect or no effect occurred immediately after Trump’s
surprise victory. The following section looks at whether this positive effect occurred in all
specified subgroups, or whether social divides can be observed in line with H3a and H3b.
Figure 3. The overall “Trump effect” on the EU’s popularity.
Note: Based on regression presented in Table 3. Control variables, including the country dummies, are set to their
mean value. * p < 0.05, ** p < 0.01, *** p < 0.001.
Subgroup analyses
Dividing the sample by the respondents’ perception of their country’s economy reveals a first
meaningful fault line: While the EU became significantly more popular among respondents
who perceived their countries’ economy as “very bad” after the election, respondents who
thought of it as “rather bad,” “rather good,” or “very good” did not significantly change their
attitudes toward the EU (Figure 4, cf. Table A1 in the Appendix for the full model statistics).
Thus, it appears that it was specifically respondents who perceived their countries as ridden by
economic turmoil that “rallied ’round the EU’s flag” in response to the Trump shock. Additional
analyses based on a subgroup division by national unemployment rate and change in
unemployment rate reconfirm this picture.67

16
Figure 4. Treatment effect of the US presidential election. Subgroup analysis by perceived economic situation of the country.
Note: Based on regression presented in Table A1. Control variables, including the country dummies, are set at their mean value. * p < 0.05, ** p < 0.01, *** p < 0.001.

17
Figure 5. Treatment effect of the US presidential election. Subgroup analysis by partisanship.
Note: Based on regression presented in Table A2. Control variables, including the country dummies, are set at their mean value. * p < 0.05, ** p < 0.01, *** p < 0.001.

18
Subgroup analyses by political orientation disclose a second meaningful divergence
(Figure 5; Table A2). In line with existing research, the EU was initially most popular among
the political center (EU popularity index score: .031), followed by the left (.008), and least
popular among the right (–.075). While it appears that the EU became slightly more popular
among the political left (RD = .007) and the center (RD = .032) in the wake of the U.S.
presidential election, these increases are not significant (p > .05). Among the right, however, a
significant upward jump occurred after Trump’s victory (RD = .114, p < .01). The right
overtook the left and remains only slightly below the center, indicating a remarkable shift in
the political landscape of EU support.68 This picture can also be reconfirmed with two
alternative indicators.69
Robustness checks
To ensure that these results are not spurious, we ran a comprehensive set of thirteen robustness
checks, including, inter alia, a placebo analysis, propensity score and nearest neighbor
matching, as well as a correction for the minor sample imbalances observed in Table 1. Due to
spatial restraints, these robustness checks—which all support the findings presented above—
are included as Supplementary Material. Here, just some key findings are highlighted. First, the
main result of a positive and significant Trump effect on the EU’s popularity could be replicated
with another data set, namely round 8 of the European Social Survey. Strikingly, even the
relative size of the Trump effect is similar, being again roughly equivalent to the effect of three
additional years of education. We further reran the analysis using a 2012 Eurobarometer (78.1)
that coincided with Barack Obama’s second election as U.S. president to test whether the
“Trump effect” is not actually a general “U.S. presidential election effect.” No effect was found
in 2012, suggesting that the significant effect observed in 2016 was uniquely connected to the
personality of Donald Trump. Finally, we replicated the analysis using a 2013 Eurobarometer
(80.1) that was carried out during the same time of the year to check whether the observed effect
is not actually a seasonal one, possibly due to the commemoration days mentioned above.
Again, no significant treatment effect was found, allowing to rule out the possibility that
memorial days (or any other intervening seasonal events) are the actual reason for the observed
change in the EU’s popularity. Thus, there is conclusive evidence that the significant increase
in the EU’s popularity observed in November 2016 actually occurred and that it was indeed
caused by the surprise election of Donald Trump as U.S. president.
Summary and discussion
This study treated the surprise victory of Donald Trump in the 2016 U.S. presidential election
as an external shock and examined whether it led to a change in the popularity of the EU in
Europe. Three main findings shall be highlighted:
1. The EU’s popularity did in fact increase immediately after the election, suggesting that
Trump’s surprise victory caused a “rally effect” in Europe (in line with H1, and
disproving H2 and H0).

