ArticlePDF Available

How to find the ‘winning formula’? Conducting simulation experiments to grasp the tactical moves and fortunes of populist radical right parties


Abstract and Figures

This article puts party strategies at the centre of the analysis of radical right-wing challengers’ fortunes. It extends most previous studies because the shifts of both established and populist parties are envisioned as a complex dynamic system, in which party leaders adaptively learn from feedback and voters continually update their party choices. We argue that agent-based modelling is a fruitful tool to systematically map out the implications of hypotheses on the behaviour of parties, voters and their interactions. Our argument is empirically illustrated by using computer simulations to examine the remarkable rise of the Dutch anti-immigration party PVV. Outcomes reveal that an adaptive strategy leads to large shifts towards the socio-economic left and considerably boost its electoral strength. The more general contribution of this article is that we show how to unravel the mechanisms by which flexible populist parties can find winning positions.
Content may be subject to copyright.
Original Article
How to find the ‘winning formula’? Conducting
simulation experiments to grasp the tactical moves
and fortunes of populist radical right parties
Jasper Muis*
and Michel Scholte
Department of Sociology, Vrije Universiteit Amsterdam, De Boelelaan 1081, Amsterdam,
HV 1081, The Netherlands.
*Corresponding author.
Both authors contributed equally to this article.
Abstract This article puts party strategies at the centre of the analysis of radical
right-wing challengers’ fortunes. It extends most previous studies because the shifts
of both established and populist parties are envisioned as a complex dynamic
system, in which party leaders adaptively learn from feedback and voters
continually update their party choices. We argue that agent-based modelling is a
fruitful tool to systematically map out the implications of hypotheses on the
behaviour of parties, voters and their interactions. Our argument is empirically
illustrated by using computer simulations to examine the remarkable rise of the
Dutch anti-immigration party PVV. Outcomes reveal that an adaptive strategy
leads to large shifts towards the socio-economic left and considerably boost
its electoral strength. The more general contribution of this article is that we show
how to unravel the mechanisms by which flexible populist parties can find winning
Acta Politica advance online publication, 14 September 2012; doi:10.1057/ap.2012.21
Keywords: agent-based simulation; adaptive behaviour; party strategies; populist
radical right; immigration and integration issues
Issues of immigration and the integration of foreigners have constituted
the most prominent and controversial field of political contention in
West European polities since the early 1990s (Koopmans et al, 2005, p. 3).
r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
At the same time, anti-immigrant parties experienced a dramatic growth in
electoral support in several European countries (Evans, 2005; Van der Brug
et al, 2005). Examples are the French Front National, the Flemish Vlaams
Blok/Vlaams Belang and the Austrian FPO
The rise of right-wing populist parties in Europe can first be seen as the
consequence of a transformation in the electorate (Ivarsflaten, 2008). The issue
of immigration has emerged as a new and dominant cultural conflict dimension
that divides the ‘winners’ and ‘losers’ of globalization (Kriesi et al, 2008).
However, the distribution and intensity of demand-side preferences cannot, by
itself, explain the electoral strength of the radical right. Already since the
seminal work of Kitschelt (1995), it has been acknowledged that the success of
these parties is to a large extent dependent on the ideological choices and
moves of both the established parties and radical right challengers. Kitschelt
(1995, p.vii) put this as follows: ‘the success of the extreme Right is contingent
upon the strategic choices of the moderate conservative parties as well as the
ability of the extreme-rightist leaders to find the electorally “winning formula”
to assemble a significant voter constituency’. He posited that successful
radical right parties combine culturally exclusionist/authoritarian positions
with pro-market positions on socio-economic policies. In this current of
thought, Goodwin (2006) argued that, for a full understanding of the successes
and failures of anti-immigrant parties, we need to take into account that
populist parties are able to actively adapt and shape their own fate (see
likewise, for example, Ignazi, 2003; Mudde, 2007).
This article focuses on how strategic positioning explains the rise of the
populist radical right. The central question is how it can find successful
positions. In line with the plea of Kitschelt (2007), our first general
contribution is to theoretically map out more sophisticated behavioural
models of parties’ strategies than generally used. A strategy defines if and how
a party ideologically shifts in the policy landscape. It does not refer to the
programmatic content or ideology a political party advocates (that is, a specific
party position, see, for example, Meguid, 2008), but to the decision rules by
which actors pursue their goals (Axelrod and Cohen, 2000).
In line with, for example, Kollman et al (1992), Laver (2005) and Bendor
et al (2011), we assume that political party leaders are adaptive, instead of
rationally forward-looking actors. This means that party leaders can learn
from feedback and adapt their behaviour in the policy field accordingly. As
party leaders continuously react to past outcomes and adjust their platform in
response to each other and their environment, the dynamics of party
competition resemble a complex adaptive system (Laver and Sergenti, 2011).
To put some empirical flesh to the theoretical bones, we will examine the
remarkable rise of the controversial Dutch anti-immigration party Party for
Freedom (PVV) headed by Geert Wilders. Because of its tough stance on
Muis and Scholte
2r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
immigration and integration issues, scholars qualify the PVV as a ‘radical right
party’ (see, for example, De Lange and Art, 2011). The PVV is an interesting
case because its electoral support has spectacularly increased since its
foundation in 2005. Thus, our study aims to provide a better understanding
of the trajectories of radical populist parties in modern democracies in general
and how these parties, or any newcomer party for that matter, sometimes
succeed to break through.
Thus, the second, more specific aim of this article is to elaborate
propositions on the role of the party strategy of both the PVV and the
established parties in the electoral successes of the Dutch radical populist right.
We will execute four simulation experiments on the basis of agent-based
modelling (hereafter: ABM), taking the Dutch political landscape in 2006 as a
baseline. ABM provides a fruitful and innovative tool for exploring the role of
party strategies. A simulation is a simple, artificial representation of the
complex dynamics of electoral competition. We manipulate and investigate the
emergent properties of that system (that is, outcomes on the macro-level, in this
case the vote share of the radical right) in order to provide insight into dynamic
adaptive processes. In sum, to paraphrase Epstein (1999), we aim to ‘grow’ the
history of the rise of the PVV.
The following sections first discuss the role of party strategies and issue
saliency for explaining anti-immigrant party success and why ABM are
necessary and fruitful. Next, we present the design of the simulation model for
the Dutch case and discuss the outcomes. The article ends with the conclusion
and discussion.
Explaining Anti-Immigration Party Success
The demand-side: The salience of immigration and integration issues
Explanations for the fortunes of populist radical right parties can be grouped
into two broad perspectives: one focusing on popular grievances, and one
on political opportunities and party characteristics. Scholars often borrow
the market metaphor from economics and label these sets of factors as
the demand-side, and the external and internal supply-side (Norris, 2005;
Mudde, 2007; Rydgren, 2007; Van der Brug and Fennema, 2007). This
distinction matches the argument of Kitschelt (1995) that it depends on three
premises whether radical right-wing parties can successfully occupy an electoral
First, the extreme right does well when societies have a post-industrial
economic structure that increases the divide between left-libertarian and
How to find the ‘winning formula’?
3r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
right-authoritarian citizens. Kriesi et al (2008) argued that globalization
processes and mass immigration have transformed the meaning of the existing
two-dimensional policy space: resistance to immigration and globalization
have become integrated into and nowadays dominate the cultural policy
dimension. They expect that an increase in saliency of the cultural dimension
in a country enhances electoral successes for right-wing populist parties. Their
findings show that economic issues have lost salience over the years in all
countries they investigated (Great Britain, France, Austria, Switzerland and
the Netherlands), except Germany. This might account for the fact that, thus
far, right-wing populist parties did not have any electoral success in this
We argue that the public debate will play a key role in shaping and
amplifying the salience of the new cultural divide among the population.
Trends in media coverage can explain why party preferences are rather volatile,
while voters had relatively stable attitudes towards immigration and
integration issues over time (Kleinnijenhuis et al, 2007, p. 12; Meuleman
et al, 2009, p. 355). The argument that the news media affects the salience of
issues corresponds to the agenda-setting thesis: issues that appear frequently in
the news tend to become the issues that voters deem important (McCombs and
Shaw, 1972; Rogers et al, 1993).
Agenda-setting does not predict, however, which political party will be
preferred when certain issues are more salient in the voter’s mind. There are
two distinct mechanisms by which issue saliency might affect individual party
choices (Van der Brug, 2004; Tavits, 2008). First, the notion of issue ownership
(Budge and Farlie, 1983) enables a link: certain parties will benefit because they
‘own’ an issue. When voters give more priority to certain issues, a party with a
stronger reputation on dealing with these issues effectively or on prioritizing
these issues will consequently become more attractive. In this current of
thought, publicity for issues that are ‘owned’ by right-wing populist parties
increases their electoral attractiveness.
