Content uploaded by Ayse Guveli
Author content
All content in this area was uploaded by Ayse Guveli on Jul 23, 2015
Content may be subject to copyright.
Journal of Elections, Public Opinion and Parties
Vol. 17, No. 1, 21–47, February 2007
ISSN 1745-7289 Print/1745-7297 Online/07/010021-27 © 2007 Elections, Public Opinion & Parties
DOI: 10.1080/13689880601132505
Voting LPF: Stratification and the
Varying Importance of Attitudes
MARCEL LUBBERS
*
& AYSE GÜVELI
**
*Radboud University Nijmegen, **Department of Sociology, University of Essex
Taylor and Francis LtdFBEP_A_213181.sgm10.1080/13689880601132505Journal of Elections, Public Opinion and Parties1745-7289 (print)/1745-7297 (online)Original Article2007Taylor & Francis
171000000February 2007Dr. MarcelLubbersM.Lubbers@maw.ru.nl
A
BSTRACT
Large differences exist between socio-cultural specialists and technocrats in the
extent they voted for the List Pim Fortuyn (LPF), showing us that support for the party was
aligned by social class. Distinguishing between the two specialist types provided the opportu-
nity to test Kitschelt’s hypothesis on the importance of communicative experiences and
capabilities in the labour market and the importance of cultural and economic ethnic threats.
Cultural ethnic threat explains social class differences in LPF voting better than economic
ethnic threat, though we cannot claim that the latter is of no relevance. We show this using
the Dutch sample of the European Social Survey (
n
= 2,260). Moreover, we show that socio-
political attitudes that affect voting for the LPF do so to a much smaller extent among lower-
educated people and non-socio-cultural specialists. The interactions between socio-political
attitudes and education and social class are significant.
Introduction
In the Dutch election campaign of 2002, the political establishment was woken up
by a newcomer: Pim Fortuyn. With his messages over immigrants and “the arrogant
politics” of the government and established parties, he was either loved or hated. In
the municipal elections earlier that year, an unexpected result in Rotterdam made
other politicians fear the worst for the general elections. After the murder of
Fortuyn, just a week before the general elections of May 2002, his party (LPF: List
Pim Fortuyn) gained huge success by winning 26 seats, just as the polls had
predicted before the murder of the politician – a result previously unknown in Dutch
politics with a brand new party. The party subsequently formed a controversial
coalition government with Christian democrats and liberals. Mainly on account of
disorganisation within the LPF, the coalition fell after only 87 days. And in the 2003
elections the LPF lost a majority of its seats, retaining only eight seats in Parliament.
Van der Brug (2003) shows that the LPF support resembles many far right-wing
electorates in European countries (Kitschelt, 1995; Lubbers
et al
., 2002). Certainly,
an unfavourable attitude towards immigrants turned out to be an important predictor
Correspondence Address:
Dr. Marcel Lubbers, Radboud University Nijmegen, Department of Sociology,
6500 HE Nijmegen, The Netherlands. Email: M.Lubbers@maw.ru.nl
Correspondence Address:
Dr. Marcel Lubbers, Radboud University Nijmegen, Department of Sociology,
6500 HE Nijmegen, The Netherlands. Email: M.Lubbers@maw.ru.nl
22
M. Lubbers & A. Güveli
of the LPF vote in 2002. Moreover, Van der Burg also showed that political distrust
correlated strongly with LPF voting. Akkerman (2005) meanwhile depicts the LPF
programme as liberal-nationalist, combining leftist ethical issues with strict neutral-
ity in the public domain (where Islam was identified as an anti-liberal religion).
Ignazi (2003) also addressed the party in his comparative work on right-wing
extremism. The party strongly dissociated itself from European extreme right-wing
parties, and accused the left-wing parties of defamation and carrying out a policy of
political correctness. Support from “the man in the street” ensued. Overnight, the
voice of ordinary people became the one to listen to. According to Fortuyn, they had
been kept in silence by the political and socio-cultural elites with their multicultural
dream. In this contribution we investigate to what extent socio-cultural specialists
were indeed less likely to vote for the LPF in 2002, and if so why. We also question
whether Pim Fortuyn gave a voice to the less privileged
per se
or whether they were
also attracted mainly by the political distrust and anti-immigrant sentiments the
party leader proclaimed.
Theory
The moderated relevance of social class in shaping political preferences has often
been corroborated in studies on simple left-wing and right-wing distinctions (Clark
et al
., 1993), but discussion within this tradition reveals that a negative trend parameter
does not hold for all countries (Brooks
et al
., 2006). Moreover, the importance of
social cleavages can be distorted due to the emergence of new political parties that
do not easily fit within the simple left/right distinction. In this way, the inclusion of
new-left parties (Gijsberts & Nieuwbeerta, 2000; Brooks
et al
., 2006) as representa-
tive of the left-libertarian policy axis (Kitschelt, 1995) may result in a smaller dimin-
ishing influence of social class than is found based in a one-dimensional model. For
the Netherlands, however, De Graaf
et al
. (2001) conclude that the formation of the
Green party did not affect the influence of social class. And, in this tradition, less
attention has been paid to the inclusion of far right-wing parties as representative of
authoritarian politics (Billiet & Swyngedouw, 1995; Kitschelt, 1995).
Research on far right-wing voting provides some evidence for an overrepresenta-
tion of the lower occupational positions. In countries where these parties form merely
splinters, it is particularly manual workers who turn out to be overrepresented
(Lubbers
et al
., 2002). For France and Belgium, with somewhat larger far right-wing
parties (Front National and Vlaams Blok, respectively), a less convincing overrepre-
sentation of lower manual workers was accompanied by an overrepresentation of
routine non-manual workers (Mayer, 1996; Lubbers
et al
., 2000).
According to Kitschelt (1995), the importance of social class for the contempo-
rary far right is reflected in the nature of these parties. Unsuccessful “racist authori-
tarian” or “welfare chauvinist” parties would be characterised by an
overrepresentation of blue-collar workers. Successful “populist parties” would not,
however, be characterised by alignment by social class; the Austrian FPÖ serves as
an example. For the “new radical right parties” – the Front Nationale being the ideal
Voting LPF
23
typical party – Kitschelt expects a cross-class alliance of voters, with blue-collar
workers and small businessmen overrepresented and white-collar professionals
underrepresented. The market liberal programme that is combined with authoritar-
ian appeals would attract the social categories. Iversflaten (2005a) also points to the
cross-class alliance of voters in explaining the success of the far right, but argues
that economic alignment is not the reason for this traditionally unnatural alliance.