19
2. There is an economic fault line in that gains in the EU’s popularity after Trump’s victory
were particularly high among respondents who perceived their country as economically
struggling (in line with H3a).
3. There is also a political fault line in that gains in the EU’s popularity after Trump’s
surprise win were particularly high among the political right (in line with H3b),
suggesting a shift in the EU’s base of support.
This study has implications regarding both the specific case it addresses and the broader,
general dynamic that underlies it. Relating to the specific case, the findings show that the
election of Trump as a right-wing nationalist with a declared aversion to supranational
institutions—including the EU—did not trigger a domino effect in the same direction in Europe.
To the contrary, a rally effect occurred, in which Europe moved closer together, rallying around
the EU’s “flag.” This indicates that an event that may at first sight appear to be a global victory
for nationalism can immediately trigger measurable sentiments of resistance in another part of
the world, actually leading to new impetus for supranationalism.
The large gains in the EU’s popularity among the political right, however, are an
important qualifier. They suggest that this increased popularity of the EU is likely not primarily
cosmopolitan or liberal in nature. Instead, the Trump effect appears to have given rise to a right-
wing variant of pro-EU stances, akin to Hannah Arendt’s idea of a non-progressive “anti-
American Europeanism.” This impression fits well with recent shifts in the positions of some
leading right-wing populist politicians in several European countries, from mere Eurosceptic
nationalism to pro-European stances with a right-wing twist. For instance, Hungary’s Prime
Minister Victor Orbán has adapted Trump’s slogan, calling to “make Europe great again.”70
During Austria’s recent election campaign, right-wing populist party FPÖ leader Heinz-
Christian Strache praised the EU as a “positive project.”71 Most recently, Czech parliamentary
election winner Andrej Babis surprisingly stated that his party, ANO, was pro-European and
that he wants to take an active role in the EU to “fight against migration and other issues.”72
Although such a rhetorical shift cannot necessarily be observed in all leaders of the far right in
Europe, the post-Trump increase in the EU’s popularity among the parts of the population
identifying as right-wing observed in this study73 could be part of a larger process of attempted
usurpation of the EU through right-wing forces. Hence, instead of a rally-’round-the-flag effect,
it would perhaps be more appropriate to speak of a conquer-the-flag effect, in which the right
aims at reshaping Europe and the EU according to its ideas, that is, as a strong and closed
fortress and inward-looking power that is fit enough to compete with Trump’s America. While
perhaps too suggestive of a unified and coordinated activity by the European right that did not
exist to this extent at this point in time, this picture would fit the common perception of the
right as more susceptible to external threats and zero-sum logics that require fierce deterrence
and unitary responses. The stable, continuing high EU popularity scores among the center and
the left, however, show that it is not clear whether the right would be successful in any such
“takeover” attempt. Who will determine the EU’s future political orientation is thus an open
question and more research is needed to confirm this shift in the EU’s base of support and to
determine its longevity. Future research should also go deeper in exploring the exact meaning
European integration has for the right today.

20
The finding that the rise in the EU’s popularity was also particularly high among
respondents who perceived the economic situation of their home country as “very bad” could
be interpreted as a positive sign for the legitimacy base of the EU. Among these respondents,
initial EU popularity values were much lower than in any other group, but the Trump effect
brought their views on the EU at least slightly closer to the more favorable ones among
respondents who were less concerned about their countries’ economy. This shows that said part
of the European population is not “lost” for the EU but in fact susceptible to changing their
opinion on European integration in response to news, in this case the external shock of Trump’s
surprise victory. Following Baum, at least parts of this group appear to be closer to the point of
ambivalence between approval and disapproval rather than inveterate and fierce opponents of
the EU.74
More generally, then, this study shows the complexity and partial unpredictability of
political chains of interaction. A shift in one direction in one part of the world does not
necessarily lead to a simple domino effect in the same direction in another part. This insight is
important, because humans seem to have an inbuilt tendency to make unidirectional
extrapolations. As shown in the discussion of a potential domino effect, such expectations of
linearity were clearly visible in Europe after Trump’s victory.75 Countervailing effects are much
harder to foresee and reveal, and in this case a natural experiment helped uncover them.
Furthermore, this study contributes to the growing literature on the impact of exceptional
historical events on public support for integration in the field of EU studies. Many of the events
studied in past research, such as the Euro and refugee crises, lasted several years. This study,
by contrast, reveals the instant impact that a singular event can have from one day to the next
on these ostensibly inert public opinion structures. While some historical changes are slow and
long term, in other instances “history, in fact, is a very sudden thing.”76
This study is not without limitations. For one thing, the Eurobarometer is relatively short
on sociodemographic variables. Although controlling for more covariates is not necessary given
the experimental setup of the study, additional individual-level information, such as a well-
constructed social-class variable, would have been useful. A more important limitation is that
it was only possible to examine the short-term Trump effect. Whether this effect persists in the
long run and whether politicians will be able to transform it into political capital that may
ultimately lead to a deepening—or, more generally, a re-shaping—of European integration can
only be speculated about. Yet, even if additional survey material became available in the future,
a long-term effect would be hard to prove causally due to the many intervening events.77
A further concern regards the normative evaluation of this shift. In the first half of the
twentieth century, Hannah Arendt warned that the unifying effects that arise from a perceived
external threat are not necessarily desirable forces. She concluded, “Americanism on one side
and Europeanism on the other side of the Atlantic, two ideologies facing, fighting, and, above
all, resembling each other as all seemingly opposing ideologies do—this may be one of the
dangers we face.”78 Care must be taken, therefore, not to glorify the positive Trump effect on
the EU’s popularity as a victory for cosmopolitan forces in response to a parochialist threat.