In this article, however, we employ a second mechanism. In line with
Benoit and Laver (2006), we argue that it is the party’s position that
ultimately counts in political competition. According to such a positional
approach, when voters deem certain issues more important, those parties will
benefit that are closest to their positionsontheseissues;atthesametime,
parties that are ideologically proximate on issues that are not prioritized
by voters will be electorally penalized. In this view, established centre–right
parties might even lose votes to a radical right party when they more strongly
stress issues of migration and multiculturalism in their platforms. They
will only undermine a populist challenger when they move towards their
position, most notably in the direction of more restrictive policies concerning
immigration and integration.
Muis and Scholte
4r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
The supply side: Political party strategies
Kitschelt’s second assumption is that when established parties converge, they
create a configuration that leaves a ‘political space’ in the electoral market that
can be exploited by new challengers. Other scholars have similarly argued that
the behaviour of the mainstream competitors critically shape the electoral
fortunes of niche parties (for example, Meguid, 2008). The ‘adaptation
hypothesis’ of Kriesi et al (2008, p. 14) holds that established parties will
reposition themselves as a result of the rising new cultural dimensions, and thus
update their preferences and identities.
However, the adaptive capacity of established parties might be limited
because of the dilemma between power and ideals. In the Netherlands, for
instance, internal tensions have emerged within the established centre–right
parties CDA and VVD about their stance on the ‘foreigners issue’ (Van
Kersbergen and Krouwel, 2008). These parties face a trade-off: they might
enhance their electoral support by advocating stricter immigration and
integration policies, but such a shift might go against their core ideological
values and thus upset certain factions of the party. McDonald et al (2004)
conclude that parties are generally principled and consistent instead of
opportunistic and vote-grabbing. Parties generally refrain from shifting
policies and are inclined to keep the same position, even if it makes them
unpopular. Likewise, Bale et al (2010) state that the so-called ‘hold’ strategy is
the default setting of parties. Laver and Sergenti (2011) label such an
‘ideologically intransigent’ party a Sticker.
Another reason why political parties might exhibit rigidity is that they are
satisficers. People often choose options that might not be optimal, but satisfy
them enough (Simon, 1955). This implies that party leaders attempt to achieve
at least a certain minimum level of electoral support, but do not move endlessly
around to maximize their vote share (Bendor et al, 2011).
The third premise of Kitschelt corresponds with internal supply-side
accounts, which stress that the strategic moves of the radical right actors
themselves are important too. They do well provided that they find the
‘winning formula’. The willingness and ability to adjust one’s policy stances to
boost electoral appeal is more strongly associated with populist leaders than
with mainstream party leaders. In the public debate, the term populism is more
often used as a label for electoral opportunism and a lack of consistent
ideological principles, than for a certain ideology (Mudde, 2004). The
flexibility to exploit whatever grievances can be mobilized is linked with an
organizational structure of strong central leadership that allows this kind of
party behaviour (Albertazzi and McDonnell, 2008). As Mazzoleni (2003, p. 5)
put it: ‘neo-populism, thanks to its chameleon-like nature, may adapt to
different contexts’. In line with Laver (2005), we use the label Hunter for a
How to find the ‘winning formula’?
5r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
vote-seeking party that continually adapts to polls on electoral support.
Hunters do not necessarily shift to moderate positions: where votes are to be
won depends on the entire party configuration. The optimal vote-seeking
option might be to stay put. Adams et al (2006) found that niche parties were
electorally punished when they brought their policy positions more closely in
line with the average preferences of the public.
A second way to learn from feedback is to adapt to the preferences of your
supporters. Ezrow et al (2011) denote such responsive behaviour as ‘partisan
constituency representation’, which corresponds with an Aggregator strategy
(Laver, 2005). This decision rule means that a party continually updates its
ideological stance by reflecting the mean policy position of its current voters.
Ezrow et al (2011) find empirical support for the application of this strategy.
They conclude that whereas niche parties appear unresponsive to shifts in the
mean voter position, they do indeed respond to shifts in the preferences of their
Regarding the decision-making process of voters, ideological shifts of parties
only matter if citizens are policy-motivated, that is, if their choices are based
on preferences for an ideological position. In this respect, Van der Brug
et al (2005) have shown that voters for the extreme right do not differ from
voters for mainstream parties. Furthermore, whether party responsiveness
is bearing fruit hinges on the assumption that voters are responsive as well.
Adams et al (2011) call this ‘policy-based partisan switching’. Their findings
show that ideological positioning matters: voters generally react to shifts in
their perception of the parties’ stances and adjust their partisan support
Simulating a Complex Adaptive System: Exploring the Rise of the
Dutch PVV
It has become clear that the idea that party strategies matter is not new at all.
However, academic work that explores the implications of hypotheses on party
tactics is scarce. We believe that this lacuna results from the fact that dynamic
assumptions of interacting actors generate a so-called complex adaptive system
(Miller and Page, 2007). This brings us to our argument that we should
envision the political landscape as a complex ecosystem inhabited by voters
and party leaders. In accordance with an evolutionary framework, we do not
assume a priori that it is advantageous or harmful for a particular mainstream
party to shift towards a certain position or not. Likewise, we do not predict
beforehand that there is a specific winning formula for a populist challenger:
the most fertile location will depend on the position and behaviour of all
competitors in that particular setting.
Muis and Scholte
6r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
A primary tool for mapping out theories that involve complex system
dynamics is ABM. This boils down to specifying how individual agents inter-
act with each other and with their environment. Subsequently, computer
simulation is used to gain insights about the emergent properties of the system
as a whole. Axelrod (1997, p. 3) explicated that ABM provides ‘a third way of
doing science’. One starts with a well-defined set of assumptions, such as
deduction. However, unlike deduction, ABM is capable of analysing conse-
quences that cannot be deduced with conventional techniques. It would simply
be too difficult or impossible to model such complex adaptive processes with
verbal theorizing or mathematical deduction. Although ABM is in itself a
solely theoretical exercise, it shares with induction its main method of revealing
these processes through the analysis of a data set, in this case data generated by
a computer simulation. ABM is particularly fruitful to explore the properties
and dynamics of multi-party competition in a multi-dimensional policy space
(Kollman et al, 1992; Laver, 2005).
To illustrate our argument, we will devote the remainder of this article to the
analysis of a concrete case. Our target is the position and vote share of the
Dutch populist radical right Party for Freedom, which was established in
March 2005. Wilders, its controversial leader, left the conservative-liberal VVD
in September 2004 because of disagreement about the party’s stance on the
accession of Turkey to the European Union. The PVV has experienced a
remarkable increase in electoral support. It achieved nine seats during its first
parliamentary elections in November 2006. In the most recent parliamentary
elections in 2010, the party tripled its number of seats and became the third
largest party in the Netherlands with 15.4 per cent of the votes. This was an
absolute increase of almost 10 per cent of the votes, more than any other party.
The continuous line in Figure 1 depicts the opinion poll support for the PVV
(Synovate, 2010). The actual Dutch political landscape in 2006 is taken as a
baseline for the simulations, which makes them empirically embedded instead
of purely theoretical (for example, Boero and Squazzoni, 2005; Muis, 2010).
The computational challenge is to specify plausible rules for the behaviour of
political parties, voters and their interactions that match the ‘true history’.
The Design of the Simulation Experiments
Voter and party positions in 2006
We use a confrontational approach, which assumes that parties and voters are
clearly in favour or against certain policy issues (Kleinnijenhuis and Pennings,
2001). If all parties and voters do so on all issues, then a landscape with clearly
demarcated party positions emerges (Kriesi et al, 2008). In line with Aarts and
How to find the ‘winning formula’?
7r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Thomassen (2008) (see also Pellikaan et al, 2007), we assume that the Dutch
policy space is three-dimensional. The first dimension is the opposition
between support for the welfare state and support for economic liberalism; the
second divide is between the opponents and advocates of multicultural policies;
and the third dimension is opposition to versus support for religious principles
in politics. All parties that were represented in Parliament in 2006 and a
representative group of voters got position scores on the socio-economic,
cultural conflict and religious dimension, represented by, respectively, the x- y-
and z-axis in a three-dimensional policy space.
Party positions are derived from selected questions of the 2006 Chapel Hill
Expert Survey (Hooghe et al, 2008), converted into a 10 to 10 scale (See
Tables 1 and 2). High scores on the socio-economic policy dimension indicate
an emphasis on the market, whereas low scores indicate an emphasis on the
state. The higher the score on the religious dimension, the more strongly a
party supports religious principles in politics. Finally, high and low scores
on the cultural conflict dimension represent, respectively, a preference for
Figure 1: Opinion polls on party preference in the Netherlands between 2004 and 2010.
Source: Synovate (2010).
Muis and Scholte
8r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
restrictive immigration and integration policies, versus a preference for
multicultural policies. As expected, the PVV has the highest score on the
cultural axis (8.53), followed by the VVD (5.18); the Green Left (GL) Party
(4.75) and the progressive liberal D66 (2.80) are most in favour of
multicultural policies. Table 2 furthermore shows that in 2006 the PVV (6.85)
and conservative liberal VVD (6.31) were the most right-wing on the economic
axis, whereas the Socialist Party (7.91) and GL (5.33) were the most left-
wing. Figure 2 shows the parties in a three-dimensional policy landscape in
Voters got empirically based scores on the same three dimensions, based on
four items from the Dutch Parliamentary Election Study 2006 (Kolk et al,
2007). The survey was held among a representative sample of the Dutch voting
population before the 2006 parliamentary elections (n¼2793). Similar to the
party scores, the answer categories are converted into a 10 to þ10 scale (see
Table 3).
