The ‘Alternative für Deutschland in the Electorate’: Between Single-Issue and Right-Wing Populist Party

Abstract

The good result of the recently formed Alternative für Deutschland (AfD) was a striking outcome of the 2013 Federal Election. This article explores why AfD supporters chose this party at the 2013 Federal Election and at the 2014 European and eastern German State (Land) Elections. At the Federal Election the AfD's electorate was composed of two groups: a minority of instrumental issue-voters that were drawn to the AfD by its emphasis and positioning on the Euro crisis, and a majority of ‘late supporters’ that decided close to Election Day and were moved more by expressive motives, most notably xenophobic sentiments like those identified in other European countries as a main source of support for right-wing populist parties. The analysis of the subsequent elections shows that, paralleling developments in the AfD's public rhetoric, the Euro crisis ceased to be important for AfD support whereas xenophobic motives became more central.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=fgrp20
Download by: [UB Mannheim] Date: 13 January 2017, At: 09:54
German Politics
ISSN: 0964-4008 (Print) 1743-8993 (Online) Journal homepage: http://www.tandfonline.com/loi/fgrp20
The ‘Alternative für Deutschland in the Electorate’:
Between Single-Issue and Right-Wing Populist
Party
Rüdiger Schmitt-Beck
To cite this article: Rüdiger Schmitt-Beck (2017) The ‘Alternative für Deutschland in the
Electorate’: Between Single-Issue and Right-Wing Populist Party, German Politics, 26:1, 124-148,
DOI: 10.1080/09644008.2016.1184650
To link to this article: http://dx.doi.org/10.1080/09644008.2016.1184650
Published online: 15 Jun 2016.
Submit your article to this journal
Article views: 534
View related articles
View Crossmark data

The ‘Alternative fu
¨r Deutschland in the
Electorate’: Between Single-Issue and Right-Wing
Populist Party
RU
¨DIGER SCHMITT-BECK
The good result of the recently formed Alternative fu
¨r Deutschland (AfD) was a
striking outcome of the 2013 Federal Election. This article explores why AfD
supporters chose this party at the 2013 Federal Election and at the 2014 Euro-
pean and eastern German State (Land) Elections. At the Federal Election the
AfD’s electorate was composed of two groups: a minority of instrumental
issue-voters that were drawn to the AfD by its emphasis and positioning on
the Euro crisis, and a majority of ‘late supporters’ that decided close to Election
Day and were moved more by expressive motives, most notably xenophobic sen-
timents like those identified in other European countries as a main source of
support for right-wing populist parties. The analysis of the subsequent elections
shows that, paralleling developments in the AfD’s public rhetoric, the Euro crisis
ceased to be important for AfD support whereas xenophobic motives became
more central.
INTRODUCTION
At the German Federal Election on 22 September 2013, two parties failed to pass the 5
per cent threshold of the electoral system by a narrow margin. One was the Free Demo-
cratic Party (FDP) which suffered a landslide loss and for the first time ever failed to
win any mandates in the Bundestag. The other was the Alternative fu
¨r Deutschland
(Alternative for Germany, AfD), a newcomer in the German party system that had
been founded just half a year before the election and immediately gained 4.7 per
cent of the second votes. For a new party, this was an excellent result. In fact, never
before had a newcomer or fringe party come so close to passing the 5 per cent threshold
at a Federal Election. At subsequent elections the AfD drew momentum from this near
success. At subsequent nation-wide and State Elections it gained enough votes to send
representatives to the respective parliaments. With 7.1 per cent of the national vote
share, it won seven of the 96 German seats in the European Parliament, which was
elected on 25 May 2014. At the State Elections in Saxony, Brandenburg and Thuringia
in August and September 2014 it obtained vote shares between 9.7 and 12.2 per cent
and accordingly won substantial numbers of mandates in each of the three state parlia-
ments. In February and May 2015 the AfD also gained mandates in the parliaments of
German Politics, Vol.26, No.1, 2017, pp.124–148
http://dx.doi.org/10.1080/09644008.2016.1184650 #2016 Association for the Study of German Politics

the West German city states Hamburg and Bremen (vote shares 6.1 and 5.5 per cent).
These successes were followed up with more eye-catching performances in Saxony-
Anhalt, Baden-Wu
¨rttemberg and Rheinland Palatinate in early 2016.
The AfD emerged in early 2013 as a single-issue party criticising the policies of the
federal government and the Bundestag parties on the Euro crisis.
1
However, its organ-
isational face became somewhat ambiguous during the Federal Election campaign, as
it added – although only as an aside – a variety of other issues to its profile. Due to the
nature of the positions articulated by the AfD, political and scholarly observers soon
began to ask whether it should be rated as a right-wing populist party comparable to
the FPO
¨in Austria, the Front National in France or the VVD in the Netherlands. To
date, such enquiries have studied the AfD only from a supply-side perspective. Inves-
tigating the AfD’s electoral platforms and other forms of campaign communication
they focused on the party as a ‘propagandiser’.
2
In spite of the party’s repeated
claims of pursuing policies outside the left – right spectrum systematic analyses
found the AfD to be clearly leaning to the right. But overall its communication did
not warrant classification as a right-wing populist party.
3
The present article complements this research with a demand-side perspective. It
studies the ‘AfD in the electorate’
4
by examining why AfD supporters preferred this
party at the 2013 Federal Election and at the 2014 European and regional elections.
Although it takes guidance from various theories of electoral behaviour, the approach
of this research is exploratory. For investigating a novel and ambiguous phenomenon
like the AfD’s surprising success at the 2013 and 2014 elections a wide-angle lens that
covers a broad range of possible predictors of party choice appears most suitable. The
article starts with a brief account of how the AfD emerged on the political stage and
campaigned during the run-up to the Federal Election. It then goes on to trace how
the party’s electoral support crystallised during the campaign and what moved
voters to support it at the polls. Moving beyond the 2013 Federal Election, the
article then explores the development of the party’s programmatic self-presentation
and its voter support at the 2014 European and regional elections. All analyses are
based on surveys conducted by the German Longitudinal Election Study (GLES).
5
To explore voting behaviour at the 2013 Federal Election, a pre – post-election panel
survey is used.
6
Both waves were conducted by telephone. The pre-election wave
was designed as a rolling cross-section (RCS) survey covering the entire election cam-
paign.
7
For the analyses of voting behaviour at the 2014 European and State Elections
cross-sectional data from non-representative online surveys are used.
8
THE AFD AT THE 2013 FEDERAL ELECTION
The Euro Crisis and the Emergence of the AfD
The sovereign debt crisis in the Eurozone that resulted from the 2008 financial and
economic crisis and drove some Euro member states to the verge of financial break-
down was clearly one of the most serious and far-reaching problems that the federal
government of CDU/CSU and FDP had to face during the legislative period preceding
the 2013 election. In connection with the bailout measures for Greece, Ireland, Portu-
gal, Spain and Cyprus, Germany assumed very considerable financial risks. The crisis
THE ‘AFD IN THE ELECTORATE’ 125

policy of the government followed the logic that financial assistance for heavily
indebted Euro states should not be refused but made conditional on quite severe aus-
terity measures.
9
Only the Left Party was clearly against this policy. The SPD and the
Greens agreed to the rescue packages and supported the government’s policy in parlia-
ment although over time they became more critical of the strictness of the govern-
ment’s austerity course.
10
Voices that on the other hand found this approach still too
generous could sporadically be heard from the ranks of the FDP, but remained incon-
sequential in terms of actual policy.
In the indebted countries of the European South the crisis policy of the European
Union (EU), the European Central Bank (ECB) and International Monetary Fund
(IMF), which had been crucially influenced by the German government, led to
massive legitimacy crises. The governments that were forced to implement the auster-
ity measures came under attack from groups affected by these measures’ consequences
and had to face enormous losses at elections.
11
But in a donor country like Germany
whose government had burdened its citizens with considerable financial risks the legiti-
macy of the crisis policy was also precarious.
12
It is therefore hardly surprising that the
established parties shied away from thematising the Euro crisis during the 2013 election
campaign. To be sure, all party manifestos included positions concerning preferred
crisis policies. The government parties CDU/CSU and FDP pleaded for a continuation
of the chosen course, while the SPD and the Greens wanted to soften the burden placed
on the debtor countries and complement the austerity course by investment pro-
grammes. The Left demanded a complete policy reversal away from austerity.
However, party manifestos remain usually unread and in the campaign communication
addressed to the general public the Euro crisis was widely neglected. Campaign strate-
gists seemed to consider this issue ‘toxic’.
13
Two features thus characterised treatment
of the Euro crisis in the campaigns of the Bundestag parties: dethematisation with
regard to the saliency dimension,
14
and an only partial coverage of the political spec-
trum with regard to the position dimension,
15
ranging from restrictive and conditional
to more generous financial assistance to the debtor countries, but excluding the option
of refusing any support at all.
16
It was exactly this vacancy in the political debate that the AfD aimed to fill at its
foundation on 6 February 2013. Under the leadership of Bernd Lucke, a professor of
macro-economics, it presented itself as a single-issue party with a rather narrow
focus on the European debt crisis.
17
Its slim political agenda was dominated by
liberal– conservative Euro-scepticism. Its platform demanded an ‘orderly dissolution’
of the European monetary zone and a return to national currencies or smaller, more
homogeneous monetary associations. Germany should insist on a unilateral withdrawal
from the Eurozone. The AfD thus emphasised an issue that was widely neglected in the
other parties’ communications and assumed a position on this issue that the established
parties had left unoccupied, thereby promising, as its name indicated, an ‘alternative’
to the policies of the established parties.
18
Although the Euro clearly dominated the rhetoric of the AfD, as an aside it also
thematised a few other issues, although far less extensively. In its electoral platform
monetary and European policy as well as fiscal and budgetary policy ranked foremost,
but it also addressed themes like ‘rule of law and democracy’, ‘old-age pensions and
family’ as well as education, energy and integration.
19
Some of these topics also
126 GERMAN POLITICS