21
Appendix
Table A1. Subgroup analysis by perceived economic situation of the country. DV: EU popularity index.
Very bad
Rather bad
Rather good
Very good
Treatment group (1=yes)
0.095*
(0.05)
0.007
(0.02)
0.006
(0.02)
0.076
(0.08)
Female (1=yes)
0.008
(0.05)
0.044
(0.02)
0.064**
(0.02)
-0.011
(0.08)
Age
-0.006**
(0.00)
-0.004***
(0.00)
-0.005***
(0.00)
-0.001
(0.00)
Occupation
White collar
(ref)
(ref)
(ref)
(ref)
Self-employed
0.000
(0.10)
0.005
(0.05)
-0.076
(0.05)
-0.025
(0.14)
Manual
-0.044
(0.08)
-0.079*
(0.03)
-0.042
(0.03)
-0.198
(0.13)
Homemaker
-0.115
(0.11)
-0.069
(0.06)
-0.045
(0.05)
0.105
(0.20)
Unemployed
-0.342***
(0.09)
-0.226***
(0.05)
-0.142*
(0.06)
-0.068
(0.23)
Retired
0.009
(0.09)
-0.035
(0.04)
-0.021
(0.04)
0.022
(0.13)
Student
0.516*
(0.21)
0.435***
(0.09)
0.140
(0.08)
0.284
(0.27)
Urban (=1, 0=rural)
-0.033
(0.05)
0.058*
(0.03)
0.111***
(0.02)
-0.042
(0.09)
Education (in years)
0.019**
(0.01)
0.014***
(0.00)
0.003
(0.00)
0.011
(0.01)
Country dummies




Constant
-0.068
(0.21)
0.135
(0.11)
0.667***
(0.11)
0.520
(0.31)
Observations
2,287
6,510
6,572
750
R2
0.19
0.14
0.14
0.16
Note: Treatment group: Interviewed after November 9. Robust standard errors in parentheses. Own calculations, based on Eurobarometer 86.2, not weighted. * p < 0.05, ** p <
0.01, *** p < 0.001.

22
Table A2. Subgroup analysis by partisanship. DV: EU popularity index.
Left
Center
Right
Treatment group (1=yes)
0.007
(0.03)
0.032
(0.03)
0.114**
(0.04)
Female (1=yes)
0.053
(0.03)
-0.013
(0.02)
0.077*
(0.04)
Age
-0.004**
(0.00)
-0.004***
(0.00)
-0.003
(0.00)
Occupation
White collar
(ref)
(ref)
(ref)
Self-employed
-0.117
(0.06)
-0.046
(0.05)
-0.076
(0.06)
Manual
-0.154***
(0.04)
-0.091*
(0.04)
-0.002
(0.05)
Homemaker
-0.215**
(0.08)
-0.063
(0.06)
-0.090
(0.10)
Unemployed
-0.388***
(0.07)
-0.164**
(0.06)
-0.355***
(0.08)
Retired
-0.105*
(0.05)
-0.031
(0.04)
0.013
(0.06)
Student
0.471***
(0.11)
0.442***
(0.10)
0.210
(0.14)
Urban (=1, 0=rural)
0.074*
(0.03)
0.045
(0.03)
0.096**
(0.04)
Education (in years)
0.019***
(0.00)
0.013***
(0.00)
0.003
(0.00)
Country dummies



Constant
0.114
(0.13)
0.245*
(0.11)
0.261
(0.16)
Observations
4,321
6,134
3,695
R2
0.18
0.13
0.16
Note: Treatment group: Interviewed after November 9. Robust standard errors in parentheses. Own calculations, based on Eurobarometer 86.2, not
weighted. * p < 0.05, ** p < 0.01, *** p < 0.001.