She claims that merely the authoritarian dimension – read unfavourable attitudes
towards migrants – causes this unusual combination of manual workers and small
businessmen.
Iversflaten’s findings (the result that economic interests are of minor importance)
contrast with the core ideas surrounding the New Radical Right (NRR) parties.
Kitschelt (1995) expected people working in technical and commercial jobs to be
more likely to vote the far right-wing because international competition in these jobs
is heavier than in socio-cultural public sphere jobs – the latter being more dependent
on state financing. People with higher-risk jobs would be more supportive of a
programme favouring tax cuts for the welfare state, which would have become a
burden for hardworking taxpayers and their internationally operating firms. Because
in many Western European countries immigrant groups are dependent on govern-
ment support to a larger extent than natives, a far right-wing party with anti-immi-
grant sentiments and a market liberal view would be perceived by natives to be in
their interest. One of the criticisms to this approach, however, is that the political
programme from the new radical right was less market liberal than Kitschelt
suggested (Lubbers
et al
., 2002; Iversflaten, 2005a). The LPF’s focus on the econ-
omy was minimal, and it was less of a concern among voters – the reason Van
Holsteyn and Irwin (2003) emphasised in their explanations of the rise of the LPF,
“it is not the economy, stupid!” Nevertheless, the attention paid to the economy in
the 2002 election programme (LPF online) focused more on a market liberal
approach than a socialist one, in accordance with Kitschelt’s expectation for
successful far right-wing parties. Fortuyn argued, for example, that taxes and social
insurance contributions should go down, and that mortgage allowances should not
be abolished.
Another way to centralise economic interests of certain social categories is to
depart from ethnic competition theory (Blalock, 1967; Quillian, 1995; Lubbers
et
al
., 2002). Here it becomes apparent that the authoritarian dimension can well be
fuelled by economic interests. In this theory, the proposition is formulated that
workers and self-employed people are more likely to vote for parties that take anti-
immigrant stances as their main focus, because these social categories fear their
position most strongly due to larger (perceived) competition with ethnic out-groups
(Quillian, 1995; Scheepers
et al
., 2002). Far right-wing parties, which claim to serve
the interests of natives by emphasising the importance of national identification and
assimilation of ethnic groups – as Fortuyn did – are expected to be a more attractive
option for those who are most likely to be in competition with such out-groups
(Blalock, 1967; Olzak, 1992; Lubbers
et al
., 2002). For Europe, Kiehl and Werner
(1999) show that ethnic minorities are (strongly) overrepresented in occupations at
24
M. Lubbers & A. Güveli
the lower end of the labour market. The Dutch SCP (2004) provides a similar, more
detailed picture for the Netherlands.
A parallel motivation derived from ethnic competition theory has been expected
and supported for the lower-educated. These groups have stronger competition,
not only over labour, but also over housing, schooling and lifestyles. In this way,
the importance of the attitude towards ethnic minorities in voting far right can still
be motivated economically – contrary to the argument of Iversflaten (2005b).
Lower-educated people, manual workers and self-employed people are expected
to be more likely to vote LPF because they are more dissatisfied with the
economy and give support to less government intervention in the economy. This
relation should then be interpreted by perceptions of perceived economic ethnic
threat.
The crux of the matter is that measurements of economic ethnic threat and
cultural ethnic threat are empirically (as tested with factor analyses) not distinguish-
able (Sniderman
et al
., 2004). Therefore, often unbalanced economic and cultural
explanations are taken into account, of which one is mostly closer to the phenome-
non to be explained than the other (Iversflaten, 2005b), warranting false interpreta-
tions. Sniderman
et al
. also found multicollinearity in their models, but nevertheless
showed that parameters of cultural ethnic threat were stronger related to stereotypes
and ethnic distance than economic ethnic threat. Thus, we may expect that even
though both cultural ethnic threat and economic ethnic threat may interpret the
differences between social classes this should be stronger so by the perceptions of
perceived economic threat. Moreover, we expect that both forms of ethnic threat are
affected by perceptions of economic dissatisfaction, but perceived economic ethnic
threat stronger than cultural ethnic threat.
Socio-cultural Elite
Contrasting the rationale of economic threat of why certain occupational categories
would be more likely to vote extreme right is Kitschelt’s argumentation on why
white-collar professionals are expected to be the least likely to do so. This would be
the result of their communicative experiences and specific capabilities. Kitschelt
(1995: 19) claims that due to lack of data, exploration of the proposition is missing.
With the development of a new class division by Güveli
et al
. (2005, 2007), in which
socio-cultural specialists and technocrats are divided, we argue that we may answer
the question of the extent to which the proposed occupational experiences are
relevant to voting for the LPF.
Güveli
et al
. (2005) adjusted the frequently used EGP class scheme initially
constructed by Erikson
et al
. (1979). They show that the large group of service
workers in Western societies are a combination of socio-cultural specialists and
more technocratically oriented specialists. Dividing the labour market into these
groups has brought new insights into the relation between voting behaviour and
social class in Western European societies since the 1960s. The changing labour
market structure was one of the reasons to reconsider the structure of social class in
Voting LPF
25
their empirical research, as the percentage of service workers has increased relative
to the proportion of manual workers.
Employment relations or work situations are considered to form the basic criteria
for distinguishing social classes (Erikson & Goldthorpe, 1992; Goldthorpe, 2000).
Goldthorpe (2000: 206), for example, claims that “class positions can be understood
as positions defined by employment relations”. Additionally, Kitschelt (1995:7)
argued that work relations are not only characterised by their competitiveness but by
their communicative experiences and capabilities as well:
Individuals who work in symbol- and client-processing organizations where
social relations are at the heart of the work process have a much stronger orien-
tation toward a reciprocal, egalitarian design of democratic politics and cultural
institutions than do individuals who work in strategic and instrumental
economic settings where they manipulate objects, documents and spreadsheets
generated by other instrumental players.