appeared in the AfD’s campaign communication, although again much less dominant
than its core theme. For instance, on its campaign posters next to criticism of the mon-
etary union voters could sometimes also encounter demands for a stricter regulation of
immigration.
20
Importantly, during the campaign the AfD’s messages gained consider-
able visibility. Generous donations from business people and other wealthy benefactors
provided the party with a campaign budget that allowed it to organise a much more
vigorous campaign than is usually feasible for small extra-parliamentary parties.
21
Pro-
minent spokespersons, among them well-known renegades from established parties of
the centre– right, university professors (especially economists) and talk-show celebri-
ties also drew the news media’s attention to the AfD.
Crystallisation of AfD Support during the Election Campaign
Casting a vote is the end of a more or less protracted process of decision making.
22
This
section explores how the preferences of those voters that eventually chose the AfD
with their second votes developed during the last 10 weeks before the 2013 Federal
Election. Voters of the parties then present in the Bundestag are used as yardstick
for comparison. At the beginning of the 2013 election campaign the AfD suffered
the competitive disadvantage of being relatively unknown. But during the run-up to
the election the party became more recognised among voters – presumably at least
in part as a consequence of its professional and publicly quite visible campaign.
This can be seen when inspecting how the shares of those respondents who were
able to classify the AfD on thermometer scales developed over time. Early in the cam-
paign about a third of all respondents claimed to have not enough knowledge to evalu-
ate this party and resorted to ‘don’t know’ answers. This changed about three weeks
before Election Day on 22 September. With increasing campaign intensity the share
of those unable to volunteer an opinion about the AfD decreased constantly to less
than 20 per cent immediately before the election.
Figure 1 shows that, paralleling this development, the great majority of those even-
tually voting for the AfD made up their minds to vote for this party not until the very
last days of the campaign. An advantage of the combined RCS-panel design of the
survey is that the figure visualises how many of those voters eventually ending up
choosing the AfD (as registered by the post-election wave) had already claimed an
intention to support this party at different stages of the election campaign. It shows
how the AfD’s electorate built up during the campaign in a stage-wise fashion. For
the sake of comparison, similar information is also given for the electorates of the
CDU/CSU, the SPD, the Greens, the Left and the FDP.
Obviously, the developmental trajectories differed substantially across parties. It
appears as if the TV debate between the CDU/CSU and SPD Chancellor candidates,
Angela Merkel and Peer Steinbru
¨ck, on 1 September 2013,
23
had sent a signal to
voters to start the process of decision making. The picture appears static up until
about that date. However, the parties had no level-playing ground at that point. In
July and August 70 per cent of the later voters of the CDU/CSU were already sure
about their voting decision while voters of the SPD and the Greens followed closely
with only slightly smaller shares of already crystallised supporters. The early
support bases of the FDP and the Left were quite a bit weaker. Only less than half
of their later voters already expressed a vote intention in favour of these parties. The
THE ‘AFD IN THE ELECTORATE’ 127

AfD, however, was even worse off at the early phase of the campaign. Less than a third
of its later voters already claimed to support this party.
At least two out of three later CDU/CSU voters had thus in July already tied them-
selves down to this party. During the last three weeks of the campaign support of the
Christian Democrats grew further by another 10 percentage points. The electorates of
the Greens and the SPD evolved similarly, although at a somewhat lower level and this
pattern was less linear in the case of the Social Democrats. Among later voters of the
FDP and the Left the mobilisation curves started at lower levels, but went up more
steeply. Nevertheless almost half of them had not yet reached their final decisions
even immediately before the election and therefore qualify as ‘last-minute deciders’.
24
It is striking how different the picture yet again looks for the electorate of the AfD.
Even 10 days before the election there was no increase at all in these persons’ vote
intentions for the AfD. Up until immediately before Election Day we see a straight
line persisting at a level of just about 30 per cent.
25
But right at the end of the campaign
a sudden boost occurred. The share of crystallised AfD voters increased so sharply that
it caught up with the FDP and the Left. Altogether, seven out of 10 AfD voters thus
arrived at their final voting decision only during the last days of the campaign or
even directly on Election Day itself.
Taken together, these findings suggest that the electorate of the AfD was composed
of two groups: a majority of about 70 per cent that decided only at the door of the
polling station or very shortly before, and a minority of about 30 per cent that had
FIGURE 1
DEVELOPMENT OF PARTY VOTERS’ VOTE INTENTIONS DURING THE 2013 FEDERAL
ELECTION CAMPAIGN
Notes: Entries are percentages of respondents expressing vote intention for respective party within each time period, to the
basis of later voters of the respective party; data are weighted by education and design weight.
Source: GLES RCS/panel survey 2013.
128 GERMAN POLITICS

made up their minds long before. When exactly this happened is impossible to tell for
lack of data. It must have been sometime between February 2013, when the AfD
emerged as a new party, and July, when the RCS survey’s period of observation
began. For easier distinction the two groups will in the remainder of this article be
addressed as ‘early supporters’ and ‘late supporters’. The following analyses will
show that at the Federal Election these two groups choose the AfD for quite different
reasons.
Backgrounds of AfD Support at the 2013 Federal Election
Dependent variables. Why did voters support the AfD at the 2013 Federal Election?
Using the RCS/Panel survey of the 2013 GLES these questions will be explored by
means of multivariate analyses. Two series of models are presented in Table 1 that
compare supporters of the AfD to those of all other parties in order to explore
which characteristics fostered preferences for the AfD. Models 1 – 3 focus on vote
intentions registered by the pre-election wave of the survey. The second line of analy-
sis consists of similarly structured panel models of voters’ actual choices (Models 4 –
6). Both dependent variables are dichotomous and contrast preferences for the AfD
(coded 1) with preferences for other parties (0). Non-voters are not taken into
account. All estimates are derived from binary logistic regression analyses. A
crucial difference between the two series of models is that analyses of vote choices
include vote intentions expressed during the election campaign (1 ¼AfD, 0 ¼
other party, non-voter or undecided) as an additional control variable. Whereas the
cross-sectional pre-election models show which features of respondents were con-
nected to a higher or lower likelihood to consider voting for the AfD during the cam-
paign, the panel models reveal which attributes fostered a final decision for the AfD
among those voters who had not yet expressed such a vote intention in the pre-election
wave. The models for vote intentions thus indicate why early supporters were drawn to
this party. In contrast, the panel models of vote choice reveal which factors were
important for the choices of late supporters.
Strategy of analysis. To explore AfD support at the 2013 Federal Election a broad
range of predictors is taken into account. The well-known notion of a ‘funnel of caus-
ality’
26
is used as a heuristic for selecting independent variables and organising the
analysis. As a model for explaining voting behaviour it links individuals’ vote
choices at a specific election over several steps of increasing generality to socio-pol-
itical lines of conflict over material and immaterial goods. Relying on this heuristic
suggests a stepwise strategy of analysis which begins with basic socio-demographic
attributes of voters (Models 1 and 4). Since socio-political cleavages typically find
expression in the form of general and stable political worldviews and loyalties, a
set of attitudinal predispositions is included in the next step (Models 2 and 5). The
specific circumstances of a particular election enter voters’ calculus by way of a
broad range of attitudes and beliefs that refer to currently prevalent issues and to
the candidates competing for votes. Furthermore, although they are not originally
included in the notion of the funnel of causality, orientations relevant for strategic
voting should also be taken into account when analysing voting behaviour in a
multi-party system like Germany. These short-term orientations are responsive to
the current flow of events, but they are also coloured by voters’ structural and
THE ‘AFD IN THE ELECTORATE’ 129