23
Acknowledgements
This research is part of the project “Cross-border Interactions and Transnational Identities,”
which is supported by the German Research Foundation (DFG) within the framework of the
DFG research unit FOR-1539 Horizontal Europeanization. For more information, see
http://www.horizontal-europeanization.eu/en. We are grateful to the participants of the research
colloquium at the Chair of Macrosociology at Otto von Guericke University Magdeburg for
useful comments and suggestions and to Cheryl Abundo for important advice. We would also
like to thank five anonymous reviewers at Perspectives on Politics for their invaluable feedback
that heavily improved the article.
Supplementary material
The Supplementary Material is available on the Perspectives on Politics website, together with
the Stata code used for the analyses presented here.

24
Notes
1 Fivethirtyeight 2016.
2 Roberts et al. 2016; Steltzner 2016.
3 Tolentino 2016.
4 Arkin 2015.
5 Reilly 2016.
6 Timm 2015.
7 European Parliament 2016; Krastev 2017, 6; Maher 2016; Quest 2016.
8 Boomgaarden and Vreese 2007, 354.
9 Silver 2017, emphasis added.
10 Jakobsson and Blom 2014; Legewie 2013.
11 Depetris-Chauvín and Durante 2017.
12 Hoffman and Bearman 2015.
13 Czymara and Schmidt-Catran 2017.
14 Donohue, Aneja and Weber 2017.
15 Ferwerda and Miller 2014.
16 Bhavnani 2009.
17 E.g., Inglehart, Rabier and Reif 1987.
18 E.g., Kuhn, van Elsas, Hakhverdian and van der Brug 2016.
19 Smetko et al. 2003; Elenbaas et al. 2012.
20 De Vries 2018.
21 Ibid.
22 Degner 2017; Hobolt and Wratil 2015; Kuhn and Stöckel 2014.
23 Harteveld et al. 2018.
24 Hobolt 2016.
25 Foos and Bischof 2018.
26 Cf. King, Keohane, and Verba 1994.
27 Mueller 1970.
28 Arian and Olzaeker 1999; Kazun 2016; Lai and Reiter 2005; Perrin and Smolek 2009.
29 Deutsch et al. 1957, 156-157.
30 McCarthy 2016.
31 European Parliament 2016; Krastev 2017, 6; Maher 2016; Quest 2016.
32 Farrell 2017.
33 Arendt 1994 [1930-1954], 416.
34 Ibid, 416.
35 Baum 2002.
36 Colaresi 2007.
37 Baker and Oneal 2001.
38 Stone 2017.
39 Leeson and Dean 2009.
40 Baldwin 1993.
41 Hobolt 2016; Stone 2017.

25
42 Mudde 2016; see also Kirkegaard 2017.
43 Wellings 2010, 2014; Vasilopoulou 2011; Halikiopoulou, Nanou and Vasilopoulou 2012.
44 Grabow and Hartleb 2013; Van Elsas and van der Brug 2015.
45 Bremmer 2017; Kirkegaard 2017; Silver 2017.
46 The Supplementary Material contains several additional subgroup analyses, based on polit-
ical sophistication (i.e., knowledge about the EU) and a split of the EU into East and West
as well as North and South. Furthermore, it presents robustness checks for the two main
subgroup analyses presented here, with national unemployment rate and change in unem-
ployment rate as two alternative (and more “objective,” country-level) indicators for the eco-
nomic situation of the country (cf. Braun and Tausendpfund 2014) and territorial identifica-
tion and globalization-related ideology (based on Teney et al. 2014) as alternative indicators
for political orientation.
47 Van Elsas and van der Brug 2015.
48 Rubin 1974.
49 Holland 1986.
50 Dunning 2012.
51 See Supplementary Material, sections 2 and 3.
52 Ibid, 121.
53 Supplementary Material, section 11.
54 Roberts et al. 2016.
55 Steltzner 2016.
56 If, however, this unlikely anticipation of Trump’s win did appear, the estimates in this re-
search are conservative rather than invalid. In the event that Europeans were anticipating
Trump’s win, and given that this anticipation affects the EU’s popularity levels, they would
probably already be higher or lower than average four days prior to the election, making the
gap to the post-election EU popularity values smaller than it would be otherwise.
57 Finseraas, Jakobsson and Kotsadam 2011; Finseraas and Listhaug 2013; Legewie 2013.
58 The reasons for creating this index rather than drawing on a single item are twofold. On the
one hand, our goal was to capture the multifaceted nature of an abstract concept like the
EU’s popularity, and research suggests using an “overinclusive” item pool in such cases
(Clark and Watson 1995). On the other hand, the standard item that is often used to measure
EU support (“Do you think membership of the EU is a good thing/ a bad thing/ neither good
nor bad”) was only available for EU candidate countries and Cyprus TCC in this particular
Eurobarometer. However, a robustness check, where we additionally use a single item that
measures trust in the EU, confirms our results (see Supplementary Material, section 9).
59 Comrey and Lee 1992.
60 E.g., Rodriguez 2014.
61 Angrist and Pischke 2008.
62 Ibid.
63 Ibid.
64 Finseraas et al. 2011; Finseraas and Listhaug 2013; Legewie 2013; see also Gelman and
Hill 2007, 213-214.
65 e.g., Hooghe and Marks 2005; Inglehart et al. 1987; Kuhn et al. 2016; Lubbers and
Scheepers 2010.
66 e.g., Kuhn 2012; Hakhverdian et al. 2013.