Calculating electoral support
The output of main interest is the electoral support for the PVV over time. We
assume that each voter prefers the party that is ideologically most proximate
considering the distance on all three dimensions (using Euclidean distances).
Over time, each month (each round) the distance for each voter towards all
parties is calculated and reported. By varying the assumptions about the party
strategies and agenda-setting, we generate different scenarios for the
development of the PVV over time, each of which can be compared with the
actual trajectory as depicted in Figure 1.
Table 1: Expert survey items on the socio-economic, cultural conflict and religious dimensions
Socio-economic dimension
1. Improving public services versus reducing taxes
2. Deregulation of markets
3. Redistribution from the rich to the poor
Cultural conflict dimension
4. Immigration policy: Support versus oppose tough policy
5. Integration of immigrants and asylum seekers: Favour multiculturalism versus assimilation
6. Ethnic minorities: Support versus oppose more rights
Religious dimension
7. Role of religious principles in politics
8. Social lifestyle (for example, homosexuality): Support versus oppose liberal policies
Note: Items had an 11-point scale, ranging from ‘strongly opposes’ to ‘strongly favours’.
How to find the ‘winning formula’?
9r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Table 2: Policy positions of Dutch political parties on socio-economic, cultural conflict and religious issues; the election outcomes in 2006 and 2010;
and outcomes of the 2006 baseline simulation model (electoral support in percentage)
Policy position (2006) Vote share
elections (in
Vote share
model (2006)
x y z 2006 2010
CDA, Christian Democrats 0.63 2.49 2.80 26.5 13.6 17.9 (0.4)
CU, Christian Party 1.74 1.34 7.33 4.0 3.2 4.1 (0.2)
D66, Progressive Liberals 0.78 2.80 7.86 2.0 6.9 16.4 (0.4)
GL, Green Left Party 5.33 4.75 6.50 4.6 6.7 5.7 (0.2)
PvdA, Labour Party 2.95 1.55 4.71 21.2 19.6 28.8 (0.5)
PVV, Populist Right-Wing Party 6.85 8.53 1.45 5.9 15.4 3.6 (0.2)
SGP, Christian Party 2.95 2.57 8.33 1.6 1.7 1.8 (0.1)
SP, Socialist Party 7.91 0.89 4.43 16.6 9.8 5.2 (0.2)
VVD, Conservative Liberals 6.31 5.18 4.81 14.7 20.5 16.3 (0.4)
Mean (weighted by vote share) 0.70 1.61 1.87 —
Others 2.9 2.4
Note: Standard deviations between brackets.
Source: Hooghe et al (2008). For the SGP, expert scores of Benoit and Laver (2006) were used (rescaled likewise).
Muis and Scholte
10 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
dimension (x-axis)
cultural conflict
dimension (y-axis)
religious dimension
Figure 2: Dutch political parties in a three-dimensional policy landscape in 2006.
Table 3: Position of Dutch voters in the three-dimensional field, based on items from NKO 2006
Socio-economic items
1. Differences in income should be decreased versus increased (v145) 4.1
2. Taxes should be cut (v204) 2.5
3. Tax reduction for people with mortgage should be abolished (v207) 4.1
4. People with a good pension should pay taxes (pay for the AOW) (v210) 1.5
Average position score 0.5
Cultural conflict items
5. One should permit more asylum seekers versus send them back (v155) 2.4
6. Foreigners should be allowed to preserve their own culture versus adjust to
Dutch culture (v185)
7. One should stop the influx of Muslims (v209) 0.7
8. Illegal citizens should get a citizenship permit (v206) 0.1
Average position score 1.6
Religious items
9. Euthanasia should be forbidden versus allowed (v135) 6.3
10. Gay marriage should be forbidden: Agree versus disagree (v211) 4.4
11. Adoption by homosexual couples should be allowed: Agree versus disagree (v200) 1.8
12. Shops should be closed on Sundays: Agree versus disagree (v205) 1.2
Average position score 3.6
Note: Items are translated from Dutch. Between brackets are the original variable labels of the
NKO 2006 data set. Some items had a 1–7 scale, others a 1–4 scale. We recoded all scores into
a10 to þ10 scale. Items v145, v204, v209 and v200 were recoded in reversed order.
Source: Kolk et al (2007).
How to find the ‘winning formula’?
11r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Varying agenda-setting
The relative importance of cultural, economic and religious issues in the eyes of
voters is implemented by adding a ‘weight’ to each policy dimension. We vary
these weights, but make sure that the three weights are constrained to sum up
to one. The distance Dbetween voter Vand party P, while considering the
salience of economic issues i
, cultural issues i
and religious issues i
expressed as:
This implies, for instance, that an increase of i
increases for voters the weight
of the proximity towards parties on the cultural dimension compared with
party’s ideological distance on the two other dimensions.
Varying party strategies
We have argued that populist parties with a strong leadership are more capable
of employing ideological flexibility than mainstream parties. We will use
counterfactual reasoning to corroborate this argument. This means that we
have to simulate both a scenario based on our assumptions and scenarios based
on alternative assumptions, and determine whether the outcome of our
scenario is more accurate. Thus, the strategies of the PVV and established
parties have to be varied per simulation experiment. If our hypothesis is
correct, then the simulated data set based on our assumptions should match
the real political developments more closely than outcomes that are based on
alternative assumptions.
The specification of the party strategies elaborates on the theoretical
background we sketched earlier. First, Satisficing is implemented by setting a
certain share of the vote as the aspiration level. When this level is reached or
surpassed, parties stop moving around. We assume that parties were satisfied
with their current platform as long as it did not yield a loss of more than 25 per
cent of the amount of supporters they initially had in 2006. Furthermore, a
Sticker stays on the party position of 2006, regardless of the opinion poll
results or policy stances of its supporters (Laver, 2005). Next, an Aggregator
moves towards the average position of all its current supporters for each of the
three dimensions of the policy field. Thus, this so-called ‘partisan constituency
model’ (Ezrow et al, 2011) means that a party adapts to the ideological stances
of their supporters by advocating their average position. Finally, a Hunter
compares the old and new amount of votes as expressed by the opinion polls.
If its last policy shift increased electoral support, it continues in the same
Muis and Scholte
12 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
direction; otherwise, it changes heading and makes an ideological move in the
opposite direction.
To avoid unrealistically large jumps across the policy field and make sure
that parties can only ‘locally’ adapt (Kollman et al, 1992), the maximum size of
an ideological move each round is set at 0.5 (on a scale from 10 to 10).
Experiment 1: Varying issue saliency
The party sizes of the baseline simulation model resemble the real political
situation in 2006, except for the considerably larger size of D66 and smaller
sizes of the CDA and SP (see Table 2). First, we explore what would happen if
the importance of either economic or cultural issues increases. As explained, we
can expect that the smaller the ideological distance between a party and the
voters, the more a party will electorally profit from an increased importance of
that particular issue. Reversely, parties that deviate strongly from voters are
put at a disadvantage. The outcomes in Table 4 show that this indeed is the
case if we put a stronger emphasis on the parties’ stances regarding economic
issues. Stressing the positions on economic issues is harmful for the PVV (from
2.8 per cent to 1.9 per cent), which holds a relatively extreme position
(x¼6.85), whereas it is advantageous for the parties (CDA, D66, CU) that are
closest to the average voter position on economic issues (x¼0.7).
Remarkably, however, the outcomes show that an increase in saliency of the
cultural dimension leads to exactly the opposite: parties with a score around
the average (y¼1.6) shrink, whereas parties with extreme positions benefit.
Although the position of the PVV deviates enormously from the average stance
of the electorate (6.9 units), the party benefits most from emphasizing
immigration and integration issues: the amount of electoral support doubles
and increases from 3.6 to 8.6 per cent. A similar conclusion applies to the GL,
which holds the second most extreme position on immigration and integration
issues and enhances its electoral strength from 5.7 to 8.1 per cent of the vote.
For an explanation on why these two most extreme parties benefit most from
an increased importance of cultural issues, we have to shift our focus from the
positions of the voters to the positions of all other parties. The PVV and GL
hardly face any competing party nearby on the cultural dimension issue, while
at the same time they are confronted with a severe competition when we
consider the positions of their competitors on the economic dimension.
Thus, a strong deviation from the mean voter position is not necessarily
disadvantageous in itself; we have to take the whole configuration of party
positions into account.
How to find the ‘winning formula’?