TABLE 1
PREDICTORS OF VOTE INTENTIONS AND VOTE CHOICES FOR THE AFD AT THE 2013 FEDERAL ELECTION
Vote intentions Vote choices
Model 1:
socio-
demographics
Model 2:
plus
attitudinal
predispositions
Model 3:
plus
short-term
factors
Model 4:
Socio-
demographics
Model 5:
plus
attitudinal
predispositions
Model 6:
plus
short-term
factors
Campaign time 1.76+1.73 2.17+–––
Pre-election vote intention AfD – – – 295.74∗∗∗ 257.13∗∗∗ 139.63∗∗∗
Gender: male 2.34∗∗∗ 2.20∗∗∗ 2.36∗∗ 1.37 1.39 1.49+
Age 3.13 4.31∗3.54 2.02
-1
1.35
-1
4.00
-1
Region: West 1.31
-1
1.37
-1
1.61
-1
+2.20
-1
∗∗∗ 2.27
-1
∗∗∗ 2.56
-1
∗∗∗
Education: completed secondary 1.23 1.39 1.41 1.12
-1
1.06
-1
1.38
Employment/occupation: worker 2.32
-1
2.25
-1
2.08
-1
1.04
-1
1.07
-1
1.42
Employment/occupation: new
middle class
1.28
-1
1.22
-1
1.58 1.68
-1
1.61
-1
1.12
-1
Employment/occupation:
unemployed
1.11 1.19 2.22 1.40 1.40 1.61
Employment/occupation: retired 2.27
-1
∗2.32
-1
∗1.19
-1
1.78
-1
1.75
-1
1.14
-1
Employment/occupation: otherwise
not gainfully employed
1.49
-1
1.39
-1
1.22 3.45
-1
∗3.45
-1
∗2.22
-1
Member of trade union 1.91
-1
∗1.48
-1
1.85
-1
+1.32 1.41 1.50+
Left–right identification 10.37∗∗∗ 2.21 3.69∗∗ 1.34
Party identification: CDU/CSU 7.41
-1
∗∗∗ 3.23
-1
∗∗∗ 4.03
-1
∗∗∗ 2.22
-1
∗
Party identification: SPD 7.94
-1
∗∗∗ 7.69
-1
∗∗∗ 3.06
-1
∗∗∗ 2.50
-1
∗∗
Party identification: FDP 3.03
-1
∗3.23
-1
+2.10
-1
1.35
-1
Party identification: Greens 10.86
-1
∗∗∗ 6.67
-1
∗∗ 1.96
-1
∗1.27
-1
Party identification: Left 3.77
-1
∗5.26
-1
∗3.12
-1
∗3.70
-1
∗
Most important problem: Euro crisis 2.28∗∗ 1.53+
Position issue: help countries
affected by Euro crisis
11.11
-1
∗∗∗ 5.26
-1
∗∗∗
130 GERMAN POLITICS

Fear of Euro crisis 1.84 3.99∗∗
Economic situation: current personal 2.07 1.30
-1
Economic situation: current
sociotropic
1.54
-1
1.08
-1
Economic situation: retrospective
sociotropic
1.21 2.13
-1
Economic situation: prospective
sociotropic
8.33
-1
∗∗ 2.78
-1
+
Most important problem: economy
general
1.13 1.34
Most important problem: labour
market
1.32
-1
1.59
-1
Most important problem: social
justice
1.54
-1
1.13
Most important problem: ‘NSA’
scandal
1.37
-1
1.14
-1
Most important problem: energy
policy
1.27
-1
1.54
-1
Most important problem: education
policy
1.67
-1
1.45
-1
Most important problem:
immigration policy
1.33 1.19
Most important problem: family
policy
1.12
-1
4.76
-1
∗
Position issue: government should
redistribute
2.38
-1
∗2.56
-1
∗
Position issue: immigrants should
adapt
1.06 5.13∗∗∗
Performance federal government 9.09
-1
∗∗∗ 1.39
-1
Chancellor preference: Merkel
(CDU/CSU)
1.04
-1
2.13
-1
∗
Chancellor preference: Steinbru
¨ck
(SPD)
1.05
-1
1.59
-1
(Continued)
THE ‘AFD IN THE ELECTORATE’ 131

TABLE 1
CONTINUED.
Vote intentions Vote choices
Model 1:
socio-
demographics
Model 2:
plus
attitudinal
predispositions
Model 3:
plus
short-term
factors
Model 4:
Socio-
demographics
Model 5:
plus
attitudinal
predispositions
Model 6:
plus
short-term
factors
Likelihood of AfD passing 5 per cent
threshold
57.00∗∗∗ 6.45∗∗∗
Following media polls 1.62+1.09
Campaign interest 1.15
-1
4.35
-1
∗∗
McKelvey & Zavoina’s Pseudo-R
2
.10 .29 .55 .23 .29 .51
(N) (4078) (4078) (4078) (3546) (3546) (3546)
Notes: Entries are odds ratios; all independent variables transformed to range 0– 1; unweighted data. ∗∗∗p,.001; ∗∗p,.01; ∗p,.05; +p,.10.
Source: GLES RCS/Panel survey 2013.
132 GERMAN POLITICS

attitudinal dispositions which serve as perceptional screens.
27
Together, they form the
third block of predictors which is included in the final steps of the respective analyses
(Models 3 and 6). Proceeding in three stages of growing complexity allows not only to
understand which factors increased the odds of voters favouring the AfD instead of
another party, but also how more specific factors mediated the impact of more
distant and general ones.
Independent variables. All independent variables are taken from the pre-election
wave. Models 1 and 4 include only predictors located at the distant end of the funnel
of causality. They encompass a variety of basic socio-demographic characteristics,
including respondents’ gender (1 ¼male, 0 ¼female), residential region (1 ¼
West Germany, 0 ¼East Germany), age, level of education (1 ¼completed sec-
ondary education, 0 ¼less), occupational status (dummy variables for workers,
new middle class, unemployed persons, pensioners and persons for other reasons
not included in the workforce; reference category: old middle class) as well as
trade-union membership (dummy variable). In addition, Model 1 also includes a
time variable (derived from the dates of interviews) in order to capture the trajectory
of AfD support during the election campaign. Models 2 and 5 build on Models 1
and 4 by adding measures of stable attitudinal predispositions. Ideological left –
right positions have been shown to be important structuring factors for German
voters’ electoral preferences.
28
Partisanship exerts an even more powerful influence
on electoral behaviour.
29
For both vote intentions and vote choices respondents’
self-placement on 11-point left – right scales as well as a set of dummy variables
indicating their party identification (CDU/CSU, SPD, FDP, Greens and Left; refer-
ence category: apartisans including ‘don’t know’) are therefore added at the second
stage of modelling.
The block of independent variables most proximate to the vote in the funnel of
causality concerns short-term orientations relating to the specific circumstances of
the 2013 Federal Election. It is added in Models 3 and 6. An important type of such
situational factors are voters’ orientations towards the issues and themes prevalent at
the time of the election. The central topic of the AfD’s rhetoric during the 2013 cam-
paign was the Euro crisis and the way it was handled by the government of the CDU/
CSU and the FDP and the opposition parties in the Bundestag. It is therefore of predo-
minant interest whether and how AfD support was responsive to perceptions and atti-
tudes concerning the Euro crisis as well as to more general economic assessments.
Several indicators allow assessing whether the AfD profited from voters’ concern
about the Euro crisis. In spite of the low attention the Bundestag parties paid to this
issue in their campaign communication, almost 20 per cent of the voters mentioned
it spontaneously when asked which was, in their opinion, the most important
problem facing the country. The analyses therefore include a dummy variable that indi-
cates whether respondents saw the Euro crisis as one of Germany’s two most important
political problems.
30
Moreover, respondents were asked to position themselves with
reference to Germany’s policy in the Euro crisis, using a five-point Likert scale on
the following statement: ‘[i]n times of the European sovereign debt crisis Germany
should grant financial help to EU member states that have strong economic or financial
problems’. Respondents were also invited to indicate how the Euro crisis affected them
emotionally, using a five-point scale from ‘no fear at all’ to ‘extremely great fear’.
THE ‘AFD IN THE ELECTORATE’ 133

Furthermore, the analysis includes a standard set of perceptions commonly referred to
in analyses of economic voting:
31
retrospective, prospective and current sociotropic
evaluations of the economy as well as assessments of respondents’ current personal
economic situation (all measured on five-point scales).
Although the Euro crisis was the dominant theme of the AfD in the months after
its foundation, it occassionally also addressed other topics. It is therefore of interest
whether support of the AfD was also spurred by other issue orientations than scepti-
cism on the European common currency and perhaps more general economic
worries. The analysis includes a series of dummy variables indicating other policy
areas mentioned by respondents in response to the most important problems’ ques-
tion as well as respondents’ positions on core issues of the two basic cleavages of
German politics.
32
To elicit positions on the socio-economic cleavage, respondents
were asked to indicate their position on redistribution, using a Likert scale refering
to the statement: ‘[t]he government should take measures to reduce income dispar-
ities’. The immigration issue was chosen to indicate positions on the libertarian–
authoritarian dimension, and the statement read: ‘[i]mmigrants should be obliged
to adapt to the culture of Germany’. In addition to specific position and valence
issues, parties’ electoral prospects may also be influenced by more general
valence assessments. The analysis therefore includes retrospective evaluations of
the performance of the federal government (11-point thermometer scale from – 5
to +5).
Besides specific or generalised issue and policy-related orientations, party prefer-
ences may also be influenced by voters’ perceptions and evaluations of the candidates
nominated by the parties and competing for their esteem and sympathy. The most
important candidates under the German electoral regime are the Chancellor candidates
of the two large parties the CDU/CSU and the SPD.
33
The final models for vote inten-
tions and choices therefore take respondents’ Chancellor preferences into account.
They include a set of two dummy variables indicating whether respondents named
the incumbent Chancellor Angela Merkel (CDU/CSU) or her challenger Peer Stein-
bru
¨ck (SPD) as their preferred future head of government (reference category: unde-
cided or none of the two candidates).
For a number of reasons, most prominently its origin in the American two-party
system, the heuristic of the funnel of causality neglects the possibility that electors
might cast their votes strategically. However, in recent years electoral research has
become more sensitive to the fact that in multi-party systems electoral choices must
not always be sincere, but can also take the form of strategic responses to expectations
of how parties will fare at the election.
34
These expectations depend on two factors –
persons’ assumptions about other voters’ likely behaviour at the polls, and their under-
standing of how electoral rules transform votes into parliamentary mandates. A crucial
element of German electoral law is the 5 per cent threshold that parties need to pass in
order to obtain seats in the Bundestag. Assuming that electors wish to avoid ‘wasted
votes’, it should make a difference for small parties like the AfD whether they are
expected to pass the 5 per cent threshold or not.
35
It is therefore of interest to
explore if and how electoral support for the AfD also depended on voters’ expectation
about whether or not it would pass the 5 per cent threshold. As up-to-date information
about the electorate’s support of particular parties is a crucial basis for such
134 GERMAN POLITICS