26
67 Supplementary Material, section 5.
68 This shift mirrors findings from other studies showing that the political support base for
regional integration may change over time, for instance in Latin America (Deutschmann and
Minkus 2018).
69 Supplementary Material, section 5.
70 Eder 2017.
71 Bartlau 2017.
72 Muller 2017.
73 In the Supplementary material, we reconfirm this effect using two additional indicators,
showing that the Trump effect on the EU’s popularity is largest among communitarians (as
opposed to cosmopolitans) and among those identifying exclusively with their nation (as
opposed to those having an at least partly European identity).
74 Baum 2002.
75 E.g., Mudde 2016; Kirkegaard 2017.
76 Philip Roth, cited in Krastev 2017, 35.
77 See the discussion of Silver 2017 in the Introduction.
78 Arendt 1994 [1930-1954], 417.

27
References
Angrist, Joshua D., and Jörn-Steffen Pischke. 2008. Mostly Harmless Econometrics: An
Empiricist’s Companion. Princeton, NJ: Princeton University Press.
Arendt, Hannah. 1994 [1930-1954]. Essays in Understanding, 1930-1954. New York:
Schocken Books.
Arian, Asher, and Sigalit Olzaeker. 1999. “Political and Economic Interactions with National
Security Opinion: The Gulf War Period in Israel.” Journal of Conflict Resolution
43(1): 58-77.
Arkin, Daniel. 2015. “Donald Trump Criticized After He Appears to Mock Reporter Serge
Kovaleski.” NBC News, November 26, 2015. Available at: nbcnews.to/2x1QhBX (last
accessed 14/9/2017).
Baker, William D., and John R. Oneal. 2001. “Patriotism or Opinion Leadership? The Nature
and Origins of the ‘Rally ’round the Flag’ Effect.” Journal of Conflict Resolution
45(5): 661-687.
Baldwin, Richard. 1993. “A Domino Theory of Regionalism.” Working Paper No. 4465.
Cambridge, MA: National Bureau of Economic Research.
Bartlau, Christian. 2017. “Merkel muss keine Angst vor Kurz haben [Merkel does not need to
fear Kurz].” ntv. Available at: bit.ly/2yI7MFC (last accessed 25/10/2017).
Baum, Matthew A. 2002. “The Constituent Foundations of the Rally‐Round‐the‐Flag
Phenomenon.” International Studies Quarterly 46(2): 263-298.
Bhavnani, Rikhil. 2009. “Do Electoral Quotas Work After They Are Withdrawn? Evidence
from a Natural Experiment in India.” American Political Science Review 103(1): 23-
35.
Boomgaarden, Hajo G., and Claes H. de Vreese. 2007. “Dramatic Real-World Events and
Public Opinion Dynamics: Media Coverage and Its Impact on Public Reactions to an
Assassination.” International Journal of Public Opinion Research 19(3): 354-366.
Braun, Daniela, and Markus Tausendpfund. 2014. “The Impact of the Euro Crisis on Citizens’
Support for the European Union.” Journal of European Integration 36(3): 231-245.
Bremmer, Ian. 2017. “2017 Might Not Be Europe’s ‘Year of the Populist’ After All.” Time.
Available at: ti.me/2wZSNGa (last accessed 11/9/2017).
Clark, Lee Anna, and David Watson. 1995. “Constructing Validity: Basic Issues in Objective
Scale Development.” Psychological Assessment 7(3): 309-319.
Colaresi, Michael. 2007. “The Benefit of the Doubt: Testing an Informational Theory of the
Rally Effect.” International Organization 61(1): 99-143.
Comrey, Andrew L., and Howard B. Lee. 1992. A First Course in Factor Analysis (2nd ed.).
Hillsdale, NJ: Erlbaum.
Czymara, Christian S., and Alexander W. Schmidt-Catran. 2017. “Refugees Unwelcome?
Changes in the Public Acceptance of Immigrants and Refugees in Germany in the
Course of Europe’s Immigration Crisis.” European Sociological Review 33(6): 735-
751.