Despite the emphasis on these occupational characteristics, most survey researchers
did not take it into account, either because of a shortage of suitable surveys or
because of the lack of a consensus on how to categorise occupations (Kitschelt,
1995: 9, 19). That said, the distinction between public service workers and others
has been included in previous research. We argue that the systematic categorisation
of higher service jobs according to skills, level of autonomy and possession of
specialised knowledge provides a better test of Kitschelt’s proposition. The adjusted
EGP class scheme provides an opportunity to test this proposition. Inspired by
Kriesi (1989), Güveli
et al
. (2007) distinguished a class of socio-cultural specialists
and one of technocrats within the service class of the EGP class scheme. Kriesi
(1989) asserts that “a basic antagonism of interest” exists between these classes.
Technocrats are supposed to preserve the integrity of the organisation they work for,
while the specialists are more client-oriented or their objective is to act within the
body of knowledge of the discipline they belong to. And Güveli
et al
. (2006) show
that the socio-cultural specialists differ from the technocrats in their socio-political,
cultural and economic preferences and behaviour.
Güveli
et al
. (2005) argue that the classes of socio-cultural specialists and the
technocrats differ in their employment relations. Next to the economic basis of
social classes, social and cultural capital and skills are becoming essential classifica-
tion devices in the employment structure and in the class mechanism in the post-
industrial societies. It is relatively harder for employers to monitor socio-cultural
specialists than technocrats in their work tasks. Furthermore, socio-cultural special-
ists have specific skills and knowledge involving social services and socio-cultural
issues. According to Güveli
et al
. (2007), the egalitarian and progressive attitudes
held by socio-cultural specialists owe much to the relative autonomy they enjoy in
their work tasks. Members of these occupations also tend to possess more humanis-
tic and value-laden knowledge, which makes them more sensitive to non-economic
issues. The relevant criteria that distinguish between technocrats and socio-cultural
26
M. Lubbers & A. Güveli
specialists lead to the hypothesis that socio-cultural specialists are less likely to vote
for the LPF than technocrats. The differences between the occupational categories
should result in different attitudes. We expect the skills, knowledge and work
experience of socio-cultural specialists to have created stronger support for the
importance of stressing social attitudes. Because of their experiences with people in
their jobs and differences between people they have to cope with, we also expect
that in particular a lower level of perceived cultural ethnic threat may explain why
socio-cultural specialists are least likely to vote LPF.
Importance of Attitudes to Explain LPF Voting
We propose that although higher-educated people and higher-status groups are less
likely to vote for the LPF, they are more likely to relate their attitudes to their voting
outcome. Politics are complicated, and political knowledge is an important determi-
nant in voting behaviour (Inglehart, 1990). It is therefore remarkable that previous
research has hardly tested to what extent attitudes are of different relevance for
different social groups in explaining far right-wing voting behaviour. Previous
research showed that people are likely to choose the parties that are closest to the
issues they support (Tillie, 1995; Van der Brug, 2003). The question is to what
extent this relationship between attitudes and party preference is equal for social
categories. We expect that for the higher-educated it is easier to find a party that
corresponds most with their attitudes. Fishbein and Ajzen (1975) also formulated
that the relation between attitudes and behaviour can be distorted because of social
norms. We expect, however, that the social pressure among the higher-educated on
attitudes towards minorities is as large as on voting LPF, and will therefore not
affect this relation. On the other hand, we might expect that people within groups
where Fortuyn was highly popular, but without unfavourable attitudes towards
ethnic minorities, were likely to vote LPF due to group pressure, also resulting in a
lower correlation between attitudes and voting LPF. From this argument and the
notion of consequential acting corresponding to their attitudes, we expect that polit-
ical distrust and ethnic threat increase the likelihood of voting for the LPF among
the higher-educated and socio-cultural specialists more strongly than among the
lower-educated and other social classes. Likewise, we formulate that a stronger
emphasis on the importance of social attitudes decreases the likelihood of voting for
the LPF to a larger extent for higher-educated people and socio-cultural specialists
than it does for lower-educated people and other social classes.
Data and Operationalisation
To test the hypotheses we use the Dutch survey from round 1 from the European
Social Survey (ESS, 2004a). The ESS project is funded by the European Commis-
sion’s 5th Framework Programme, the European Science Foundation and academic
funding bodies in participating countries. The ESS team emphasises the exception-
ally high standards of the design and operation of the project. Since the response
Voting LPF
27
rate for the Netherlands is quite high (at 68%, higher than other Dutch surveys) and
the number of respondents is large enough to distinguish between more categories
than in the standard EGP class scheme (
n
= 2260), we believe the data are well
suited to test our hypotheses. For a comprehensive guide to the data and sampling
procedure, see the ESS data documentation (Jowell
et al
., 2003; ESS, 2004b).
Social Class
Social class is the crucial independent variable in our analyses. In the data, respon-
dents’ occupations are coded as ISCO88 values. These values are entered into the
EGP classification using Ganzeboom’s recoding programme (Erikson
et al
., 1979;
Ganzeboom & Treinman, 1996). As the numbers of farmers and farm workers are
quite small in the Netherlands, we subsumed them into one single category. Retired
or unemployed people were categorised according to their last occupation. People
who never have had an occupation were taken as a separate category of “other”.
To work with a more detailed occupational classification, we used the categorisa-
tions as proposed by Güveli
et al
. (2005, 2007). In these categories, the service class
(Classes I and II of the EGP class scheme) is divided into higher and lower techno-
crats and socio-cultural specialists. This distinction was based on level of autonomy
and possession of specialised service or socio-cultural knowledge. Next to this
standard division used by Güveli
et al
., we applied the same principle of client-
based specialisations to routine non-manual workers, even though for this category
the second criterion of autonomy cannot be applied. We distinguished lower service
workers from lower sales workers, which are subsumed in the lower routine non-
manual jobs in the EGP division. Typical occupations from these distinguished
categories are presented in Table 1.
The other occupational categories are comparable to the ones in the original EGP
class scheme (self-employed people with and without employees; manual
supervisors; skilled manual workers; unskilled manual workers; farmers and farm
labourers). Table 2 presents the distribution of the social class measurement. Here
we find that technocrats and socio-cultural specialists make up large categories in
social class divisions, showing that labour in the Netherlands is to a large extent
Table 1.
Typical jobs in occupational categories of service and non-manual workers
Category Typical jobs
Higher technocrats Computer professionals, engineers, higher managers
Lower technocrats Technicians, safety inspectors, lower managers
Socio-cultural specialists Psychologists, physicians, social scientists, teachers,
nurses, social workers, artists
Routine non-manual workers Clerks
Lower service workers Nursing associates, police officers
Lower sales workers Salespersons, demonstrators
28
M. Lubbers & A. Güveli
specialised. Other large categories are routine non-manual workers and manual
unskilled workers, both making up around 10% of the population of people aged 18
and older.