expectations the models not only take these expectations into account, but also whether
the respondents followed the results from opinion polls (dummy variable), and whether
they were generally interested in the election campaign (five-point scale from ‘very
interested’ to ‘not interested at all’).
FINDINGS
The first line of Table 1 confirms the descriptive finding of Figure 1 that voters’ like-
lihood to consider voting for the AfD increased during the election campaign (although
only with p,.10). Socio-demographic background attributes explain support for the
AfD overall only to a limited extent. The effect patterns of Models 1 and 4 show a mix
of similarities and differences. For instance, the AfD gained more support among men
than among women. Moreover, it also was preferred more often by voters from the new
states of East Germany. However, the gender effect was much more pronounced for
vote intentions, whereas the reverse was true for voters’ region of residence. According
to Model 1, pensioners were less prone to consider voting for the AfD. Regarding
vote choices, on the other hand, persons not gainfully employed for other reasons
than being unemployed or retired appeared rather less inclined to support this party
(Model 4). However, these relationships disappear once short-term orientations are
taken into account. Regarding vote intentions, trade-union membership decreased
voters’ likelihood to favour the AfD, but the effect reversed its sign for the AfD’s
late supporters.
Inspecting Models 2 and 5 we see that with regard to both vote intentions and vote
choice AfD support had a clear ideological profile. Matching the party platform’s ideo-
logical profile, voters that placed themselves right of centre were more strongly than
others drawn to the AfD. For early supporters this relationship was more pronounced
than for late supporters. However, its evaporation for both vote intentions and choices
indicates that support for the AfD rooted in general ideological leanings was mediated
by more specific orientations (Models 3 and 6). Party identification, by way of contrast,
mostly remains a significant predictor even when controlling for short-term factors.
Persons expressing a vote intention for the AfD during the campaign as well as
those casting their ballot for it on Election Day came predominantly from the ranks
of apartisans.
Short-term factors were much more important for both vote intentions and vote
choices than socio-demographic and attitudinal predispositions. As is to be expected
in view of the AfD’s core issue, beliefs and attitudes concerning the Euro crisis
appear very influential on whether or not voters preferred this party. In close correspon-
dence to its strong emphasis on this issue, voters who mentioned it spontaneously as
the country’s most important problem were more strongly attracted by the AfD.
However, this concerns especially early supporters, whereas the relationship is
weaker and only marginally significant for late supporters. The first panel of Figure
2visualises the substantial implications of the difference between the effects of
these problem perceptions on AfD vote intentions (dotted lines) and vote choices
(solid lines) by means of average marginal effects,
36
computed on the basis of
Models 3 and 6.
37
Viewing the Euro crisis as Germany’s most important problem
increased the likelihood to develop a vote intention for the AfD by about two
THE ‘AFD IN THE ELECTORATE’ 135

FIGURE 2
PREDICTED PROBABILITIES OF AFD SUPPORT AT THE 2013 FEDERAL ELECTION FOR
SELECTED BELIEFS AND ATTITUDES
Notes: Predicted probabilities of vote intentions (dotted lines) and voting decisions (solid lines) in favour of AfD with 5 per
cent confidence intervals, unweighted data.
Source: GLES RCS/panel survey 2013.
136 GERMAN POLITICS

percentage points, whereas the corresponding effect on vote choices was only about
half as large. Closely corresponding to the AfD’s rhetoric, scepticism or outright oppo-
sition to financial help for indebted Euro countries on the part of voters also led to
strongly increased odds of preferring that party at the polls. Although this relationship
was of considerable strength also for late supporters, it was again more pronounced for
those voters that already early on had decided to vote for it. According to the estimates
displayed in the second panel of Figure 2, a strict refusal of financial help resulted in a
five-fold increased chance of a vote intention in favour of the AfD than strong support
for a generous crisis policy.
Affective responses to the Euro crisis also had some impact on AfD support,
although – in stark contrast to the valence and position-based issue effects of saliency
perceptions and policy opinions – only with regard to those voters that flocked to this
party late in the campaign. For early vote intentions fearful reactions to the crisis made
no difference. In contrast, the probability for voters that found the crisis emotionally
disturbing to switch to the AfD late in the campaign amounted to 7 per cent and
was thus four percentage points higher than for carefree voters (see the third panel
of Figure 2). Among the four dimensions of general economic evaluations included
in Models 3 and 6 only prospective sociotropic assessments were relevant for
support of the AfD. Regardless of orientations specifically related to the Euro crisis,
the party thus also took advantage of economic pessimism among voters. Prior to
the 2013 Federal Election only a few citizens expected a deterioration of the country’s
economic situation in the year to come.
38
But those who did were strongly drawn to the
AfD, although especially early in the campaign. Model 3 predicts a probability of 5 per
cent to prefer the AfD for very pessimistic voters compared to only 1 per cent for very
optimistic voters. Regarding vote choices the difference between voters with negative
and positive economic expectations are much smaller and statistically only marginally
significant (see the fourth panel of Figure 2).
Besides orientations related to the AfD’s core topic, several other beliefs and
opinions concerning the material substance of politics were also relevant for voters’
inclination to support this party. For early supporters two further policy-related
relationships appear significant in Model 3. Persons dissatisfied with the performance
of the federal government of the CDU/CSU and the FDP and socio-economically con-
servative persons that were opposed to redistribution policies tended to consider voting
for the AfD during the campaign. Compared with completely satisfied voters, extre-
mely discontented voters’ probability to express a vote intention for the AfD was
four percentage points higher (see the fifth panel of Figure 2). However, retrospective
performance evaluations did not play a role for voters who found their way to the AfD
only late (Model 6). This was different for socio-economic conservatism. As can be
seen in the sixth panel of Figure 2, the difference between strict opponents and
ardent supporters of redistributive policies amounted to two to three percentage
points for both vote intentions and voting decisions. Moreover, preferences for the
AfD were considerably fostered by immigration-sceptical attitudes, although only
among voters that made up their minds late in the campaign. Compared with fervent
proponents of a multiculturalist immigration policy, the vehement opponents’ prob-
ability of switching to the AfD immediately before the election increased by about
four percentage points. In striking contrast, the seventh panel of Figure 2 shows
THE ‘AFD IN THE ELECTORATE’ 137

only a flat line for vote intentions. For the preferences of the AfD’s early supporters
immigration policy was completely irrelevant.
Candidate-related attitudes were important for late supporters, but not for early
supporters, and they concerned only Angela Merkel of the CDU/CSU. According to
Model 6, those who disliked the prospect of the incumbent Chancellor continuing
in office more often chose the AfD. Lastly, we see evidence in both Models 3 and
6 that strategic considerations were important for AfD support. In line with the
‘wasted vote’ hypothesis, voters’ expectation that the AfD would make it across the
5 per cent threshold emerges as an extremely powerful predictor of the party’s
early support, and to a lesser extent also of late support. To be sure, only a few
voters (about 8 per cent) expected the AfD to enter the Bundestag at least ‘probably’,
but among these persons the party found remarkable support during the election cam-
paign as well as in the final spurt that led it close to actually passing the electoral
threshold. The last panel of Figure 2 shows that voters who were certain that the
AfD would pass the 5 per cent threshold had a probability of 11 per cent to switch
to the AfD compared to 4 per cent for those being convinced that it would not gain
any seats in the parliament. For early support, the impact of these expectations was
even larger.
Opinion polls are a useful source of information that helps voters to develop expec-
tations on likely election outcomes.
39
However, in line with the AfD’s low support in
the electorate during most of the campaign that is evident from Figure 1, the data pub-
lished by pollsters did not give any reason to expect the party to pass the 5 per cent
threshold.
40
Correspondingly, Model 6 does not show a significant connection
between AfD votes and voters’ attention to media polls. Moreover, it is quite remark-
able that late switches to the AfD happened much more likely among those with no or
only little interest in the election campaign than among those that followed it closely.
For its early support, campaign interest did not make a difference.
The AfD emerged in early 2013 as a single-issue party, but in view of its cam-
paign communication the question soon was raised whether it should rather be classi-
fied as a right-wing populist party. Its campaign communications appeared
ambiguous in this regard. Right-wing populist motives, in particular scepticism
about immigration, were increasingly present in the party’s public rhetoric, although
the Euro crisis remained clearly dominant throughout the pre-election period.
According to the data presented above as ‘party in the electorate’ the AfD was simi-
larly two-faced. It started out with an instrumental single-issue electorate that was
mainly drawn to it by the high saliency it placed on the Euro crisis and its
liberal– conservative position with regard to how it should be dealt with. However,
close to Election Day these early supporters were supplemented with a major
influx of voters responsive to a motive that in other countries has been identified
as a mainspring of support for right-wing populist parties – an exclusivist stance
on immigration.
41
Although this motive was not central in the party’s campaign
rhetoric, it was one of the main predictors of late supporters’ votes for the AfD.
And this group of voters, in turn, contributed many more votes to the AfD’s favour-
able outcome than the early supporters for whose choices immigration-sceptical atti-
tudes were entirely irrelevant.
138 GERMAN POLITICS