13r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Table 4: Party support (in percentage vote share) while varying the agenda-setting of socio-economic (x) and cultural conflict (y) issues
(Experiment 1)
Relative increase in salience of cultural issues of: Relative increase in salience of socio-economic issues of:
25% 50% 75% 100% 25% 50% 75% 100%
CDA 18.2 (0.4) 18.2 (0.5) 18.4 (0.5) 18.8 (0.5) 19.4 (0.4) 20.3 (0.4) 20.6 (0.5) 20.7 (0.5)
CU 4.0 (0.2) 3.9 (0.2) 3.7 (0.2) 3.6 (0.2) 4.3 (0.2) 4.6 (0.2) 4.8 (0.2) 5.0 (0.2)
D66 14.8 (0.4) 13.4 (0.4) 12.2 (0.4) 11.1 (0.4) 18.8 (0.4) 20.8 (0.5) 22.2 (0.5) 23.0 (0.5)
GL 6.3 (0.2) 7.0 (0.3) 7.6 (0.3) 8.1 (0.3) 5.4 (0.2) 5.1 (0.2) 4.8 (0.2) 4.7 (0.2)
PvdA 26.8 (0.5) 25.5 (0.5) 24.7 (0.5) 24.1 (0.5) 28.8 (0.5) 28.2 (0.5) 27.7 (0.6) 27.5 (0.5)
PVV 4.9 (0.2) 6.3 (0.2) 7.8 (0.3) 8.6 (0.3) 2.8 (0.2) 2.3 (0.2) 2.0 (0.2) 1.9 (0.2)
SGP 1.9 (0.1) 1.9 (0.1) 1.9 (0.1) 1.9 (0.1) 1.8 (0.1) 1.8 (0.1) 1.9 (0.1) 1.9 (0.1)
SP 5.2 (0.2) 5.1 (0.2) 4.6 (0.2) 4.4 (0.2) 5.2 (0.2) 5.3 (0.2) 5.4 (0.2) 5.4 (0.2)
VVD 17.9 (0.4) 18.8 (0.4) 19.2 (0.4) 19.4 (0.4) 13.5 (0.4) 11.6 (0.3) 10.6 (0.3) 9.9 (0.3)
Note: Standard deviations between brackets.
Muis and Scholte
14 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Experiment 2: Varying the strategy of the PVV
In subsequent scenarios, the PVV is inclined to choose new positions in
the policy field, while other parties are Stickers. Obviously, if Wilders is
ideologically rigid as well, he would not be able to enhance his electoral
attractiveness over time. He would then consistently receive about 3.6 per cent
of the votes. The impact of ideological flexibility for the radical populist right
is rather large (see Table 5): the Aggregating PVV can generally achieve the
support from more than a third of the electorate (34.5 per cent). The results
also show that the PVV moves considerably (7.3 points) to the economic left
although it also gets more moderate on immigration policies (4.8 points). On
religious policies, the PVV becomes more secular (3 points). The strategy pulls
the PVV to an ‘electoral niche’ that consists of voters on the economic left of
the conservative liberal VVD and the cultural right of the Christian Democrats
(CDA) and Social Democrats (PvdA). The location of this fertile niche
corresponds the empirical findings of Van der Brug and Van Spanje (2009,
p. 309), who state that ‘large groups of citizens are not represented by any
parties, in particular those who are left-wing on socio-economic issues and
right-wing on cultural issues’.
In the second scenario, we consider what happens if the PVV applies the
Hunter rule. Counter-intuitively, the Aggregator strategy, which has no built-
in incentive to increase its vote share, is generally far more successful for
the PVV than the Hunter, which does have such an incentive and achieves
17 per cent. In both scenarios, the inclination of the PVV to move to the
Table 5: PVV support (in percentage vote share) and position over time while varying the PVV
party strategy (Experiment 2)
Rounds Percentage of PVV x y z
PVV=Aggregator 5 7.1 (1.3) 5.4 (0.1) 8.4 (0.3) 1.6 (0.2)
10 16.8 (3.5) 3.2 (0.3) 7.7 (0.7) 2.2 (0.5)
15 25.2 (4.1) 1.4 (0.6) 6.5 (1.1) 3.2 (0.6)
20 32.9 (4.5) 0.0 (1.0) 4.7 (1.6) 4.2 (0.9)
34.5 (4.9) 0.4 (1.3) 3.8 (1.9) 4.5 (1.2)
30 34.5 (4.9) 0.4 (1.3) 3.7 (2.1) 4.5 (1.2)
PVV=Hunter 5 3.7 (1.4) 6.7 (0.6) 8.5 (0.6) 1.7 (0.7)
10 6.3 (4.4) 6.0 (1.4) 8.2 (1.4) 2.3 (1.7)
15 9.4 (6.7) 5.1 (2.1) 7.8 (2.1) 2.8 (2.4)
20 12.1 (8.0) 4.2 (2.5) 7.2 (2.6) 3.1 (2.6)
25 14.8 (8.9) 3.4 (2.8) 6.7 (3.0) 3.4 (2.6)
30 17.2 (9.7) 2.7 (2.8) 6.2 (3.3) 3.6 (2.4)
Stable situation is reached; Standard deviations between brackets.
How to find the ‘winning formula’?
15r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
economic left is much stronger than the tendency to move to cultural right.
A Hunting PVV occupies a cultural position (y¼6.2) that remains more anti-
immigrant than the VVD. The outcomes show furthermore that in that case
the PVV moves, on average, somewhat to the economic left (x¼2.7), and
becomes less religious (z¼3.5). The PVV mainly steals votes from the VVD,
by triggering the same electoral potential on the socio-economic left and
cultural right side of the VVD.
In sum, as a result of an apparent electoral niche in the Dutch political
landscape, both adaptive scenarios lead to a steady and impressive growth of
the electoral strength of the PVV, as we have observed in reality. Although we
empirically based our party position scores on the situation in 2006 and did not
provide empirical data on changes in positions, the hypothesized shift seems
clearly in line with the actual ideological trajectory of the PVV that scholars
and journalists have observed. For instance, Vossen (2011) concludes that one
of the most import ideological changes of Wilders was the replacement of the
harsh neo-liberal tone in the initial programme by a much more leftist stance.
In contrast to previous statements, the PVV does not advocate the reduction
of minimum wage levels and social security benefits anymore. Wilders
protested against governmental plans to raise the official retirement age and
announced to look for cooperation with the trade unions.
Experiment 3: Varying all party strategies
Third, we have explored 10 different scenarios with varying strategies of the
established parties. Table 6 presents an overview of the predicted party size and
position of the PVV. Table 7 displays the results of two of these scenarios in
more detail.
The share of the PVV decreases significantly, and depends on the timing of the
response and type of strategy used by the mainstream parties. The Aggregator
appears to be the ‘dominant’ strategy here. Both the PVV and established parties
are better off with this strategy than with the Hunter or Sticker strategy,
regardless of the strategy used by the competing parties. We predict that when all
other parties would immediately have used the Aggregator strategy from 2006
onwards, the PVV would have stayed considerably smaller (8.5 per cent)
in comparison with the situation in which it faces Hunting competitors
(15.6 per cent). The three largest mainstream parties CDA, VVD and PvdA lose
the least voters to an Aggregating PVV if they copy their strategy, as they are
then rapidly pulled to the same electoral niche.
This outcome can be due to the fact that the Hunter strategy entails a lot
of risks because it uses ‘blind’ adaptation. It now and then makes a ‘mistake’
by moving in the wrong direction – when it would have been better to stay put
Muis and Scholte
16 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
or moderate one’s stance. For instance, the first move of the PVV is sometimes
towards an even stronger anti-immigrant stance than it already has, which
diminishes its electoral attractiveness.
The crucial role of timing in political competition becomes evident when we
consider the impact of Satisficing. We assumed that a party only undertakes
action after it has lost a total of more than 25 per cent of its votes compared
with the amount it initially had in 2006. When the established parties only
decide to reposition themselves after they are faced with substantial electoral
losses, the amount of support that the PVV is able to mobilize is much larger.
For instance, when Wilders adopts the most profitable strategy (Aggregator)
and the established parties immediately react when the PVV is established, the
differences in the fortunes of the PVV are 9 (instead of 16) and 16 (instead of
26) per cent for Aggregating and Hunting mainstream parties, respectively.
Thus, the simulations show that the PVV gained votes because the governing
parties were hesitating in responding. Given the actual election results in
Table 6: Party support (in percentage vote share) and position while varying party strategies
(Experiment 3). Average situation after 30 rounds
Percentage of
CDA 14.4 (2.6) 0.3 (1.0) 2.4 (1.4) 1.0 (0.3)
CU 4.6 (0.7) 1.7 (0.1) 1.3 (0.0) 7.3 (0.1)
D66 14.7 (1.5) 0.7 (0.7) 1.8 (0.8) 7.5 (0.8)
GL 7.2 (1.0) 5.3 (0.1) 4.8 (0.2) 6.5 (0.1)
PvdA 22.0 (1.7) 1.4 (0.5) 0.1 (1.2) 3.8 (0.6)
PVV 16.4 (1.3) 1.4 (1.1) 6.1 (1.1) 4.6 (1.1)
SGP 2.1 (0.8) 2.7 (0.0) 3.2 (0.0) 9.6 (0.0)
SP 5.2 (2.8) 7.5 (0.4) 0.7 (1.2) 5.7 (1.1)
VVD 13.5 (1.1) 3.8 (1.1) 4.2 (0.7) 5.6 (0.8)
CDA 12.5 (5.3) 0.3 (2.1) 2.5 (2.4) 0.7 (2.1)
CU 8.5 (4.0) 1.2 (2.5) 1.7 (2.9) 2.5 (2.7)
D66 14.0 (6.4) 0.1 (2.2) 1.1 (2.6) 5.6 (1.7)
GL 9.4 (5.2) 3.5 (2.7) 3.2 (2.8) 5.8 (1.9)
PvdA 13.8 (5.9) 1.8 (2.3) 0.7 (2.5) 4.9 (1.6)
PVV 15.6 (6.3) 0.6 (2.5) 5.2 (2.7) 5.2 (1.8)
SGP 6.5 (3.2) 1.2 (3.0) 3.2 (3.1) 4.4 (2.6)
SP 8.4 (5.2) 5.1 (2.9) 0.2 (3.0) 5.3 (1.9)
VVD 11.3 (5.9) 3.4 (2.9) 4.3 (2.6) 4.5 (1.8)
Notes: Satisficing means: not losing more than 25 per cent of start amount at the first round (see
Table 4) is acceptable. Standard deviations between brackets.