Other Social Background Characteristics
Education was measured with the Dutch measurement of educational categories.
As we only use the Dutch data from the ESS project, we argue that this measure-
ment provides more detailed information than the available standardised educa-
tion measurement (either by level or in years). Educational level runs in seven
categories from “primary or less” to “tertiary scientific”. As research into voting
behaviour has shown that religion is an important predictor, we have also
included a measure for this (Billiet, 1995). Research on far right-wing voting
behaviour has predicted and corroborated that non-religious people are more
likely to vote for these parties (Billiet & Swyngedouw, 1995; Lubbers
et al
.,
2000; Van der Brug, 2003). To include a measurement of religiosity, we used
church attendance. Four categories divide between “never attends”, “seldom
attends”, “attends once a month” and “attends once a week”. Age (18 and older)
and its quadratic term are included too, as from disintegration theories it is
expected that younger people are more likely to vote for the far right. Finally,
gender is included, since research has shown that men are much more likely to
vote for the far right (Rippl & Siepel, 1998; Lubbers
et al
., 2002; Van der Brug,
2003; Givens, 2004).
Table 2.
Descriptive statistics of social class
Social position category
N
%
Higher technocrats 333 14.7
Lower technocrats 438 19.4
Socio-cultural specialists 379 16.8
Routine non-manual workers 204 9.3
Lower service workers 144 6.1
Lower sales workers 50 2.2
Self-employed with employees 59 2.6
Self-employed without employees 62 2.7
Manual supervisors 96 4.2
Manual skilled workers 101 4.5
Manual unskilled workers 233 10.3
Farmers and labourers 45 2.0
Never had a job/not classified 116 5.1
Total 2260 100.0
Source
: ESS Netherlands 2002, Jowell
et al.
, 2003.
Voting LPF
29
Party Preference
To estimate the likelihood of LPF voting, we used responses from the question
asking which party people voted for in the last general elections. As the survey was
conducted between the general elections of May 2002 and those of January 2003, it
referred to the situation where the LPF had performed best. The percentage of LPF
voters (excluding non-voters) in the data was 13.6% (
n
= 261), something of an
underrepresentation, given that the actual result was 17%. As is usual in survey data,
the percentage of non-voters (13.5%) is lower than the actual turnout. In the 2002
elections, 21% of electors did not vote. To test our hypotheses we constructed a
dichotomous variable “voted for the LPF” versus “not voted for the LPF”. In this
measurement we excluded those who did not vote in the elections.
Socio-political Attitudes
For the measurement of
perceptions of economic ethnic threat
(with a higher score
referring to more threat), we used three bipolar items that appeared to be reliable
measurements. (“immigrants take jobs away versus create new jobs”, “immigrants
take more services in versus take more services out than they put in”, and “immigra-
tion is bad versus good for the economy”) and three items with answer categories
running from strongly agree to strongly disagree (Cronbach’s alpha = 0.75) (“aver-
age salaries and wages are generally brought down by people coming to live and
work here”, “people who come to live here generally harm the economic prospects
of the poor more than the rich” and “if people who have come to live and work here
are unemployed for a long time, they should be made to leave”).
For
perceptions of cultural ethnic threat
(again with a higher score referring to
more threat) we used three items. The item “it is better for a country if almost every-
one shares the same customs and traditions” was also used by Iversflaten (2005b).
The two other items were the bipolar scale whether “immigrants undermine the
cultural life versus enrich the cultural life” and “if a country wants to reduce
tensions it should stop immigration”. The alpha for this scale of cultural threat was
0.69. Economic and cultural threats are factorially not distinguishable. Nevertheless,
comparable to the procedure of Sniderman
et al
. (who use single items for economic
and cultural threat) we continue with the distinction between the scales to test
whether they produce different effects and interpretations.
Political distrust
was measured with four items (“politicians in general care what
people like respondent think”; “politicians are interested in votes rather than in
people’s opinions”; “trust in country’s parliament”; and “trust in politicians”). The
items were transformed into items with similar scale lengths, running from 0 to 10,
where 10 means “no trust”. After factor analyses showed single dimensionality and
reliability analyses provided satisfactory statistics (Cronbach’s alpha = 0.76), we
computed one scale of political distrust by taking the mean of the scores on the four
items. The
importance of social attitudes
was measured with three items, asking
respondents to what extent they believed it to be important “to understand different
30
M. Lubbers & A. Güveli
people”, “that people are treated equally and have equal opportunities”, and “to help
people and care for others’ well-being” (Cronbach’s alpha = 0.60). We took the
mean from the three items to compute the scale.
Satisfaction with the economy
,
referring to the Netherlands at the time of the interview, was measured as such with
a single question. The extent to which people object to
government intervention in
the economy
was also measured with one item, since no other related measurements
were available to construct an appropriate scale. We also take into account left/right
placement, as it turned out to be of relevance in previous research, even though the
placement will be partly the result of positions people take on the addressed
attitudes.
Left/right placement
was measured with a single scale running from 0
(extreme left) to 10 (extreme right) on which people could indicate their position.
All the attitudes were transformed into
z
-scores.
Descriptive Analyses
Before testing our hypotheses, we describe to what extent social class categories
voted for the LPF. Table 3 presents the results, also showing the differences
between categories in non-voting, because this also varies and is the reason for the
different
n
of the categories compared with Table 2. The non-voting pattern is some-
what different than that of LPF voting. The abstention level of the self-employed is
low. Socio-cultural specialists score on the lower end too, as with technocrats.
Reported non-voting is more common among manual workers, farmers and people
who never had a job, of whom between one in four and one in five did not vote.
Large differences are found in LPF support between social class categories. It is
lowest for socio-cultural specialists (5.5%) and highest among self-employed people
without employees (26.8%). Within the higher occupational categories, we also find
large differences in LPF support. The percentage found for socio-cultural specialists
doubles and triples for higher and lower technocrats, respectively. Routine non-
manuals workers, lower services workers and sales workers do not differ very
strongly in their support for the LPF. All categories score around the average of
13.6%. The percentage of LPF voters among manual workers and their supervisors
lies around 17%.