THE AFD AT THE 2014 EUROPEAN AND STATE ELECTIONS
The AfD’s Self-Presentation after the 2013 General Election
At the Federal Election of September 2013 the AfD failed to gain seats in the Bundes-
tag. However, the campaign gave it a lot of publicity and the margin of its failure was
small enough to send a signal to voters that the party might be a viable competitor at
future elections. Partly as a self-fulfilling prophecy (due to the facilitating role of
voters’ expectations concerning the party’s likelihood to surpass the electoral
TABLE 2
PREDICTORS OF VOTE INTENTIONS FOR THE AFD AT THE 2014 EUROPEAN ELECTION AND
STATE ELECTIONS
European
Election
State Elections Saxony,
Brandenburg, Thuringia
Gender: male 1.63 1.19
-1
Age 2.94
-1
3.30
Region: West 2.50
-1
∗–
Education: completed secondary 1.08
-1
1.20
Employment/occupation: worker 1.57 1.24
-1
Employment/occupation: new middle class 1.69 2.05
-1
+
Employment/occupation: unemployed 1.29 1.37
-1
Employment/occupation: retired 2.09 2.53
-1
+
Employment/occupation: otherwise not gainfully employed 1.77
-1
2.11
-1
Member of trade union 1.01 1.47
Left-right identification 17.95∗∗ 17.67∗∗∗
Party identification: CDU/CSU 6.33
-1
∗∗ 5.10
-1
∗∗∗
Party identification: SPD 4.50
-1
∗∗ 6.02
-1
∗∗∗
Party identification: FDP – 5.71
-1
∗
Party identification: Greens – 18.25
-1
∗∗
Party identification: Left 10.64
-1
∗5.08
-1
∗∗∗
Most important problem: Euro crisis 3.42+1.16
Position issue: help countries affected by Euro crisis – 1.88
-1
Position issue: exclude indebted countries from Euro zone 5.38∗–
Position issue: reintroduce D-Mark 4.01∗–
Economic situation: current personal 2.88 1.22
Economic situation: current sociotropic 1.96 5.60
-1
∗
Economic situation: retrospective sociotropic 20.66∗∗ 3.52∗
Economic situation: prospective sociotropic 12.55
-1
∗4.46
-1
∗
Position issue: government should redistribute – 1.72
-1
Position issue: immigrants should adapt – 2.86∗
Position issue: restrict immigration 6.02+-
Performance state government 10.71
-1
∗∗ 1.79
-1
Preferred head of state government: CDU – 1.27
-1
Preferred head of state government: SPD – 1.61
-1
Preferred head of state government: Left – 1.25
-1
Likelihood of AfD passing 5 per cent threshold – 2.65∗
Campaign interest – 1.10
-1
Political interest 15.68∗∗ –
McKelvey & Zavoina’s Pseudo-R
2
.62 .39
(N) (671) (1118)
Notes: Entries are odds ratios; all independent variables transformed to range 0 –1; unweighted data. PId FDP
and Greens excluded for European Election due to insufficient numbers of cases. ∗∗∗p,.001; ∗∗p,.01; ∗p
,.05; +p,.10.
Source: GLES tracking survey T24; cumulated GLES State Election boosts Saxony, Brandenburg, Thuringia
2014.
THE ‘AFD IN THE ELECTORATE’ 139

threshold), history confirmed this view. In all six parliaments elected until 2015, the
AfD gained mandates.
In the immediate aftermath of the Federal Election, it had been unclear whether the
AfD would continue to present itself as a single-issue party mainly concerned about
European monetary policies or whether it would rather move towards a stronger and
more general commitment to right-wing populist topoi.
42
Impressionistic evidence
suggests that the AfD gradually opted for the latter course. Accompanied by quite
intense struggling within the party leadership it abandoned its near exclusive focus
on the Euro crisis and broadened its thematic spectrum. It now addressed a wider
set of issues and shifted its emphasis towards views that more closely resembled the
typical menu of right-wing populist ideas and arguments, although it still never
went so far that it unequivocally qualified as a right-wing populist party.
43
‘Courage for Germany’ was the slogan the AfD adopted for its European Election
campaign. It expressed a more general commitment to re-nationalising European poli-
tics, although stopping short of taking the step from ‘soft’, that is, sectoral, to ‘hard’,
that is, fundamental and principled Euroscepticism by requesting the European Union
to be dismantled or Germany to terminate its membership.
44
While retaining its pre-
vious emphasis on domestic implications of European unification, this allowed the
AfD to broaden its outlook. At the three East German State Elections four months
later, the Euro crisis was finally relegated to the background. Instead, the AfD
mainly campaigned on a range of issues more or less directly related to general
themes of open borders and immigration, such as dangers for German society presum-
ably resulting from Islam, trans-border crime as well as ‘unregulated’ immigration and
the presumed overload of the German welfare system resulting from it. It also thema-
tised issues of life-style politics, for instance by denouncing gay marriage and advocat-
ing the ‘three-child family’. At the same time, anti-establishment rhetoric paired with
demands for more direct democracy to give ‘the’ people more say – another typical
element of right-wing populism
45
– gained ground in the party’s communication.
46
Backgrounds of AfD Support at the 2014 Elections
How did voters respond to this shift in the party’s self-presentation? A look at vote
intentions at the four elections that took place in 2014 concludes our exploration of
AfD support. It is based on online surveys conducted in the GLES project prior to
these elections. The surveys are not based on random samples, but this should not
seriously distort statistical relationships between voters’ preferences and background
attributes in multivariate models.
47
The models displayed in Table 2 are set up in
such a way that they correspond as closely as possible to Model 3 in Table 1. This
could be achieved somewhat more completely for the State Elections than for the Euro-
pean Election. Still, important predictors are missing in both models so that the ana-
lyses are not fully comparable and have therefore primarily heuristic value. To
obtain sufficient numbers of cases the data for the three State Elections have been
merged into one integrated data set that covers the three elections about equally.
In important, yet varying respects, the backgrounds of AfD support at the 2014
European and State Elections resembled the profile of AfD preferences at the 2013
Federal Election. Apartisans yet again emerged as the strongest source of AfD
support. Moreover, at the European Election East Germans were yet again more
140 GERMAN POLITICS

strongly inclined to vote for the AfD than West Germans. More importantly, the effect
of ideological leanings on vote intentions for the AfD appears not only relevant but
indeed far stronger at all 2014 elections than at the Federal Election in the previous
year. Most notably, in stark contrast to the Federal Election it does not evaporate
once short-term attitudes are taken into account. Even in the full model that includes
all short-term predictors, rightist voters appear very strongly drawn to the AfD.
However, these differences should be interpreted with caution due to the variation
in survey modes and coverage.
At the Federal Election, early support for the AfD had appeared mainly as a con-
sequence of instrumental issue-voting pertaining to the Euro crisis and the economy
more generally. Voting behaviour at the 2014 European Election displayed a similar
pattern, whereas this was not the case for the East German State Elections. At the Euro-
pean Election, but not at the three State Elections, viewing the Euro crisis as Ger-
many’s most important problem yet again seems to have stimulated votes for the
AfD. Likewise, Euro-related policy positions were important at the European Election,
but not at the State Elections. The State Election surveys contained the same item as the
2013 Federal Election survey which registered respondents’ position on helping
indebted Euro countries. According to Table 2, its effect is negative like the one dis-
played in Table 1, but it attains not even marginal statistical significance. The Euro-
pean Election survey included different items to register attitudes on the European
common currency. Their effects on AfD preferences are strong and statistically signifi-
cant. Persons in favour of excluding indebted countries from the Eurozone and of
returning to the D-Mark as national currency were considerably more likely to vote
for the AfD.
Pessimistic assessments of the general economic prospects played a role at both
elections, but much more strongly at the European Election than at the State Elections.
In a similar vein, assessments of the federal government’s performance were important
only at the European Election, but not at the State Elections. Voting behaviour at the
European Election thus bore resemblance to the AfD’s early supporters’ behaviour at
the 2013 Federal Election in its rather instrumental and issue-focused character,
whereas these motives were almost or completely irrelevant at the State Elections in
Saxony, Brandenburg and Thuringia. Economic concerns were important at these elec-
tions, but only generally, whereas considerations specifically pertaining to the Euro
crisis seem to have disappeared from the calculus that led voters to support the AfD.
To elicit the role of immigration-sceptical attitudes at the East German State Elec-
tions the same item as at the Federal Election could be used. According to Table 2,itis
much more strongly related to AfD preferences than opinions and beliefs concerning
the Euro crisis. In the European Election survey again, a different instrument was
used for measuring attitudes on immigration, but it is similarly strongly related to
AfD support. Voters who were in favour of limiting immigration displayed a strong
inclination to prefer the AfD. Hence, whereas AfD support at the European Election
resembled early supporters’ preferences at the Federal Election quite strongly with
regard to its issue-specific instrumentality, it was also to some extent responsive to
xenophobic attitudes which had emerged late in the Federal Election campaign as a
new important predictor of choices for the AfD. At the East German State Elections,
AfD support was even more clearly determined by such orientations. In essence,
THE ‘AFD IN THE ELECTORATE’ 141