How to find the ‘winning formula’?
17r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Table 7: Overview of PVV position and party support (in percentage vote share) in different scenarios. Average situation after 30 rounds
Experiment Agenda-
setting: x:y
Strategy other
PVV in
1 1:1 Sticker Stickers 3.6 (0.2) 6.9 8.5 1.5
1:2 Sticker Stickers 1.9 (0.2) 6.9 8.5 1.5
2:1 Sticker Stickers 8.6 (0.3) 6.9 8.5 1.5
2 1:1 Aggregator Stickers 34.5 (4.9) 0.4 (1.3) 3.7 (2.1) 4.5 (1.2)
1:1 Hunter Stickers 17.2 (9.7) 2.7 (2.8) 6.2 (3.3) 3.6 (2.4)
3 1:1 Aggregator Aggregator 8.5 (1.6) 3.4 (0.9) 5.4 (1.7) 1.8 (0.7)
1:1 Aggregator Satisficing Aggregators 16.4 (1.3) 1.4 (1.1) 6.1 (1.1) 4.6 (1.1)
1:1 Aggregator Hunter 15.6 (6.3) 0.6 (2.5) 5.2 (2.7) 5.2 (1.8)
1:1 Aggregator Satisficing Hunters 26.4 (6.5) 0.5 (1.6) 4.2 (2.3) 4.7 (1.3)
1:1 Hunter Aggregator 7.9 (4.7) 3.2 (3.2) 5.9 (2.8) 2.9 (2.0)
1:1 Hunter Satisficing Aggregators 13.4 (5.0) 2.0 (2.3) 6.7 (2.6) 3.3 (2.0)
1:1 Hunter Hunters 15.6 (6.3) 0.6 (2.5) 5.2 (2.7) 5.2 (1.8)
1:1 Hunter Satisficing Hunters 16.4 (9.4) 2.5 (2.8) 6.2 (3.2) 3.5 (2.4)
1:1 Sticker Aggregator 3.0 (0.8) 6.9 8.5 1.5
1:1 Sticker Hunter 3.9 (2.8) 6.9 8.5 1.5
4 1.5:1 Aggregator Sticker 34.0 (3.9) 0.1 (1.3) 3.6 (2.0) 4.3 (1.1)
2:1 Aggregator Sticker 32.5 (3.6) 0.4 (1.3) 3.6 (1.9) 4.3 (1.0)
1:1.5 Hunter Sticker 19.4 (10.7) 2.6 (3.0) 6.2 (2.8) 3.5 (2.4)
1:2 Hunter Sticker 19.3 (10.8) 3.0 (3.1) 6.1 (2.7) 3.4 (2.3)
Notes: Satisficing means: not losing more than 25 per cent of the start amount at the first round (see Table 4) is acceptable. Standard deviations
between brackets.
Muis and Scholte
18 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
2010 (the PVV received 15.4 per cent), our outcomes support the claim
that mainstream parties aim to represent the stances of their supporters, but
because they only become active when they clearly have lost votes, the adaptive
populist newcomer could considerably boost its electoral appeal over time.
An explanation for the fact that established parties nevertheless display inertia
is that it is the easiest response: it does not require innovation and does not
provoke, at least at first, internal disunity (Bale et al, 2010).
At the same time, however, the newcomer is only able to become a serious
challenge to the established parties when it adopts ideological flexibility.
If we in contrast assume that the PVV is ideologically rigid, it would be
inexplicable how the party could have grown so steadily since its foundation.
Wilders cannot enlarge his vote share in any scenario in which he adopts
a Sticker strategy. It then hardly matters what all other political parties do:
the vote share of the PVV does not significantly change over time and remains
between 3 and 4 per cent.
Experiment 4: Varying issue saliency and the strategy of the PVV
Finally, we explore assumptions about the party strategy of the PVV and issue
saliency simultaneously (see the last four rows in Table 7). The outcomes thus far
have shown that both increasing the importance of immigration and integration
topics and adaptive strategies generate a surge of the PVV. The combination
of both saliency and strategy, however, does not lead to extra successes.
Nevertheless, an adaptive PVV can prevent a decrease in support when
voters attach more significance to socio-economic issues and consider
immigration and integration issues relatively less important (see Table 8). As
Aggregator or Hunter, the PVV achieves roughly similar outcomes (32.5 and
19.3 per cent, respectively) compared with the scenario in which voters attach
equal importance to the cultural conflict and socio-economic dimension (see
Experiment 2: the PVV achieved 34.5 and 17.2 per cent). Thus, in contrast
to an agenda-setting argument that suggests a straightforward positive
association between immigration news and amount of support for Wilders,
our simulations suggest that populist leaders are not necessarily put at a
disadvantage in case of diminishing mass media publicity on immigration
issues, if they strategically reposition themselves.
Conclusion and Discussion
The innovative contribution of this article is that it maps out the mechanisms
that underlie the long-cherished theories on the impact of party strategies for
How to find the ‘winning formula’?
19r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Table 8: PVV support and position while varying both agenda-setting and party strategy (Experiment 4)
Rounds Percentage
of PVV
x y z Percentage
of PVV
xy z
PVV=Aggregator Socio-economic issue agenda-setting: x=1.5 Socio-economic issue agenda-setting: x=2
5 5.2 (0.6) 5.7 (0.4) 8.0 (0.4) 1.2 (0.4) 3.8 (0.5) 6.0 (0.5) 7.8 (0.4) 0.9 (0.4)
10 14.8 (2.1) 3.9 (0.8) 7.0 (1.0) 2.1 (1.1) 11.0 (2.1) 4.5 (0.9) 6.5 (1.0) 1.6 (1.30)
15 25.4 (2.1) 2.2 (1.1) 5.7 (1.4) 3.2 (1.2) 20.5 (2.4) 3.1 (1.2) 5.1 (1.4) 2.3 (1.6)
20 32.3 (2.9) 0.6 (1.4) 4.1 (1.7) 4.1 (1.2) 29.6 (3.1) 1.4 (1.6) 3.9 (1.6) 4.1 (1.3)
25 34.0 (3.9) 0.1 (1.3) 3.6 (1.9) 4.3 (1.1) 32.3 (3.2) 0.5 (1.4) 3.6 (1.8) 4.3 (1.0)
30 34.0 (3.9) 0.1 (1.3) 3.6 (2.0) 4.3 (1.1) 32.5 (3.6) 0.4 (1.3) 3.6 (1.9) 4.3 (1.0)
PVV=Hunter Cultural conflict issue agenda-setting: y=1.5 Cultural conflict issue agenda-setting: y=2
5 6.5 (2.5) 6.7 (0.6) 8.5 (0.7) 1.7 (0.7) 8.6 (3.2) 6.8 (0.6) 8.4 (0.7) 1.7 (0.7)
10 8.9 (5.3) 6.0 (1.4) 8.2 (1.4) 2.3 (1.6) 10.2 (5.2) 6.1 (1.4) 8.1 (1.3) 2.3 (1.5)
15 11.8 (8.0) 5.0 (2.1) 7.7 (1.9) 2.8 (2.3) 13.0 (8.0) 5.3 (2.1) 7.6 (1.8) 2.7 (2.0)
20 14.5 (9.2) 4.1 (2.6) 7.2 (2.4) 3.1 (2.5) 14.7 (9.2) 4.4 (2.7) 7.1 (2.1) 3.0 (2.3)
25 17.1 (10.1) 3.2 (2.9) 6.7 (2.7) 3.3 (2.4) 17.0 (10.1) 3.7 (3.0) 6.6 (2.4) 3.3 (2.4)
30 19.4 (10.7) 2.6 (3.0) 6.2 (2.8) 3.5 (2.4) 19.3 (10.8) 3.0 (3.1) 6.1 (2.7) 3.4 (2.3)
Note: Standard deviations between brackets.