28
Degner, Hanno. 2017. “Public Attention, Governmental Bargaining, and Supranational
Activism: Explaining European Integration in Response to Crises.” Journal of
Common Market Studies, online first.
Depetris-Chauvín, Emilio, and Ruben Durante. 2017. “One Team, One Nation: Football,
Ethnic Identity, and Conflict in Africa.” CEPR Discussion Paper DP12233.
Deutsch, Karl, Sidney Burrell, Robert Kann, Maurice Lee, Jr., Martin Lichterman, Raymond
Lindgren, Francis Loewenheim, and Richard Van Wagenen. 1957. Political
Community and the North Atlantic Area. Princeton, NJ: Princeton University Press.
Deutschmann, Emanuel, and Lara Minkus. 2018. “Swinging Leftwards: Public Opinion on
Economic and Political Integration in Latin America, 1997–2010.” Latin American
Research Review 53(1): 38-56.
De Vries, Catherine. 2018. Euroscepticism and the Future of European Integration. Oxford:
Oxford University Press.
Donohue, John, Abhay Aneja, and Kyle D. Weber. 2017. “Right-to-Carry Laws and Violent
Crime: A Comprehensive Assessment Using Panel Data and a State-Level Synthetic
Controls Analysis.” Stanford Law and Economics Olin Working Paper 508.
Dunning, Thad. 2012. Natural Experiments in the Social Sciences: A Design-Based Approach.
Cambridge, UK: Cambridge University Press.
Eder, Florian. 2017. “Viktor Orbán: Make Europe (But Not the EU) Great Again.” Politico.
Available at: politi.co/2lcVphS (last accessed 25/10/2017).
Ellenbaas, Matthijs, Claes H. De Vreese, Hajo G. Boomgaarden, and Andreas R.T. Schuck.
2012. “The Impact of Information Acquisition on EU Performance Judgments.”
European Journal of Political Research 51(6): 728-755.
European Parliament, ed. 2016. “US Elections: ‘We Are Not Very Sure What Exactly Trump’s
Foreign Policy Agenda Is’.” Available at: bit.ly/2yBnDI7 (last accessed 12/9/2017).
Farrell, Henry. 2017. “Thanks to Trump, Germany says it can’t rely on the United States. What
does that mean?” Available at: wapo.st/2sk52vA (last accessed 12/5/2018).
Ferwerda, Jeremy, and Nicholas Miller. 2014. “Political Devolution and Resistance to Foreign
Rule: A Natural Experiment.” American Political Science Review 108(3): 642-660.
Finseraas, Henning, Niklas Jakobsson, and Andreas Kotsadam. 2011. “Did the Murder of Theo
van Gogh Change Europeans’ Immigration Policy Preferences?” Kyklos 64(3): 396-
409.
Finseraas, Henning, and Ola Listhaug. 2013. “It Can Happen Here: The Impact of the Mumbai
Terror Attacks on Public Opinion in Western Europe.” Public Choice 156(1-2): 213-
228.
Fivethirtyeight, ed. 2016. “National Polls.” Available at: 53eig.ht/2xhNQec (last accessed
8/9/2017).
Foos, Florian, and Daniel Bishof. 2018. “Can the tabloid media create Eurosceptic attitudes?
A quasi-experiment on media influence in England.” Available at: bit.ly/2xn78QU
(last accessed 10/4/2018).
Gelman, Andrew, and Hill, Jennifer. 2007. Data Analysis Using Regression and
Multilevel/Hierarchical Models (Vol. 1). New York: Cambridge University Press.
Grabow, Karsten, and Florian Hartleb (2013). Europa – nein danke? Studie zum Aufstieg
rechts- und nationalpopulistischer Parteien in Europa [Europe – No Thanks? A Study