AfD votes at the three State Elections held in 2014 appear primarily as expression of a
general ideological tendency towards the right, a negative current and prospective
assessment of the overall economic situation and immigration-scepticism. Compared
to the beginning of the AfD’s electoral history early in the 2013 Federal Election cam-
paign the background of this party’s support has thus changed considerably – parallel-
ing and presumably responding to the party’s redefinition of its programmatic profile.
With regard to the AfD’s development after the 2013 Federal Election, demand-side
changes thus corresponded to those on the supply side.
CONCLUSION
Against the background of the European sovereign debt crisis, the AfD appeared on the
political stage early in the election year 2013 as a single-issue party. Its manifesto was
slim and primarily focused on criticism of the European currency union. However,
during the Federal Election campaign its public rhetoric as an aside also alluded to
right-wing populist motives, most notably a critical stance on immigration. After the
2013 Federal Election it moved further in that direction, downplaying the Euro
crisis and focusing its campaign communication more strongly on xenophobic topoi
and other motives from the repertoire of right-wing populism. Systematic analyses
of its rhetoric found it inappropriate to categorise the AfD as a full-fledged right-
wing populist party.
48
Rather, the party seems to have presented itself sufficiently
ambiguously to open up the possibility of being seen differently by different voters.
The analysis of voting behaviour at the 2013 Federal Election presented in this
article suggests that the party profited from this ambiguity at the polls. As ‘party in
the electorate’ the AfD’s appearance matched the two-faced character displayed by
the party in its public communication – although already with reversed weights at
the 2013 Federal Election, and with a clear trend further away from Euro concerns
to more general xenophobic motives thereafter.
That the 2013 electorate of the AfD consisted of two groups – a smaller group of
voters who already at an early stage of the campaign were determined to choose the
AfD, and a much larger group of late supporters that made up their minds in favour
of this party only during the final days of the campaign or even on Election Day
itself – is a striking outcome of this research. Just as important is the fact that early
and late supporters preferred the AfD for quite different reasons. Of course, like any
new party the AfD drew advantage from the increasing electoral availability of
voters.
49
Weakly anchored voters were an important source of its support. It obtained
considerably more votes from apartisans than from individuals with firm party attach-
ments. Likewise, voters from the states of post-socialist East Germany were more
strongly drawn to the AfD than voters from the ‘old’ Federal Republic. But beyond
these similarities the ‘two electorates’ of the AfD differed considerably.
Right from the start of the election campaign, the AfD was supported by a small but
stable group of voters that preferred it mainly because of its core issue. In several
respects, these early supporters appear as instrumental issue voters. Findings suggest
that they found the AfD appealing because of its self-presentation as a single-issue
party predominantly concerned about the Euro crisis and its domestic economic impli-
cations. As the Bundestag parties remained mostly silent about this issue, the AfD
142 GERMAN POLITICS

became attractive for voters for whom it was highly salient. Moreover, the party gained
strong support from voters that preferred a much more restrictive policy course with
regard to financial aid for indebted countries of the Eurozone than advocated by any
of the parties in the Bundestag. In a situation where these parties almost unanimously
agreed on the necessity of financial help and only discussed its conditions, the AfD –
true to its name – became an attractive alternative for voters that were opposed to any
such help. From both a valence and a positional point of view
50
the AfD thus appeared
as an attractive choice for voters concerned about its central theme, the crisis of the
European common currency.
The impact of general economic pessimism was also more pronounced among
persons who knew very early that they would vote for the AfD. That early rather
than late moves towards the AfD were more likely among persons dissatisfied with
the performance of the federal government also fits into this picture. Voters more gen-
erally displaying a liberal– conservative mindset on socio-economic issues also tended
to express a vote intention in favour of the AfD. This suggests that attitudes on the Euro
issue were not unrelated to the traditional cleavages of German politics. Importantly,
there also seems to have been a pronounced strategic component to early supporters’
vote intentions for the AfD, since their preferences depended very strongly on the
party’s perceived electoral viability. Given what polls indicated with regard to the
size of the AfD’s support base, voters had little reason to expect the party to surpass
the 5 per cent threshold. But some did nonetheless, and within their ranks early
support for this party was especially probable.
Overall, however, the early supporters of the AfD made up only about a third of the
party’s electorate. Its good result at the polls was mainly brought about by the second
group of voters which was twice as large and joined its ranks only immediately before
or even on Election Day itself. These late supporters’ ultimate conversion towards the
AfD seems to have been driven mostly by other, arguably rather expressive motives.
To begin with, while valence and positional considerations concerning the Euro
crisis were not irrelevant for these voters’ choices, they were clearly less important.
At the same time, fearful reactions to this event brought the AfD votes from late sup-
porters while they made no difference for early supporters.
Even more importantly, immigration-sceptical attitudes contributed significantly to
late supporters converting to the AfD, whereas they were completely irrelevant for its
early supporters. This suggests that the AfD’s campaign rhetoric made voters aware of
the immigration-sceptic stances that the party assumed alongside its economically
couched, ‘soft’ Euroscepticism, even though it put much more emphasis on the
latter. This let the AfD appear an attractive choice for persons averse to inclusivist
immigration policies. In a range of studies xenophobic attitudes of this kind have
emerged as the most powerful predictor of support for right-wing populist parties in
Europe.
51
It appears that the AfD was able to tap into a voter reservoir that just like
in comparable countries exists in Germany as well, but could thus far not express its
sentiments at the polls because the legacy of German history prevented a party of
that type taking root in the party system.
52
Also paralleling findings from studies of
electoral support of right-wing populist parties, the AfD was more attractive for
male voters and for voters with right-of-centre ideological leanings.
53
The influence
of the latter, however, was mediated by short-term attitudes. That AfD late supporters’
THE ‘AFD IN THE ELECTORATE’ 143

choices at the Federal Election were expressive rather than instrumental is further
underlined by the finding that last-minute conversions to this party were more likely
for voters who had little or no interest in the election campaign. Importantly, the find-
ings of this study indicate that this group contributed many more votes than the early
supporters for whose choices immigration-sceptical attitudes were entirely irrelevant.
When the AfD first stepped onto the electoral stage in early 2013, it presented itself
as a single-issue party concerned about the policies of the Bundestag parties with
regard to the Euro crisis. However, during the Federal Election campaign it added
right-wing populist themes, particularly immigration-sceptical motives, to its rhetoric,
although it still retained its main focus on the Euro crisis. After the Federal Election, it
more clearly began to shift its emphasis, and at the three State Elections held in East
Germany about a year after the Federal Election, the AfD eventually campaigned
mainly on xenophobic topoi whereas the Euro crisis and even European unification
more generally no longer played a prominent role. Presumably responding to these
changes on the supply side the driving forces of AfD support in the electorate gradually
shifted as well. From instrumental issue-oriented voting based on voter concerns about
the Euro crisis that were insufficiently addressed by the Bundestag parties, the basis of
its support moved to xenophobic attitudes and thus over time came to resemble even
more clearly the profile of electoral support for right-wing populist parties in neigh-
bouring countries. At the 2013 Federal Election, the influx of late-deciding supporters
of this type was the main reason for the favourable result of the AfD. At the European
Election in May 2014 concerns about both the Euro crisis and xenophobic motives
were important sources of votes. At the three eastern German State Elections, at
long last, the former source of electoral support vanished whereas the latter dominated.
Over four stages and with regard to both the party organisation and the ‘party in the
electorate’ the AfD has covered a considerable distance on the road from single-
issue to right-wing populist party.
However, not only in its public rhetoric, but also as ‘party in the electorate’ at least
up until 2015 it did not yet qualify as a full-fledged right-wing populist party. For one,
xenophobic attitudes were not the only source of its support. It also was considered an
attractive choice by voters with liberal – conservative views on socio-economic issues.
In some respects the AfD’s support profile also did not correspond to expectations from
extant research on the typical demand-side conditions of votes for right-wing populist
parties.
54
In particular, this concerns the demographics of AfD voters. Neither workers
nor unemployed persons were markedly more likely than other voters to choose this
party. We also found no effect of lower levels of formal education. While xenophobic
attitudes are often seen as the attitudinal core of electoral support for right-wing popu-
list parties, authoritarian and populist attitudes are also considered important ingredi-
ents of such behaviour.
55
Whether orientations of this kind also played a role for the
AfD’s support at the 2013 and 2014 elections could not be examined for lack of appro-
priate data.
After its electoral successes in 2014 the AfD mainly made it into the news through
bitter infighting about its future course. In July 2015, an extraordinary party convention
finally brought the split of the organisation. The founding leader Bernd Lucke was
voted down by a clear majority of the present party members, and replaced by the
chairwoman of the Saxon party branch Frauke Petry, a staunch advocate of a
144 GERMAN POLITICS