Table 3 also gives the average scores on the attitudes (
z
-scores). Socio-cultural
specialists stand out in their social attitudes, whereas self-employed people with
employees and farmers and farm labourers stress the importance of social attitudes
to a lower degree. Satisfaction with the economy is higher among farmers, and
among both the higher technocrats and socio-cultural specialists, whereas satisfac-
tion is lower among routine non-manual workers and skilled manual workers.
Opposition to government intervention in the economy is more widespread among
farmers and self-employed people, and less so among socio-cultural specialists.
Economic ethnic threat is clearly below average among technocrats and socio-
cultural specialists. This also holds for cultural ethnic threat, with the differences
that the socio-cultural specialists deviate somewhat more extreme from the mean of
zero as compared to their position on economic ethnic threat, whereas for the higher
Voting LPF
31
Table 3.
Descriptive statistics: percentage of LPF voters, percentage of non-voters, and average scores on socio-political attitudes (
z
-scores) by
social class
Social class category % votes
for LPF
N
% non-
voters
N
Importance
of social
attitudes Satisfaction
economy Contra gov.
intervention
Economic
ethnic
threat
Cultural
ethnic
threat Political
distrust Left/right
placement
Higher technocrats 13.5 296 9.9 333 0.03 0.12
−
0.05
−
0.23
−
0.15
−
0.23 0.07
Lower technocrats 16.9 384 10.7 438 0.00 0.01
−
0.10
−
0.05
−
0.05
−
0.08 0.06
Socio-cultural specialists 5.5 346 7.1 379 0.16 0.13
−
0.20
−
0.32
−
0.42
−0.31 −0.21
Routine non-manual
workers 13.2 159 20.1 204 −0.13 −0.20 0.03 0.19 0.17 0.21 −0.06
Lower service workers 13.2 114 19.4 144 0.10 −0.09 0.02 −0.03 −0.02 0.08 −0.12
Lower sales workers 15.2 46 6.0 50 0.00 −0.04 0.02 0.11 −0.01 −0.03 0.05
Self-employed with
employees 16.7 54 3.4 59 −0.21 −0.08 0.09 0.19 0.30 0.28 0.27
Self-employed without
employees 26.8 56 9.9 62 0.05 −0.02 0.20 0.14 0.09 0.19 0.09
Manual supervisors 16.7 84 10.4 96 −0.16 0.07 0.07 −0.08 0.04 0.07 0.13
Skilled manual workers 17.7 79 20.8 101 −0.05 −0.23 0.11 0.42 0.41 0.37 0.04
Unskilled manual
workers 17.0 182 20.2 233 −0.06 −0.06 0.14 0.24 0.22 0.34 −0.03
Farmers and labourers 8.8 34 20.0 45 −0.29 0.26 0.30 0.32 0.28 −0.08 0.16
Never had a job/not
classified 9.4 85 25.0 116 0.01 −0.12 0.39 0.44 0.58 0.28 0.20
Total 13.6 1919 13.5 2260 0 0 0 0 0 0 0
Source: ESS Netherlands 2002, Jowell et al., 2003.
32 M. Lubbers & A. Güveli
technocrats the reversed is found. Economic and cultural ethnic threats are preva-
lent, particularly amongst skilled manual workers, farmers and farm labourers, and
people who never had a job. With regard to political distrust, again technocrats and
socio-cultural specialists score at the lower end, and manual workers at the opposite
end. Finally, socio-cultural specialists place themselves furthest to the left. Self-
employed people with employees place themselves furthest to the right.
Results
Our first exercise is to compare the relevance of the detailed social class measure-
ment in voting for the LPF compared with the traditional EGP classification.
Second, we control the effect of the best-fitting social class determinant for other
relevant determinants of LPF voting behaviour, i.e. education, gender, age and
church attendance. Hereafter we include socio-political attitudes to find out how
much these explain differences between social categories. Finally, we show to what
extent the interactions as expected exist between education and social class and
socio-political attitudes. We use logistic regression to test our hypotheses.
Table 4 presents two models to predict LPF voting. The first model provides
differences in LPF voting between social class categories according the EGP classi-
fication. The second model presents differences between social class categories
according to the proposed more detailed classification, distinguishing between
socio-cultural specialists and technocrats. The fit statistics at the lower end of
the table show that the model fit improves significantly at the p < 0.01 level in the
second model only, where the more detailed class measurement is included. In the
first model, inclusion of the EGP measurement only improves model fit signifi-
cantly at the p < 0.10 level. In other words, this traditional model discriminates less
strongly between categories in the likelihood of LPF voting. When we turn to the
parameters in the first model, we see this corroborated. Self-employed people
without employees are clearly more likely to vote LPF than higher controllers, the
reference category. Moreover, unskilled and skilled manual workers are more likely
to vote LPF than higher controllers. Hence, based on this model we should conclude
that social class is a minor predictor and that self-employed people without employ-
ees are more likely to vote LPF. This is comparable to findings in previous research.
The second model includes the more detailed class scheme. This model has a
much better fit, and is highly significant. With the socio-cultural specialists as the
reference category in this model, we see that almost all social categories are signifi-
cantly more likely to vote the LPF compared to the socio-cultural specialists. The
exceptional position of the socio-cultural specialists makes the predictor of social
class highly relevant, as they are much less likely to vote LPF than all other catego-
ries. In particular the self-employed are much more likely to vote LPF as compared
to the socio-cultural specialists, with the odds ratios for the self-employed with and
without employees being 3.44 and 6.30, respectively. But also the well-educated
category of higher technocrats has a significantly higher likelihood of voting LPF.
The outcome is very close to the prediction Kitschelt made for the new radical-right
Voting LPF 33
Table 4. EGP versus more categories of social class predicting the likelihood of voting LPF
Model 1 B s.e. Odds Model 2 B s.e. Odds
Social class: Social class:
Higher controllers (Ref) Socio-cultural specialists (ref)
Higher technocrats 0.99** 0.29 2.69
Lower controllers 0.07 0.20 1.07 Lower technocrats 1.26** 0.27 3.51
Routine non-manual workers 0.16 0.28 1.17 Routine non-manual workers 0.97** 0.33 2.64
Lower sales and service workers 0.11 0.26 1.11 Lower service workers 0.95** 0.36 2.58
Lower sales workers 1.13** 0.47 3.09
Self-employed with employees 0.42 0.40 1.53 Self-employed with employees 1.24** 0.44 3.44
Self-employed without employees 0.96** 0.34 2.60 Self-employed without employees 1.84** 0.38 6.30
Manual supervisors 0.39 0.34 1.48 Manual supervisors 1.24** 0.38 3.44
Skilled manual workers 0.60∼0.32 1.82 Skilled manual workers 1.31** 0.38 3.71
Unskilled manual workers 0.48* 0.22 1.62 Unskilled manual workers 1.26** 0.31 3.53
Farmers and labourers −0.30 0.62 0.74 Farmers and labourers 0.51 0.65 1.67
Others −0.14 0.40 0.87 Never had a job/not classified 0.58 0.44 1.79
Intercept −2.04 −2.84
−2 log likelihood 1509.34 1470.74
Chi-square model improvement;
compared to null model (df) 16.55∼(10) 38.96** (12)
Nagelkerke R20.01 0.04
**p < 0.01; *p < 0.05; ∼p < 0.10.
Source: ESS Netherlands 2002, Jowell et al., 2003, own computations; N = 1888.