Muis and Scholte
20 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
both established parties and populist challengers. We have argued that ABM
provides a fruitful tool in this endeavour. It improves our theoretical
framework for explaining the rise of radical right contenders because the
implications of adaptive actions and reactions of all actors can be system-
atically formulated. It can thus generate the aggregate outcomes (in this case:
party sizes and party positions) that emerge from the complex dynamic
interactions inherent electoral competition.
To illustrate our general argument, we took a closer look at the surge of
the Dutch anti-immigration party PVV, which has spectacularly increased its
electoral popularity since its foundation. In sum, our simulation experiments
give two major insights.
First, our outcomes show that the PVV does well when voters attach more
weight to cultural conflict topics, rather than the socio-economic or religious
divide. We assumed that agenda-setting shapes the parties’ fortunes as follows:
when immigration and integration policies are at the top of the media agenda,
voters judge parties more strongly by their positions on these issues. Likewise,
when the economy is prominent in the news, voters predominantly evaluate
parties by their positions on the socio-economic dimension. Our outcomes
explain why it is understandable that Geert Wilders tends to publicly stress
his views on citizenship, ethnic relations and immigration, and attack the
viewpoints of other parties on particularly these issues: given the Dutch
configuration of policy positions, he electorally profits when voters attach
more importance to the policy positions of political parties on these issues
and neglect others (cf. Kleinnijenhuis and Krouwel, 2007, p. 30). Interestingly,
the policy positions are already sufficient to explain the fact that popular
support for Wilders significantly increases when the debate in the mass media
focuses more on Islam and immigration; it is unnecessary to supplement this
with issue ownership notions.
Second, the simulation outcomes support our argument that populist radical
right parties, or any party for that matter, should not be viewed as passive
variables in order to understand its fortunes. In line with the hypothesis of
Ignazi (2003), we showed that the difference between a marginal and successful
radical right party is its strategic flexibility to exploit whatever favourable
circumstances are available. In our simulation scenarios, the party strategy of
the PVV has a much stronger impact on popularity than issue saliency. An
increase in prominence of socio-economic issues is not necessarily harmful for
the PVV at all – under the condition that Wilders adapts and moves to a more
moderate socio-economic position compared with his initial outspoken liberal
pro-market stance. In 2006, the PVV was positioned right socio-economically,
which is relatively disadvantageous when voters deem the parties’ view on
socio-economic policies more important. Both adaptive learning by vote-
maximizing and by copying the stances of one’s supporters lead to a move
How to find the ‘winning formula’?
21r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
towards a moderate leftist socio-economic policy position, thus moving away
from the party’s ‘Achilles-heel’. This policy shift contradicts Kitschelt’s (1995)
original claim that the ‘winning formula’ of the radical right is the combination
of a liberal pro-market position on socio-economic policies with a culturally
exclusionist position.
Our central focus on adaptive learning fills a remarkable gap in the literature
on the populist radical right. Surprisingly, although policy positions and
ideological distances are central concepts in theories of party competition,
scholars seldom identify the mechanisms by which successful populist leaders
are sometimes apparently able to find successful positions in the policy space,
whereas many other attempts fail to do so. We have focused on policy position
shifts, but it should be noted that political parties, especially governing parties,
influence the media agenda and thus shape fluctuations in issue salience
(Brandenburg, 2002). Future models could also include the deliberate, tactical
moves of political actors with regard their distribution of emphasis on different
issues instead of their position shifts on these issues only. This article is only
a modest start, but we believe that the application of agent-based simulations
can be very fruitful, as they can provide realistic accounts of how party leaders
can improve their electoral attractiveness by learning from feedback.
About the Authors
Jasper Muis is an Assistant Professor at the Sociology Department of the Vrije
Universiteit Amsterdam.
Michel Scholte is a Research Master student in Sociology at the Graduate
School of Social Sciences at the Vrije Universiteit Amsterdam.
1 We used Netlogo. The syntax is available through the authors on request.
2 We only had to exclude the PvdD because the 2006 Chapel Hill Survey did not include this party.
We also left out all other parties that did not gain seats in 2006 (for example, EenNL). One could
include these parties, but additional non-policy related parameters would be necessary to account
for their marginality, like a bias of voters (Adams, 2001) or a lack of media attention (Muis,
3 The party position scores we used correlate strongly with the expert scores collected by Benoit
and Laver (2006) (Pearson’s r¼0.97; n¼21). Hereby, we used the three items taxes versus
spending, immigration and social. Obviously, Benoit and Laver did not measure the position of
PVV, as the party did not yet exist at the time of their expert survey. Therefore, we decided to
rely on the Chapel Hill data.
Muis and Scholte
22 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
4 Every run is done with a slightly different sample of citizens: on each dimension, the individual
voter’s positions are randomly chosen from a normal distribution with each voter’s score as mean
and standard deviation of 1.
5 Change heading is done by making a random turn between 90 and 270 degrees.
6 All generated outcomes are the averages of 1000 model runs.
Aarts, K. and Thomassen, J. (2008) Dutch voters and the changing party space 1989–2006. Acta
Politica 43(2/3): 203–234.
Adams, J. (2001) A theory of spatial competition with biased voters: Party policies viewed
temporally and comparatively. British Journal of Political Science 31(1): 121–158.
Adams, J., Clark, M., Ezrow, L and Glasgow, G. (2006) Are niche parties fundamentally different
from mainstream parties? The causes and the electoral consequences of Western European
parties’ policy shifts, 1976–1998. American Journal of Political Science 50(3): 513–529.
Adams, J., Ezrow, L. and Somer-Topcu, Z. (2011) Is anybody listening? Evidence that voters do
not respond to European parties’ policy statements during elections. American Journal of
Political Science 55(2): 370–382.
Albertazzi, D. and McDonnell, D. (2008) Introduction: The sceptre and the spectre. In:
D. Albertazzi and D. McDonnell (eds.) Twenty-First Century Populism. The Spectre of Western
European Democracy. Houndmills, UK: Palgrave McMillan, pp. 1–11.
Axelrod, R. (1997) The Complexity of Cooperation. Agent-Based Models of Competition and
Collaboration. Princeton, NJ: Princeton University Press.
Axelrod, R. and Cohen, M.D. (2000) Harnessing Complexity. Organizational Implications of a
Scientific Frontier. New York: The Free Press.
Bale, T., Green-Pedersen, C., Krouwel, A., Luther, K.R. and Sitter, K.R. (2010) If you can’t beat
them, join them? Explaining social democratic responses to the challenge from the populist
radical right in Western Europe. Political Studies 58(3): 410–426.
Bendor, J., Diermeier, D., Siegel, D. and Ting, M. (2011) A Behavioral Theory of Elections.
Princeton, NJ: Princeton University Press.
Benoit, K. and Laver, M. (2006) Party Policy in Modern Democracies. London: Routledge.
Boero, R. and Squazzoni, F. (2005) Does empirical embeddedness matter? Methodological issues
on agent-based models for analytical social science. Journal of Artificial Societies and Social
Simulation 8(4): 1–31.
Brandenburg, H. (2002) Who follows whom? The impact of parties on media agenda formation in
the 1997 British general elections campaign. Harvard Journal of Press and Politics 7(3): 34–54.
Budge, I. and Farlie, D.J. (1983) Explaining and Predicting Elections. Issue Effects and Party
Strategies in Twenty-Three Democracies. London: Allen & Unwin.
De Lange, S.L. and Art, D. (2011) Fortuyn versus Wilders: An agency-based approach to radical
right party building. West European Politics 34(6): 1229–1249.
Epstein, J. (1999) Agent-based computational models and generative social science. Complexity
4(5): 41–60.
Evans, J. (2005) The dynamics in social change in radical right-wing populist party support.
Comparative European Politics 3(1): 76–101.
Ezrow, L., De Vries, C., Steenbergen, M. and Edwards, E. (2011) Mean voter representation and
partisan constituency representation. Do parties respond to the mean voter position of to their
supporters? Party Politics 17(3): 275–301.
How to find the ‘winning formula’?
23r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Goodwin, M. (2006) The rise and faults of the internalist perspective in extreme right studies.
Representation 42(4): 347–364.
Hooghe, L. et al (2008) Reliability and validity of measuring party positions: The Chapel Hill
expert surveys of 2002 and 2006. Unpublished manuscript.
Ignazi, P. (2003) Extreme Right Parties in Western Europe. Oxford: Oxford University Press.
Ivarsflaten, E. (2008) What unites right-wing populists in Western Europe? Re-examining grievance
mobilization models in seven successful cases. Comparative Political Studies 41(1): 3–23.
Kitschelt, H. (1995) The Radical Right in Western Europe. Ann Arbor, MI: University of Michigan.
Kitschelt, H. (2007) Growth and persistence of the radical right in postindustrial democracies:
Advances and challenges in comparative research. West European Politics 30(5): 1176–1206.
Kleinnijenhuis, J. and Krouwel, A. (2007) The nature and influence of party profiling websites.
Paper presented at the Politicologenetmaal 2007, 31 May–1 June, Antwerpen, Belgium.