‘national– conservative’ course. Tellingly, when polled about the ‘major problems of
the country’ attendees of the party conference placed ‘uncontrolled immigration’
ahead of the Euro crisis. Leading proponents of a ‘liberal’ agenda, among them the
AfD’s founder Lucke himself, immediately left the party to found a new one, labelled
ALFA, the ‘Allianz fu
¨r Fortschritt und Aufbruch’ (‘Alliance for Progress’). The AfD’s
new leadership’s quickly asserted that the party will not move to the right and its plat-
form will remain unchanged. But in all likelihood the schism will ultimately lead to a
clarification of the party’s profile. It is bound to lose the programmatic ambiguity from
which it profited at the 2013 Federal Election and the 2014 European Election, and
consequently also (at least) one of its two electorates. Like the proverbial sorcerer’s
apprentice, by covertly appealing to xenophobic sentiments the AfD’s founding
leaders raised the spectre of right-wing populism which eventually they were unable
to keep under control. It remains to be seen whether the AfD will go all the way
and transform itself into a full-fledged right-wing populist party, and whether it will
be able to maintain a stable base of electoral support among German voters.
ACKNOWLEDGEMENTS
I am indebted to Klaus Armingeon, Josephine Ho
¨rl, Oskar Niedermayer, Julia Parthey-
mu
¨ller and Robert Rohrschneider for advice and constructive comments on a previous
version of this article. I gratefully acknowledge the hospitality of the Research School
of Social Sciences of the Australian National University Canberra which allowed me to
complete work on this article during a Visiting Fellowship.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.
ABOUT THE AUTHOR
Ru
¨diger Schmitt-Beck is Professor of Political Science and Political Sociology at the
University of Mannheim. He is chairperson of the German Society for Electoral
Research (DGfW) and one of the principal investigators of the German Longitudinal
Election Study (GLES).
NOTES
1. See Robert Grimm, ‘The Rise of the German Eurosceptic Party Alternative fu
¨r Deutschland, between
Ordoliberal Critique and Popular Anxiety’, International Political Science Review 36/3 (2015),
pp.264– 78.
2. On the notion of ‘party as propagandiser’, see Paul Allen Beck, Party Politics in America (New York:
Longman, 1997), p.14.
3. See Nicole Berbuir, Marcel Lewandowsky and Jasmin Siri, ‘The AfD and Its Sympathisers: Finally a
Right-Wing Populist Movement in Germany?’, German Politics 24/2 (2015), pp.154–78; Simon
Tobias Franzmann, ‘Die Wahlprogrammatik der AfD in vergleichender Perspektive’, MIP 20/1
(2014), pp.115–24; Kai Arzheimer, ‘The AfD: Finally a Successful Right-Wing Populist Eurosceptic
Party for Germany?’, West European Politics 38/3 (2015), pp.535–56; Oskar Niedermayer, ‘Eine neue
Konkurrentin im Parteiensystem? Die Alternative fu
¨r Deutschland’, in Oskar Niedermayer (ed.), Die
THE ‘AFD IN THE ELECTORATE’ 145

Parteien nach der Bundestagswahl 2013 (Wiesbaden: Springer VS, 2015), pp.175– 207; Marcel
Lewandowsky, ‘Eine rechtspopulistische Protestpartei? Die AfD in der o
¨ffentlichen und politikwis-
senschaftlichen Debatte’, Zeitschrift fu
¨r Politikwissenschaft 25/1 (2015), pp.119– 34.
4. On the notion of ‘party in the electorate’, see Beck, Party Politics in America, pp.12– 13.
5. The GLES is funded by the Deutsche Forschungsgemeinschaft (German Research Foundation, DFG)
under its long-term programme. All GLES data are freely available from http://www.gesis.org/gles.
For general information on the project, see http://www.gles.eu and Ru
¨diger Schmitt-Beck, Hans Rattin-
ger, Sigrid Roßteutscher and Bernhard Weßels, ‘Die deutsche Wahlforschung und die German Longi-
tudinal Election Study (GLES)’, in Frank Faulbaum and Christof Wolf (eds), Gesellschaftliche
Entwicklungen im Spiegel der empirischen Sozialforschung (Wiesbaden: VS-Verlag fu
¨r Sozialwis-
senschaften, 2010), pp.141–72.
6. Hans Rattinger, Sigrid Roßteutscher, Ru
¨diger Schmitt-Beck, Bernhard Weßels and Christof Wolf,
Rolling Cross-Section-Wahlkampfstudie mit Nachwahl-Panelwelle (GLES 2013; GESIS Datenarchiv,
Ko
¨ln, 2014: ZA5703 Datenfile Version 1.0.0, doi:10.4232/1.11803) (pre-election wave: N ¼7882,
post-election wave: N ¼5353). For further information on this survey, see Julia Partheymu
¨ller,
Ru
¨diger Schmitt-Beck and Christian Hoops, Kampagnendynamik bei der Bundestagswahl 2013: Die
Rolling Cross-Section-Studie im Rahmen der ‘German Longitudinal Election Study’, MZES Working
Paper No. 154 (Mannheim: Mannheim Centre for European Social Research, 2013), available from
http://www.mzes.uni-mannheim.de/publications/wp/wp-154.pdf (accessed on 1 March 2015).
7. See Richard Johnston and Henry E. Brady, ‘The Rolling Cross-Section Design’, Electoral Studies 21/2
(2002), pp.283– 95. Data collection started on 8 July 2013 and ended on 21 Sept. 2013. Random
samples of on average about 104 voters were interviewed every day.
8. Hans Rattinger, Sigrid Roßteutscher, Ru
¨diger Schmitt-Beck, Bernhard Weßels and Christof Wolf,
Langfrist-Online-Tracking, T24 (GLES) (GESIS Datenarchiv, Ko
¨ln, 2014: ZA5724 Datenfile
Version 1.0.0, doi: 10.4232/1.11963) (N ¼1044; conducted 9 –23 May 2014); Langfrist-Online-Track-
ing zur Landtagswahl in Sachsen 2014 (GLES) (GESIS Datenarchiv, Ko
¨ln, 2014: ZA5738 Datenfile
Version 1.0.0, doi: 10.4232/1.12047) (N ¼503; conducted 15 –30 Aug. 2014); Langfrist-Online-Track-
ing zur Landtagswahl in Brandenburg 2014 (GLES) (GESIS Datenarchiv, Ko
¨ln, 2014: ZA5738 Daten-
file Version 1.0.0, doi:10.4232/1.12054) (N ¼507; conducted 29 Aug. – 13 Sept. 2014); Langfrist-
Online-Tracking zur Landtagswahl in Thu
¨ringen 2014 (GLES) (GESIS Datenarchiv, Ko
¨ln, 2014:
ZA5740 Datenfile Version 1.0.0, doi: 10.4232/1.12057) (N ¼504; conducted 29 Aug. – 13 Sept.
2014). The author is indebted to Rebecca Steffen for cumulating the three State Election datasets.
9. See Simon Bulmer, ‘Germany and the Eurozone Crisis: Between Hegemony and Domestic Politics’,
West European Politics 37/6 (2014), pp.1244 – 63.
10. See Hubert Zimmermann, ‘A Grand Coalition for the Euro: The Second Merkel Cabinet, the Euro Crisis
and the Elections of 2013’, German Politics 23/4 (2014), pp.322 – 36; Grimm, ‘The Rise of the German
Eurosceptic Party’.
11. See Hanspeter Kriesi, ‘The Political Consequences of the Financial and Economic Crisis in Europe:
Electoral Punishment and Popular Protest’, Swiss Political Science Review 18/4 (2012), pp.518 – 22;
Michael S. Lewis-Beck, Marina Costa Lobo and Paolo Bellucci, ‘Economic Crisis and Elections:
The European Periphery’, Electoral Studies 31/3 (2012), pp.469–642; Sonja Alonso, ‘Wa
¨hlen ohne
Wahl. Demokratie und die Staatsschuldenkrise in der Eurozone’, in Wolfgang Merkel (ed.), Demokratie
und Krise (Wiesbaden: Springer VS, 2015), pp.245– 74.
12. See Fritz W. Scharpf, Legitimacy Intermediation in the Multilevel European Polity and Its Collapse in
the Euro Crisis, MPfG Discussion Paper No. 12/6 (Ko
¨ln: Max-Planck-Institut fu
¨r Gesellschafts-
forschung, 2012).
13. Daniel Bro
¨ssler, ‘Merkels Gift wirkt’, Su
¨ddeutsche Zeitung, 11 Aug. 2013, p.4.
14. See Hans-Dieter Klingemann, Richard I. Hofferbert and Ian Budge, Parties, Policies, and Democracy
(Boulder, CO: Westview, 1994).
15. See Anthony Downs, An Economic Theory of Democracy (Boston, MA: Addison Wesley, 1965).
16. See Robert Rohrschneider and Stephen Whitefield, ‘Party Positions about European Integration in
Germany: An Electoral Quandary?’, this issue.
17. See Niedermayer, ‘Eine neue Konkurrentin im Parteiensystem?’.
18. Grimm, ‘The Rise of the German Eurosceptic Party’.
19. See http://www.Alternativefuer.de/partei/wahlprogramm/ (accessed 20 Dec. 2013); see also Berbuir
et al., ‘The AfD and Its Sympathisers’; Niedermayer, ‘Eine neue Konkurrentin im Parteiensystem?’.
20. See Berbuir et al., ‘The AfD and Its Sympathisers’.
21. See Mona Krewel, ‘Die Wahlkampagnen der Parteien und ihr Kontext’, in Ru
¨diger Schmitt-Beck, Hans
Rattinger, Sigrid Roßteutscher, Bernhard Weßels, Christof Wolf et al., Zwischen Fragmentierung und
Konzentration: Die Bundestagswahl 2013 (Baden-Baden: Nomos, 2014), pp.35– 45.
146 GERMAN POLITICS