34 M. Lubbers & A. Güveli
type of extreme right-wing parties. It is not the cross-alliance of manual workers and
self-employed people, though. More important is that all categories are almost as
likely to vote for the LPF, with the exception of socio-cultural specialists. Further-
more, we find that the distinction between service and sales workers does not add
much to the understanding of LPF voting. Both categories vary to a similar extent
from the socio-cultural specialists.
Now that we have ascertained that a more detailed social class measurement,
including the distinction between socio-cultural specialists and technocrats, deter-
mines voting for the LPF more effectively than the traditional EGP measurement,
we continue our analyses controlling for other relevant predictors. In Model A of
Table 5 we include education, gender, age and church attendance alongside to social
class.1
Comparable with previous research, we find that the higher the level of education,
the less likely one is to vote LPF. What is remarkable though is that the model fit
increases less than expected due to the inclusion of education. Additional tests show
that social class contributes more strongly to the model fit than does education. The
improvement in model fit in Model A can, to a large extent, be attributed to the
measurement of church attendance. In keeping with previous research, we find that
people who attend church once a week or more are less likely to vote LPF than
people who never attend church. Remarkably absent are the effects of age and
gender. Gender effects are widely reported in far right-wing research, with men
being more likely to cast such a vote (Givens, 2004), and contrasts the finding from
Van der Brug (2003). The small difference between men and women (the parameter
is still in the expected direction) is comparable to other less outspoken far right-
wing parties. Thus, the LPF seems more comparable with Alpine far right-wing
parties such as the FPÖ (McGann & Kitschelt, 2005) than to the Belgian, French
and Scandinavian extreme right parties (Evans, 2005). In Model A, social class
parameters decrease – compared with those presented in Table 4 – implying that
educational differences and differences in Church attendance explain a part of the
differences between social categories when voting LPF, particularly between socio-
cultural specialists, manual workers, and self-employed people. However, the
difference between socio-cultural specialists and higher technocrats remains highly
significant, and we conclude that, next to education, the division of social classes
into technical and socio-cultural specialists accounts best for the explanation of LPF
voting.
Effects from Socio-political Attitudes
In Model B in Table 5 we include three of the socio-political attitudes: social attitudes,
satisfaction with the economy, and opposition to government intervention in the econ-
omy. All of the three terms are significant, implying that as expected, people who
strongly subscribe the importance of social attitudes are less likely to vote LPF. More-
over, people who are more satisfied with the economy are less likely to vote LPF, and
stronger opposition to government intervention in the economy increases the likeli-
Voting LPF 35
Table 5. Logistic regression predicting the likelihood of voting LPF
Model A Model B Model C Model D
Bs.e. Bs.e. Bs.e. Bs.e.
Social class:
Socio-cultural specialists (ref)
Higher technocrats 0.88** 0.30 0.88** 0.30 0.84** 0.32 0.77* 0.32
Lower technocrats 0.99** 0.28 0.98** 0.28 0.88** 0.29 0.80* 0.30
Routine non-manual workers 0.59∼0.35 0.52 0.35 0.44 0.37 0.37 0.37
Lower service workers 0.68∼0.38 0.64∼0.38 0.71∼0.40 0.62 0.40
Lower sales workers 0.67 0.49 0.62 0.49 0.56 0.51 0.63 0.51
Self-employed with employees 0.95* 0.45 0.81∼0.45 0.42 0.49 0.35 0.50
Self-employed without employees 1.42** 0.40 1.35** 0.41 1.20** 0.42 1.16** 0.43
Manual supervisors 0.83* 0.40 0.83* 0.40 0.74∼0.42 0.62 0.43
Skilled manual workers 0.87* 0.41 0.82* 0.41 0.59 0.43 0.50 0.44
Unskilled manual workers 0.76* 0.34 0.73* 0.34 0.58 0.11 0.57 0.36
Farmers and labourers 0.24 0.67 0.21 0.69 0.16 0.69 0.14 0.70
Never had a job/not classified 0.48 0.47 0.35 0.48 0.09 0.50 0.02 0.50
Education −0.20** 0.05 −0.17** 0.05 −0.10∼0.06 −0.07 0.06
Gender: male 0.14 0.15 0.10 0.16 0.10 0.17 0.04 0.17
Age 0.01 0.03 0.01 0.03 0.01 0.03 0.01 0.03
Age-quadratic term 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Church attendance
Once a week −1.35** 0.31 −1.32** 0.32 −1.61** 0.33 −1.63** 0.33
Once a month −0.44 0.27 −0.32 0.27 −0.25 0.28 −0.33 0.29
Seldom −0.19 0.16 −0.15 0.16 −0.17 0.17 −0.17 0.17
36 M. Lubbers
Table 5. Continued.
Model A Model B Model C Model D
Bs.e. Bs.e. Bs.e. Bs.e.
Never (reference)
Importance of social attitudes (z-score) −0.18** 0.07 −0.06 0.07 −0.03 0.07
Satisfaction economy (z-score) −0.24** 0.07 −0.10 0.08 −0.12 0.08
Contra gov. intervention in economy (z-score) 0.25** 0.07 0.16* 0.07 0.16* 0.07
Left/right scale (z-score) 0.62** 0.08 0.55** 0.08
Political distrust (z-score) 0.41** 0.09 0.35** 0.09
Economic ethnic threat (z-score) 0.30** 0.08
Cultural ethnic threat (z-score) 0.57** 0.08
Intercept −2.23 −1.82 −2.13 −2.04
−2 log likelihood 1426.56 1392.63 1262.13 1225.71
Chi-square model improvement; compared to
previous model (df) 44.19** (7) 33.93** (3) 130.50** (3) 166.92** (3)
Nagelkerke R20.08 0.11 0.22 0.25
**p < 0.01; *p < 0.05; ∼p < 0.10.
Source: ESS Netherlands 2002, Jowell et al., 2003, own computations; N = 1888.