Kleinnijenhuis, J. and Pennings, P. (2001) Measurement of party positions on the basis of party
programmes, media coverage and voter perceptions. In: M. Laver (ed.) Estimating the Policy
Positions of Political Actors. London: Routledge, pp. 162–182.
Kleinnijenhuis, J. et al (2007) Nederland vijfstromenland. De rol van de media en de stemwijzers bij de
verkiezingen van 2006. Amsterdam, the Netherlands: Bert Bakker.
Kolk, H., Aarts, K., Rosema, M. and Brinkman, M. (2007) Nationaal Kiezersonderzoek 2006 NKO
2006 [data file]. Den Haag, the Netherlands: Centraal Bureau voor de Statistiek.
Kollman, K., Miller, J. and Page, S. (1992) Adaptive parties in spatial elections. American Political
Science Review 86(4): 929–937.
Koopmans, R., Statham, P., Giugni, M. and Passy, F. (2005) Contested Citizenship. Immigration
and Cultural Diversity in Europe. Minneapolis, MN: University of Minnesota Press.
Kriesi, H., Grande, E., Lachat, R., Dolezal, M., Bornschier, S. and Frey, T. (2008) West European
Politics in the Age of Globalization. Cambridge, UK: Cambridge University Press.
Laver, M. (2005) Policy and the dynamics of political competition. American Political Science
Review 99(2): 263–281.
Laver, M. and Sergenti, E. (2011) Party Competition: An Agent Based Model. New York: Princeton
University Press.
Mazzoleni, G. (2003) The media and the growth of neo-populism in contemporary democracies. In:
G. Mazzoleni, J. Stewart and B. Horsfield (eds.) The Media and Neo-Populism. A Comparative
Analysis. Westport, CT: Praeger.
McCombs, M.E. and Shaw, D. (1972) The agenda-setting function of the mass media. Public
Opinion Quarterly 69(4): 813–824.
McDonald, M., Budge, I. and Pennings, P. (2004) Choice versus sensitivity: Party reactions to
public concerns. European Journal of Political Research 43(6): 845–868.
Meguid, B. (2008) Party Competition between Unequals: Strategies and Electoral Fortunes in
Western Europe. Cambridge, UK: Cambridge University Press.
Meuleman, B., Davidov, E and Billiet, J. (2009) Changing attitudes toward immigration in Europe,
2002–2007: A dynamic group conflict theory approach. Social Science Research 38(4): 352–365.
Miller, J. and Page, S. (2007) Complex Adaptive Systems. An Introduction to Computational Models
of Social Life. Princeton, NJ: Princeton University Press.
Mudde, C. (2004) The populist Zeitgeist. Government and Opposition 39(4): 542–563.
Mudde, C. (2007) Populist Radical Right Parties in Europe. Cambridge, UK: Cambridge University
Muis, J. (2010) A simulation of political stability and change in the Netherlands (1998–2002).
Empirically testing agent-based models of political party competition with media effects. The
Journal of Artificial Societies and Social Simulation 13(2): 4.
Norris, P. (2005) Radical Right: Voters and Parties in the Electoral Market. New York: Cambridge
University Press.
Muis and Scholte
24 r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
Pellikaan, H., De Lange, S. and Van der Meer, T. (2007) Fortuyn’s legacy: Party system change in
the Netherlands. Comparative European Politics 5(3): 282–302.
Rogers, E.M., Dearing, J.M. and Bregman, D. (1993) The anatomy of agenda-setting research.
Journal of Communication 3(2): 68–84.
Rydgren, J. (2007) The sociology of the radical right. Annual Review of Sociology 33(1): 241–262.
Simon, H. (1955) A behavioral model of rational choice. Quarterly Journal of Economics 69(1):
Synovate. (2010) Archief Politieke Barometer. Amsterdam: Synovate,
content.asp?targetid=686, accessed 1 May 2010.
Tavits, M. (2008) Policy positions, issue importance, and party competition in new democracies.
Comparative Political Studies 41(1): 48–72.
Van der Brug, W. (2004) Issue ownership and party choice. Electoral Studies 23(2): 209–233.
Van der Brug, W. and Fennema, M. (2007) What causes people to vote for a radical right party?
A review of recent work. International Journal of Public Opinion Research 19(4): 474–487.
Van der Brug, W., Fennema, M. and Tillie, J. (2005) Why some anti-immigrant parties fail and
others succeed. A two-step model of aggregate electoral support. Comparative Political Studies
38(5): 537–573.
Van der Brug, W. and Van Spanje, J. (2009) Immigration, Europe, and the ‘new’ cultural
dimension. European Journal of Political Research 48(3): 309–334.
Van Kersbergen, K. and Krouwel, A. (2008) A double-edged sword! The Dutch centre-right and
the ‘foreigners issue’. Journal of European Public Policy 15(2): 398–414.
Vossen, K. (2011) Vom konservativen Liberalen zum Nationalpopulisten. Die ideologische
Entwicklung des Geert Wilders. In: F. Hartleb and F. Wielenga (eds.) Populismus in den
Niederlanden und Deutschland im Vergleich.Mu
¨nster, Germany: Waxmann.
How to find the ‘winning formula’?
25r2012 Macmillan Publishers Ltd. 0001-6810 Acta Politica 1–25
... Moreover, he extended the previous models by including the role of the media. The first study to apply ABM in research on party populism [10] explored how populist radical right parties position themselves in a political space to find their "winning formula". The model includes the importance voters attribute to various issues and differentiates between a limited number of party strategies, mostly following [7]. ...
... To do so, they may apply different strategies to adapt their positions on policy issues. We implemented the four strategies outlined by [10]: ...
In this paper we report on the development of an agent-based model (ABM) simulating the behaviour of voters and the positioning of political parties in Austria. The aim is to create what-if scenarios taking into account contextual changes, such as political crises as well as changes in parties’ policy positions and voters’ attitudes. Drawing on data from the Austrian National Election Study (AUTNES) and the Chapel Hill Expert Survey (CHES), we are able to map both demand- and supply-side characteristics. We present first results of the simulation analysis of applied strategies of voters and parties. This way, we are able to create first what-if scenarios that show how results of elections would change, if voters applied different strategies when deciding which party to vote for. In developing a simulation for the case of Austria as a reference model, we lay the foundation for more universal applications of ABM in political science.KeywordsVoting behaviourParty competitionAgent-based simulation
... The role of the media is particularly salient in disseminating the emotionalised message of right-wing populists (Mazzoleni 2008;Muis and Scholte 2013;Krämer 2014). As Wahl-Jorgensen (2019, 9) observes, the articulation of emotion in news media contributes to an emergence of shared repertoires of emotion, which serve to both establish and challenge emotional regimes. ...
Full-text available
This paper proposes that the emotion of shame is key to understanding the appeal of Poland’s ruling populist right-wing party, Law and Justice (Prawo i Sprawiedliwość, or PiS). PiS employs shame in its strategy of emotion regulation in mediated party communication in pro-government media outlets. We suggest that there are two pillars of shame that underpin support for PiS: (1) the collective shame that originates from the perceived cultural inferiority of Poland in relation to the West/Europe, and (2) the individual shame of failing to achieve material prosperity in the context of the post-communist economic transformation. We examine how shame and pride have been instrumentalised in the coverage of the 2019 European Parliament elections by the right-wing media outlet The paper demonstrates how consistently invoked both economic and cultural shame and highlighted the antagonism between PiS and the opposition by identifying the latter with a ‘pedagogy of shame.’
... Simulating generative processes is not new and is, for example, applied in the Siena algorithm (Snijders, 2017b) that uses stochastic simulations to find probability functions underlying network dynamics. The use of ABM for simulations with empirical input-and output reference, however, is less conventional, except for some notable models explaining the success of political parties (Muis, 2010;Muis & Scholte, 2013). ...
Linkage analyses use data from panel surveys and content analyses to assess media effects under field conditions and are able to close the gap between experimental and survey-based media effects research. Results from current studies and simulations indicate, however, that these studies systematically under-estimate real media effects as they aggregate measurement errors and reduce the complexity of media content. In response to these issues, we propose a new method for linkage analysis which applies agent-based simulations to directly assess short-term media effects using empirical data as guideposts. Results from an example study modeling opinion dynamics in the run-up of a Swiss referendum show that this method outperforms traditional regression-based linkage analyses in detail and explanatory power. In spite of the time-consuming modeling and computation process, this approach is a promising tool to study individual media effects under field conditions.
... Muis implemented an ABM of party competition with media effects for the Dutch case by using empirical data on party positions, public opinion polls, and election outcomes from May 1998 to May 2002 for calibration. In a follow-up study, Muis and Scholte (2013) were able to reproduce the rise of the Dutch Party for Freedom with the model. Based on this work, the researchers observed that the key mechanism for populists winning ground was their flexibility in finding winning positions, while established parties often stuck to their ideological stances. ...