29
on the Rise of Right-wing and Nationalist Populist Parties in Europe]. Berlin: Konrad
Adenauer Foundation.
Hakhverdian, Armen, Erika Van Elsas, Wouter Van der Brug, and Theresa Kuhn. 2013.
“Euroscepticism and Education: A Longitudinal Study of 12 EU Member States,
1973–2010.” European Union Politics 14(4): 522-541.
Halikiopoulou, Daphne, Kyriaki Nanou, and Sofia Vasilopoulou. 2012. “The Paradox of
Nationalism: The Common Denominator of Radical Right and Radical Left
Euroscepticism.” European Journal of Political Research 51(4): 504-539.
Harteveld, Eelco, Joep Schaper, and Sarah L. De Lange. 2018. “Blaming Brussels? The Impact
of (News about) the Refugee Crisis on Attitudes towards the EU and National
Politics.” Journal of Common Market Studies 56(1): 157-177.
Hobolt, Sara B. 2016. “The Brexit Vote: A Divided Nation, A Divided Continent.” Journal of
European Public Policy 23(9): 1259-1277.
Hobolt, Sara B., and Christopher Wratil. 2015. “Public Opinion and the Crisis: The Dynamics
of Support for the Euro. Journal of European Public Policy 22(2): 238-256.
Hoffman, Mark A., and Peter S. Bearman. 2015. “Bringing Anomie Back In: Exceptional
Events and Excess Suicide.” Sociological Science 2: 186-210.
Holland, Paul W. 1986. “Statistics and Causal Inference.” Journal of the American Statistical
Association 81(396): 945-960.
Hooghe, Liesbet, and Gary Marks. 2005. “Calculation, Community and Cues: Public Opinion
on European Integration.” European Union Politics 6(4): 419-443.
Inglehart, Ronald, Jacques‐René Rabier, and Karlheinz Reif. 1987. “The Evolution of Public
Attitudes Toward European Integration: 1970–1986.” Journal of European
Integration 10(2-3): 135-155.
Jakobsson, Niklas, and Svein Blom. 2014. “Did the 2011 Terror Attacks in Norway Change
Citizens’ Attitudes Toward Immigrants?” International Journal of Public Opinion
Research 26(4): 475-486.
Kazun, Anastasia. 2016. “Framing Sanctions in the Russian Media: The Rally Effect and
Putin’s Enduring Popularity.” Demokratizatsiya 24(3): 327-350.
Kirkegaard, Jacob. 2017. “Fears of Right-Wing Populism May Be Overblown in Europe.”
Peterson Institute for International Economics. Available at: bit.ly/2y5dNLR (last
accessed 12/9/2017).
Krastev, Ivan. 2017. After Europe. Philadelphia: University of Pennsylvania Press.
Kuhn, Theresa. 2012. “Why Educational Exchange Programmes Miss Their Mark: Cross‐
border Mobility, Education and European Identity.” Journal of Common Market
Studies 50(6): 994-1010.
Kuhn, Theresa, Erika van Elsas, Armen Hakhverdian, and Wouter van der Brug. 2016. “An
Ever Wider Gap in an Ever Closer Union: Rising Inequalities and Euroscepticism in
12 West European Democracies, 1976–2008.” Socio-Economic Review 14(1): 27-35.
Kuhn, Theresa, and Florian Stoeckel. 2014. “When European Integration Becomes Costly: The
Euro Crisis and Public Support for European Economic Governance.” Journal of
European Public Policy 21(4): 624-641.