Voting LPF 37
hood of voting LPF. Surprisingly, the inclusion of the three attitudes hardly explains
the differences in chances between the socio-cultural specialists and the other social
class categories, as the parameters in Model B are largely comparable with the param-
eters in Model A. Only the larger likelihood of the self-employed people voting LPF
as compared with the socio-cultural specialists has decreased slightly. The effect of
education is also somewhat smaller in Model B, implying that the attitudes partly
explain why a higher education results in a lower likelihood of voting LPF.
In the subsequent Model C, we find that LPF voting is strongly determined by left/
right-wing placement, political distrust, and economic ethnic threat. Of these three
attitudes, it is left/right placement that has the strongest influence (B = 0.62),
followed by political distrust (B = 0.41). In Model D, we replace perceived economic
ethnic threat by perceived cultural ethnic threat. Now we find that the strongest influ-
ence is from cultural ethnic threat (B = 0.57). This effect is larger than the effect from
left/right placement in this model (B = 0.55). Even though economic and cultural
ethnic threats are not factorially distinct, the different consequences of the measure-
ments are strong. We cannot conclude that economic ethnic threat has no influence
on LPF voting, but we can conclude that the cultural ethnic threat has more.
Let us turn to the comparison of the models in explaining the effects of the other
attitudes and differences between social categories. We do so both by comparing the
changes in the parameters between Models B, C and D, respectively, and by the
results of additional regression analyses of the attitudes (see Appendix 1). The effect
of social attitudes on voting LPF is explained by the attitudes “ethnic threat”, “polit-
ical distrust” and “left/right placement”. Appendix 1 shows us that social attitudes
reduce both economic and cultural ethnic threat – almost to the same extent.
Moreover people with a social attitude distrust politics less and place themselves
further to the left. The influence of satisfaction with the economy on voting LPF is
explained by the other attitudes as well. Importantly though, satisfaction with the
national economy has a quite strong effect on perceived economic ethnic threat (B =
−0.18; Appendix 1) and a more modest effect on perceived cultural threat (B = −0.09;
Appendix 1). However, greater satisfaction with the economy leads to a decrease in
political distrust. Based on these findings we have to conclude that economic
evaluations do affect ethnic threat and consequently LPF voting – the cultural threat
explanation turns out to be of greater importance. The effect of opposition to
government intervention remains significant after controlling for ethnic threat,
political distrust and left/right placement.
The effect of education on voting LPF is largely explained by political distrust,
ethnic threat, and left/right placement. Interestingly, education affects cultural ethnic
threat more than economic ethnic threat (Appendix 1). Consequently, the effect of
education is explained better in Model D as compared with Model C. Social class
differences are also partly explained by the introduction of political distrust, ethnic
threat, and left/right placement. In Models C and D – when attitudes are controlled
for – it is the same group of categories that remain significantly more likely to vote
for the LPF than the socio-cultural specialists: higher and lower technocrats, and self-
employed people without employees. But it is the cultural ethnic threat that provides
38 M. Lubbers & A. Güveli
a better explanation of the differences between the social class categories voting LPF
as compared with the economic ethnic threat. We can see this comparing Models C
and D, where in Model D the parameters of social class have diminished somewhat
more. The exception is the category of the unskilled manual workers. In Appendix 1,
we see this finding corroborated. The effects of social class are more pronounced on
cultural ethnic threat than on economic ethnic threat. Higher and lower technocrats
as well as self-employed people with employees are more likely to perceive a larger
cultural ethnic threat than the socio-cultural specialists, whereas they do not differ in
their perceived economic ethnic threat. For unskilled manual workers and farmers, it
remains the case that they perceive a larger economic ethnic threat than the socio-
cultural specialists, but not a significantly larger cultural ethnic threat.
Interactions
To answer the question of whether socio-political attitudes increase the likelihood of
voting LPF for higher- and lower-educated people, socio-cultural specialists and
other social classes to the same extent, we added interaction effects to the last model
presented in Table 5. The results of the interactions with education are presented in
Table 6, and the interactions between socio-cultural specialists and attitudes are
shown in Table 7.
The interactions between education and attitudes are significant, except in the case
of cultural ethnic threat. The main effect of education is its effect for people with an
average attitude (the value 0 on the z-scores of the attitudes). The main effect of the
socio-political attitude is within the category of the lowest-educated people (the
value 0 on the education variable). The interaction term can be interpreted as the
change in the effect of socio-political attitude when education increases. Most
remarkable is that three of the four socio-political attitudes have no significant effect
on voting LPF for the least educated people. This changes rapidly when education
level increases. For lower-educated people, for whom the likelihood of voting LPF is
larger, it does not matter much whether people perceive an economic ethnic threat or
not. For the higher-educated this is highly relevant. Higher-educated people without
perceptions of economic ethnic threat are far less likely to vote for the LPF; for those
who perceive an economic ethnic threat, the likelihood of voting LPF is comparable
to the likelihood of the lower-educated people with a strong economic ethnic threat
perception. This interaction effect is depicted in Figure 1.
Figure 1. Effects of economic ethnic threat on voting LPF for the lowest and highest educational categories. Source: ESS Netherlands 2002.
We make similar findings for political distrust and social attitudes. These are
highly discriminatory among the higher-educated, but far less so among lower-
educated categories. As education increases, so does the importance of political
distrust and social attitudes as determinants of LPF voting. Regarding a cultural
ethnic threat, we find no significant interaction effect. The term is positive, and
hence comparable with the interaction between education and an economic ethnic
threat, but it is not sufficiently strong to reach statistical significance. So, for both
the lower- and higher-educated it holds that a stronger cultural ethnic threat
increases the likelihood to vote LPF to a similar extent (Figure 2).
Voting LPF 39
Figure 2. Effects of cultural ethnic threat on voting LPF for the lowest and highest educational categories. Source: ESS Netherlands 2002.