22. See Paul F. Lazarsfeld, Bernard Berelson and Hazel Gaudet, The People’s Choice: How the Voter
Makes Up His Mind in a Presidential Campaign (New York/London: Columbia University Press,
1968); Richard Johnston, Andre
´Blais, Henry E. Brady and Jean Crete, Letting the People Decide:
Dynamics of a Canadian Election (Stanford, CA: Stanford University Press, 1992); Andrew Gelman
and Gary King ‘Why Are American Presidential Election Campaign Polls So Variable When Voters
Are So Predictable?’, British Journal of Political Science 23/4 (1993), pp.409–51; Robert
S. Erikson and Christopher Wlezien, The Timeline of Presidential Elections: How Campaigns Do
(and Do not) Matter (Chicago, IL: University of Chicago Press, 2012); Thomas Plischke, Wann
Wa
¨hler entscheiden. Abla
¨ufe von Entscheidungsprozessen und der Zeitpunkt der Wahlentscheidung
(Baden-Baden: Nomos, 2014).
23. See Ju
¨rgen Maier, Thorsten Faas and Isabella Glogger, ‘Das TV-Duell’, in Schmitt-Beck et al.,
Zwischen Fragmentierung und Konzentration, pp.281– 92.
24. See Steven H. Chaffee and Rajiv Nath Rimal, ‘Time of Voting Decision and Openness to Persuasion’,
in Diana C. Mutz, Paul M. Sniderman and Richard A. Brody (eds), Political Persuasion and Attitude
Change (Ann Arbor, MI: University of Michigan Press, 1996), pp.267 – 91.
25. Thirty-five per cent of those not yet supporting the AfD intended to vote for another party (CDU/CSU:
10 per cent; SPD, FDP and the Left: 6 per cent each; Greens: 4 per cent; other parties: 3 per cent), 5 per
cent contemplated abstaining, 28 per cent were undecided.
26. See Angus Campbell, Philip E. Converse, Warren E. Miller and Donald E. Stokes, The American Voter
(New York: Wiley, 1960), pp.24 –32; Warren E. Miller and J. Merrill Shanks, The New American Voter
(Cambridge, MA: Harvard University Press, 1996), pp.189–211; Michael S. Lewis-Beck, William
G. Jacoby, Helmut Norpoth and Herbert F. Weisberg, The American Voter Revisited (Michigan: Uni-
versity of Michigan Press, 2009), pp.22 – 8.
27. See Larry Bartels, ‘Beyond the Running Tally: Partisan Bias in Political Perceptions’, Political Behav-
ior 24/2 (2002), pp.117–50.
28. See Anja Neundorf, ‘Die Links-Rechts-Dimension auf dem Pru
¨fstand: Ideologisches Wa
¨hlen in Ost-
und Westdeutschland 1990–2008’, in Ru
¨diger Schmitt-Beck (ed.), Wa
¨hlen in Deutschland, PVS-Son-
derheft 45 (Baden-Baden: Nomos, 2012), pp.227–50.
29. See Harald Schoen and Cornelia Weins, ‘Der sozialpsychologische Ansatz zur Erkla
¨rung von Wa
¨hler-
verhalten’, in Ju
¨rgen W. Falter and Harald Schoen (eds), Handbuch Wahlforschung (Wiesbaden: VS-
Verlag, 2005), pp.206–25.
30. The variable was generated by means of an automatic analysis of verbatim answers. If they contained
one or several of the following strings as words or word parts they were attributed to the category ‘Euro
crisis’: Euro;Schulden;Finanz;Griechen;Rettung;Banken;Waehrungsunion;Haushalt;Krise.
31. See, for example, Charlotte Kellermann and Hans Rattinger, ‘Wahrnehmungen der Wirtschaftslage und
Wahlverhalten’, in Hans Rattinger, Oscar W. Gabriel and Ju
¨rgen W. Falter (eds), Der gesamtdeutsche
Wa
¨hler. Stabilita
¨t und Wandel des Wa
¨hlerverhaltens im wiedervereinigten Deutschland (Baden-Baden:
Nomos, 2007), pp.329–56.
32. See, for example, Hanspeter Kriesi, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier
and Timoteos Frey, West European Politics in the Age of Globalization (Cambridge: Cambridge Uni-
versity Press, 2008).
33. See Franz U. Pappi and Susumu Shikano, ‘Personalisierung der Politik in Mehrparteiensystemen am
Beispiel deutscher Bundestagswahlen seit 1980’, Politische Vierteljahresschrift 42/3 (2001),
pp.355– 87; on candidate assessments at the 2013 Federal Election, see Aiko Wagner, ‘Spitzenkandi-
daten’, in Schmitt-Beck et al., Zwischen Fragmentierung und Konzentration, pp.267– 80.
34. See, for example, Paul R. Abramson, John H. Aldrich, Andre
´Blais, Matthew Diamond, Abraham
Diskin, Indiri H. Indridason, Daniel J. Lee and Renan Levine, ‘Comparing Strategic Voting under
FPTP and PR’, Comparative Political Studies 43/1 (2010), pp.61–90.
35. See Thomas Gschwend, ‘Ticket-Splitting and Strategic Voting under Mixed Electoral Rules: Evidence
from Germany’, European Journal of Political Research 46/1 (2007), pp.1 –23; Sascha Huber, Thomas
Gschwend, Michael F. Meffert and Franz Urban Pappi, ‘Erwartungsbildung u
¨ber den Wahlausgang and
ihr Einfluss auf die Wahlentscheidung’, in Oscar W. Gabriel, Bernhard Weßels and Ju
¨rgen W. Falter
(eds), Wahlen und Wa
¨hler. Analysen aus Anlass der Bundestagswahl 2005 (Wiesbaden: VS-Verlag,
2009), pp.561– 84.
36. See Michael Hanmer and Kerem Ozan Kalkan, ‘Behind the Curve: Clarifying the Best Approach to Cal-
culating Predicted Probabilities and Marginal Effects from Limited Dependent Variable Models’,
American Journal of Political Science 57/1 (2013), pp.263– 77.
37. Because of the large share of late supporters within the actual electorate of the AfD the chances calcu-
lated for voting decisions are on average higher than for vote intentions.
38. See Markus Steinbrecher, ‘Wirtschaftliche Entwicklung und Eurokrise’, in Schmitt-Beck et al.,
Zwischen Fragmentierung und Konzentration, pp.225– 38.
THE ‘AFD IN THE ELECTORATE’ 147

39. See Thorsten Faas, Christian Mackenrodt and Ru
¨diger Schmitt-Beck, ‘Polls That Mattered: Effects of
Media Polls on Voters’ Coalition Expectations and Party Preferences in the 2005 German Parliamentary
Election’, International Journal of Public Opinion Research 20/3 (2008), pp.299–325.
40. See Krewel, ‘Die Wahlkampagnen der Parteien und ihr Kontext’, p.37.
41. See Pippa Norris, Radical Right: Voters and Parties in the Electoral Market (Cambridge: Cambridge
University Press, 2005); Cas Mudde, Populist Radical Right Parties in Europe (Cambridge: Cambridge
University Press, 2007); Kai Arzheimer, Die Wa
¨hler der extremen Rechten 1980 – 2002 (Wiesbaden:
VS-Verlag, 2008); Hanspeter Kriesi, Edgar Grande, Martin Dolezal, Marc Helling, Dominic Ho
¨glinger,
Swen Hutter and Bruno Wu
¨est, Political Conflict in Western Europe (Cambridge: Cambridge Univer-
sity Press, 2012).
42. See Niedermayer, ‘Eine neue Konkurrentin im Parteiensystem?’.
43. See Arzheimer, ‘The AfD’.
44. See Arzheimer, ‘The AfD’; Grimm, ‘The Rise of the German Eurosceptic Party’.
45. See Mudde, Populist Radical Right Parties.
46. See Matthias Lohre, ‘Die Angst-Partei’, Die Zeit, 7 Aug. 2014, p.6; Josef Joffe, ‘Protest- statt Prof-
Partei’, Die Zeit, 18 Sept. 2014, p.12; Grimm, ‘The Rise of the German Eurosceptic Party’, pp.272 – 3.
47. See Ina E. Bieber and Evelyn Bytzek, ‘Online-Umfragen: Eine geeignete Erhebungsmethode fu
¨r die
Wahlforschung? Ein Vergleich unterschiedlicher Befragungsmodi am Beispiel der Bundestagswahl
2009’, Methoden-Daten-Analysen 6/2 (2012), pp.185– 211.
48. Berbuir et al., ‘The AfD and Its Sympathisers’; Franzmann, ‘Die Wahlprogrammatik der AfD’; Arzhei-
mer, ‘The AfD’.
49. See Stefano Bartolini and Peter Mair, Identity, Competition and Electoral Availability (Cambridge:
Cambridge University Press, 1990).
50. See Donald E. Stokes ‘Spatial Models of Party Competition’, American Political Science Review 57/2
(1963), pp.368–77; Donald E. Stokes, ‘Valence Politics’, in Dennis Kavanagh (ed.), Electoral Politics
(Oxford: Clarendon Press,1992), pp.141– 64.
51. See Norris, Radical Right; Mudde, Populist Radical Right Parties; Arzheimer, Die Wa
¨hler der extre-
men Rechten; Kriesi et al., Political Conflict in Western Europe.
52. See Berbuir et al., ‘The AfD and Its Sympathisers’; Arzheimer, ‘The AfD’.
53. See Mudde, Populist Radical Right Parties; Arzheimer, Die Wa
¨hler der extremen Rechten.
54. See Norris, Radical Right; Mudde, Populist Radical Right Parties; Arzheimer, Die Wa
¨hler der extre-
men Rechten.
55. See Mudde, Populist Radical Right Parties.
148 GERMAN POLITICS