30
Lai, Brian, and Dan Reiter. 2005. “Rally ’round the Union Jack? Public Opinion and the Use
of Force in the United Kingdom, 1948–2001.” International Studies Quarterly 49(2):
255-272.
Leeson, Peter T., and Andrea M. Dean. 2009. “The Democratic Domino Theory: An Empirical
Investigation.” American Journal of Political Science 53(3): 533-551.
Legewie, Joshua. 2013. “Terrorist Events and Attitudes Toward Immigrants: A Natural
Experiment.” American Journal of Sociology 118(5): 1199-1245.
Lubbers, Marcel, and Peer Scheepers. 2010. “Divergent Trends of Euroscepticism in Countries
and Regions of the European Union.” European Journal of Political Research 49(6):
787-817.
Maher, Richard. 2016. “European Leaders Would See a Donald Trump Victory as Total
Calamity.” Huffington Post. Available at: bit.ly/2ytKLJ2 (last accessed 12/10/2017).
McCarthy, Niall. 2016. “Europeans Are Terrified by the Prospect of Trump Becoming
President.” Forbes. Available at: bit.ly/2zZTosy (last accessed 12/10/2017).
Mudde, Cas. 2016. “Europe’s Far Right Has Been Boosted by Trump’s Win—For Now.”
November 9, 2016. The Guardian. Available at: bit.ly/2fDOeYb (last accessed
12/9/2017).
Mueller, John E. 1970. “Presidential Popularity from Truman to Johnson.” American Political
Science Review 64(1): 18-34.
Muller, Robert. 2017. “Czech Vote Winner Babis Wants Active EU Role, Not Favoring
Government with Extremists.” Reuters. Available at: reut.rs/2yO1YfS (last accessed
25/10/2017).
Perrin, Andrew J., and Sondra J. Smolek. 2009. “Who Trusts? Race, Gender, and the
September 11 Rally Effect Among Young Adults.” Social Science Research 38(1):
134-145.
Quest, Richard. 2016. “Trump’s Challenge: Will He Help Hold the European Union
Together?” CNN. Available at: cnn.it/2fSfyHh (last accessed 12/9/2017).
Reilly, Katie. 2016. “Here Are All the Times Donald Trump Insulted Mexico.” Time, August
31, 2016. Available at: ti.me/2rfFX7m (last accessed 13/9/2017).
Roberts, Dan, Sabrina Siddiqui, Ben Jacobs, Lauren Gambino, and Amanda Holpuch. 2016.
“Donald Trump Wins Presidential Election, Plunging US into Uncertain Future.” The
Guardian. Available at: bit.ly/2ek6lq8 (last accessed 13/9/2017).
Rodriguez, Robert G. 2014. “Reassessing the Rise of the Latin American Left.” Midsouth
Political Science Review 15(1): 53–80.
Rubin, Donald B. 1974. “Estimating Causal Effects of Treatments in Randomized and
Nonrandomized Studies.” Journal of Educational Psychology 66(5): 688-701.
Silver, Nate. 2017. “Donald Trump Is Making Europe Liberal Again.” June 14, 2017. Available
at: 53eig.ht/2tkqf7g (last accessed 12/9/2017).
Smetko, Holli A., Wouter van der Brug, and Patti M. Valkenburg. 2003. “The Influence of
Political Events on Attitudes Towards the European Union.” British Journal of
Political Science 33(4): 621-634.
Steltzner, Holger. 2016. “Herausforderung Trump. [Challenge Trump]” Frankfurter
Allgemeine. Available at: bit.ly/2zt7iTu (last accessed 12/9/2017).

31
Stone, Jon. 2017. “More Europeans Than Ever Say They Feel Like Citizens of the EU.” August
2, 2017. Independent. Available at: ind.pn/2vwasrg (last accessed 12/9/2017).
Teney, Céline, Onawa Promise Lacewell, and Pieter De Wilde. 2014. “Winners and losers of
globalization in Europe: attitudes and ideologies.” European Political Science Review
6(4): 575-595.
Timm, Jane C. 2015. “Trump Calls for Discrimination Against Muslims.” MSNBC, December
7, 2015. Available at: on.msnbc.com/1NCbBit (last accessed 11/9/2017).
Tolentino, Jia. 2016. “Donald Trump’s Unconscious, Unending Sexism.” The New Yorker,
October 10, 2016. Available at: bit.ly/2vYfxVV (last accessed 14/9/2017).
Van Elsas, Erika and Wouter Van der Brug. 2015. “The Changing Relationship between Left-
right Ideology and Euroscepticism, 1973–2010.” European Union Politics 16(2):
194–215.
Vasilopoulou, Sofia. 2011. “European Integration and the Radical Right: Three Patterns of
Opposition.” Government and Opposition 46(2): 223-244.
Wellings, Ben. 2010. “Losing the Peace: Euroscepticism and the Foundations of Contemporary
English Nationalism.” Nations and Nationalism 16(3): 488-505.
Wellings, Ben. 2014. “English nationalism and Euroscepticism overlap and support each other
in important ways.” LSE British Politics and Policy Blog. Available at:
https://bit.ly/2xnG1Fu (accessed 20/4/2018).