Full-text available
In various branches of the social sciences, agent-based models (ABMs) have long been applied to enhance researchers’ understanding of complex systems and processes. However, in communication science, this approach is rarely used. In this article, we argue that ABMs have the potential to advance communication research in general, and computational communication science (CCS) in particular, by helping scholars address two major gaps. First, by generating emergent global phenomena from individual interactions, ABMs make it possible to explicitly link micro and macro perspectives in communication research. Second, by formalizing theories, ABMs offer mechanism-based explanations for observed empirical patterns in data. To familiarize more communication scholars with this approach, this article provides a systematic overview of the potentials, applications, and challenges of ABMs in communication science. Special attention is paid to the criteria of reliability and validity. Full text available at:
Many simulation models of social influence are for the theoretical exploration of the outcomes resulting from certain mechanisms. They therefore tend to be relatively focussed on one mechanism at a time—the KISS approach. Here we take a more KIDS approach, looking at the interaction of two mechanisms within an evidence-led simulation of political behaviour in Austria 2013–2017. In this simulation there is not only the mutual social influence of attitudes (within a 7D space), but this social influence is constrained by a social network. However, one can also allow this social network to adapt based on the interactions between agents, so the social attitudes and social networks co-evolve. In this model, we find that (a) whether the social network is allowed to adapt is more important to the outcomes than the particular kind of social network it is initialized with, but also that (b) (given all the other mechanisms, parameters and structures in this model) a changing social network seems essential to getting outcomes that are qualitatively similar to the patterns in the observed polling data.
Simulation tools aimed at enhancing cross-sectoral cooperation can support the transition from a traditional transport planning approach based on predictions towards more integrated and participatory urban mobility planning. This shift entails a broader appraisal of urban dynamics and transformations in the policy framework, capitalizing on new developments in urban modelling. In this paper, we argue that participatory social simulation can be used to solve these emerging challenges in mobility planning. We identify the functionalities that such a tool should have when supporting integrated mobility planning. Drawing on a transdisciplinary case study situated in Potsdam, Germany, we address through interviews and workshops stakeholders’ needs and expectations and present the requirements of an actionable tool for practitioners. As a result, we present three main challenges for participatory, simulation-based transport planning, including: (1) enhancement of the visioning process by testing stakeholders’ ideas under different scenarios and conditions to visualise complex urban relationships; (2) promotion of collective exchange as means to support stakeholder communication; and (3) credibility increase by early stakeholder engagement with model development. We discuss how our participatory modelling approach helps us to better understand the gaps in the knowledge of the planning process and present the coming steps of the project.KeywordsSustainable urban mobility planningIntegrated planning approachParticipatory modelling
Combining Games and Agent-Based Models (ABMs) in a single research design (i.e. GAM design) shows potential for investigating complex past, present, or future social phenomena. Games offer engaging environments that can help generating insights into social dynamics, perceptions, and behaviours, while ABMs support the representation and analysis of complexity. We present here the first attempt to “discipline” the interdisciplinary endeavour of developing a GAM design in which an ABM is transformed into a game, thus the two becoming intertwined in one application. When doing this, we use as a GAM design exemplar the process of developing Quantum Leaper, a proof-of-concept video game made in Unity software and based on the NetLogo implementation of the well known “Artificial Anasazi” ABM. This study aims to consolidate the methodology component of the GAM field by proposing the GAM Reflection Framework, a tool that can be used by GAM practitioners, ABM modellers, or game designers looking for methodological guidance with developing an agent-based model that is a game (i.e. an agent-based game).
Scarcity of on-street parking in cities centers is a known factor motivating drivers to drive slowly (“to cruise”) while searching for an available parking place and is associated with negative externalities e.g., congestion, accidents, fuel waste and air pollution. Finding the correct prices is suggested to bring cruising to a sustainable level. However, current research methods based on surveys and simulations fail to provide a full understanding of drivers’ cruising preference and their behavioral response to price changes. We used the PARKGAME serious game, which provides a real-world abstraction of the dynamic cruising experience. Eighty-three players participated in an experiment under two pricing scenarios. Pricing was spatially designed as “price rings”, decreasing when receding from the desired destination point. Based on the data, we analyzed search time, parking distance, parking location choice and spatial searching patterns. We show that such a pricing policy may substantially reduce the cruising problem, motivating drivers to park earlier—further away from the destination or in the lot, especially when occupancy levels are extremely high. We further discuss the policy implications of these findings.KeywordsSerious gamesCruisingDriver behaviorParking search
Full-text available
This article analyzes the appeal of populist radical right (PRR) politics in the US after the election of Donald Trump. Specifically, I seek to explain how new media helps politicians representing the PRR secure support in Republican primaries. Using an online survey of 1052 Arizona Republicans in the lead-up to the August 2018 Senate primary, I evaluate support for three candidates: Rep. Martha McSally, former Maricopa County Sheriff Joe Arpaio, and Kelli Ward, a physician. The findings highlight a bifurcation in the drivers for support of PRR candidacies: Skepticism of immigration drives the Arpaio vote, while use of social media news and belief in party convergence mobilize Ward’s support. The results demonstrate that support for PRR politicians in the Arizona primary is concentrated in two groups, anti-immigrant and anti-establishment, and that the anti-establishment voters are more likely to access news on social media. These findings indicate that social media news consumption does shape voter perceptions about mainstream parties favorably for the PRR.
Cambridge Core - Political Sociology - Cultural Backlash - by Pippa Norris
Full-text available
Agent-based models of political party competition in a multidimensional policy space have been developed in order to reflect adaptive learning by party leaders with very limited information feedback. The key assumption is that two categories of actors continually make decisions: voters choose which party to support and party leaders offer citizens a certain policy package. After reviewing the arguments for using agent-based models, I elaborate two ways forward in the development of these models for political party competition. Firstly, theoretical progress is made in this article by taking the role of the mass media into account. In previous work it is implicitly assumed that all parties are equally visible for citizens, whereas I will start from the more realistic assumption that there is also competition for attention in the public sphere. With this addition, it is possible to address the question why new parties are seldom able to successfully compete with political actors already within the political system. Secondly, I argue that, if we really want to learn useful lessons from simulations, we should seek to empirically falsify models by confronting outcomes with real data. So far, most of the agent-based models of party competition have been an exclusively theoretical exercise. Therefore, I evaluate the empirical relevance of different simulations of Dutch party competition in the period from May 1998 until May 2002. Using independent data on party positions, I measure the extent to which simulations generate mean party sizes that resemble public opinion polls. The results demonstrate that it is feasible and realistic to simulate party competition in the Netherlands with agent-based models, even when a rather unstable period is investigated.
Why do some political parties flourish, while others flounder? In this book, Meguid examines variation in the electoral trajectories of the new set of single-issue parties: green, radical right, and ethnoterritorial parties. Instead of being dictated by electoral institutions or the socioeconomic climate, as the dominant theories contend, the fortunes of these niche parties, she argues, are shaped by the strategic responses of mainstream parties. She advances a new theory of party competition in which mainstream parties facing unequal competitors have access to a wider and more effective set of strategies than posited by standard spatial models. Combining statistical analyzes with in-depth case studies from Western Europe, the book explores how and why established parties undermine niche parties or turn them into weapons against their mainstream party opponents. This study of competition between unequals thus provides broader insights into the nature and outcome of competition between political equals.
The extreme right has consolidated its presence across Western Europe. This book presents a compilation of studies on the ideological meanings and political/partisan expressions of the extreme right, their post-war evolution, and the reasons behind the success and failure of various parties. It highlights the rise of a new type of parties that are anti-system rather than neo-fascist.
The core puzzle which this book resolves is to explain why radical right parties have advanced in a diverse array of democracies--including Austria, Canada, Norway, France, Italy, New Zealand, Switzerland, Israel, Romania, Russia, and Chile--while failing to make comparable gains in similar societies elsewhere, such as Sweden, Britain, and the United States. This book expands our understanding of support for radical right parties by presenting an integrated new theory which is then tested systematically using a wealth of cross-national survey evidence covering almost forty countries.
This paper proposes a model that takes the dynamic agent-based analysis of policy-driven party competition into a multiparty environment. In this, voters continually review party support and switch parties to increase their expectations; parties continually readapt policy positions to the shifting affiliations of voters. Different algorithms for party adaptation are explored, including "Aggregator" (adapt party policy to the ideal policy positions of party supporters), Hunter (repeat policy moves that were rewarded; otherwise make random moves), Predator (move party policy toward the policy position of the largest party), and "Sticker" (never change party policy). Strong trends in the behavior of parties using different methods of adaptation are explored. The model is then applied in a series of experiments to the dynamics of a real party system, described in a published opinion poll time series. This paper reports first steps toward endogenizing key features of the process, including the birth and death of parties, internal party decision rules, and voter ideal points.
Public relations practitioners have long attempted to persuade publics through mass media campaigns. Social scientists, likewise, have searched for the most persuasive mix of messages and media. Most of this research, however, indicates that the media seldom change attitudes and behavior, and that persuasion is, therefore, an unrealistic objective for a media campaign. Led by Maxwell McCombs, the author of this article, communication researchers have discovered a logical and more realistic effect of media: While the media do not tell people what to think, they tell people what to think about. That is, the media determine which issues—and which organizations—will be put on the public agenda for discussion. In this article, McCombs summarizes research on agenda-setting and then discusses its implications for public relations.