Socio-cultural specialists were less likely to vote LPF than the many other social
class categories. We could only partly find an explanation for this in the different
attitudes. But do these attitudes determine LPF voting for socio-cultural specialists
similarly, as compared with the other categories? All interaction terms indicate that
the effect of attitude is stronger within the category of socio-cultural specialists,
although the interactions with ethnic threat are not significant (see Table 7). Among
the socio-cultural specialists, voting LPF is more likely than among other social class
categories when one perceives an ethnic threat to be stronger, distrusts politics more
strongly, or when one stresses to a lower degree the importance of social attitudes.
Social Class and Voting Other Parties
To compare the extent to which it applies specifically for voting for the LPF that the
distinction between socio-cultural specialists and technocrats is a good predictor
variable, we calculated kappa indices that provide the relation between social class
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
-2SD -1SD 0 +1SD +2SD
< economic ethnic threat>
< logit of vothing LPF >
Lowest educated
Highest educated
Figure 1. Effects of economic ethnic threat on voting LPF for the lowest and highest
educational categories. Source: ESS Netherlands 2002.
Table 6. Interaction between education and economic and cultural ethnic threat, political
distrust and social attitudes on voting LPF
Economic
ethnic threat Cultural ethnic
threat Political distrust Importance of
social attitudes
Bs.e. Bs.e. Bs.e. Bs.e.
Education (0–6) −0.12* 0.06 −0.08 0.06 −0.14* 0.06 −0.12* 0.06
Socio-political
attitude (z-score) 0.07 0.13 0.43** 0.15 0.13 0.15 0.15 0.13
Education * socio-
political attitude 0.09* 0.05 0.05 0.05 0.12** 0.05 −0.08* 0.04
Controlled for all other characteristics from Model C of Table 5.
Source: ESS Netherlands 2002, Jowell et al., 2003; N = 1888.
40 M. Lubbers & A. Güveli
and voting, controlling for number of categories being used (Gijsberts &
Nieuwbeerta, 2000).2 We provide kappa indices both for the traditional EGP social
class measurement as well as for the measurement proposed by Güveli et al.,
distinguishing between technocrats and socio-cultural specialists. We calculated
these kappa indices for the bivariate relation, hence regressing LPF voting on social
class.3 The results are presented in Table 8.
Table 8 shows the indices for voting for various parties and party combinations.
We find that the class voting index based on the distinction between socio-cultural
specialists and technocrats (EGPG) is highest for LPF voting (0.26), although when
it comes to abstention from voting the index is even higher. Social class turns out to
be related particularly strongly to abstention (0.46). For all parties and party combi-
nations other than the LPF, the class voting index based on the distinction proposed
by Güveli et al. is lower. After the LPF, it is strongest for the social democratic PvdA
-2
-1.5
-1
-0.5
0
0.5
1
-2SD -1SD 0 +1SD +2SD
< cultural ethnic threat >
< logit of voting LPF >
Lowest educated
Highest educated
Figure 2. Effects of cultural ethnic threat on voting LPF for the lowest and highest
educational categories. Source: ESS Netherlands 2002.
Table 7. Interaction between socio-cultural specialists and economic and cultural ethnic
threat, political distrust and social attitudes on voting LPF
Economic
ethnic threat Cultural ethnic
threat Political distrust Importance of
social attitudes
Bs.e. Bs.e. Bs.e. Bs.e.
Socio-cultural
specialists (compared
to higher technocrats)
−0.85** 0.33 −0.87** 0.35 −0.98** 0.35 −1.00** 0.35
Socio-political
attitude 0.28** 0.08 0.56** 0.08 0.38** 0.08 −0.00 0.08
Socio-cultural
specialists * socio-
political attitude
0.15 0.26 0.34 0.26 0.65* 0.30 −0.57* 0.27
Controlled for all other characteristics from Model C of Table 5.
Source: ESS Netherlands 2002, Jowell et al., 2003; N = 1888.
Voting LPF 41
(0.22). The relation between class and voting is especially low for the Christian
parties.
Comparing traditional EGP class divisions and the proposed distinction between
socio-cultural specialists and technocrats, (EGPG) shows that EGPG provides a better
explanation for voting LPF and Greens as well as for abstention. The LPF and the
Greens can be considered to be the parties at the extremes of the authoritarian dimen-
sion as worked out by Kitschelt (1995). For the other parties, the EGPG index is some-
what lower or comparable to the class voting index based on the EGP classification.
Conclusions and Discussion
In this study we contributed to the research on political parties that emphasise
national identity and assimilation of immigrants, by focussing on explanations of
voting LPF. First, we addressed the question whether the distinction between social
class categories by their level of autonomy and specialised knowledge contributes in
explaining voting for the LPF. Second, we tested explanations of why these and
previously corroborated social class differences exist, by focussing on the
importance of social attitudes, and the distinction between economic and cultural
ethnic threat. Third, we showed to what extent attitudinal explanations interact with
social-structural characteristics.
We found large differences between socio-cultural specialists on the one hand
and technocrats on the other, the latter being much more supportive of the LPF.
Although small businessmen were found to be most likely to vote LPF, the
differences between technocrats, routine non-manual workers and even manual
Table 8. Class voting indices (kappa)
EGPG EGP
LPF 0.26 0.12
Liberal (VVD) 0.16 0.23
Liberals (VVD, D66) 0.16 0.22
Christian Democrats (CDA) 0.07 0.09
Christians (CDA, SGP, CU) 0.07 0.09
Old left (PvdA, SP) 0.17 0.17
Social Democrats (PvdA) 0.22 0.22
Greens 0.14 0.11
Left (PvdA, Greens, SP) 0.13 0.14
Cultural (Greens, D66) 0.12 0.14
Abstention 0.46 0.43
Source: ESS Netherlands 2002, Jowell et al., 2003, own computations.
EGPG: EGP classification added with the distinction between socio-cultural specialists and
technocrats, as proposed by Güveli et al. (2005).
EGP: Erikson, Goldthorpe and Portocarero social class classification.
42 M. Lubbers & A. Güveli
workers turned out to be small. Taking the differences between socio-cultural
specialists and technocrats into account instead of merely distinguishing between
lower and higher service class workers (or lower and higher controllers) provides a
better explanation of LPF voting behaviour. More remarkably, this measurement
even competes with educational attainment in the strength of its effect on LPF
voting. With these findings we refute the idea of the LPF being a party that would
not be aligned by social class. Kitschelt’s formulation that occupational categories
should be divided not only according to lines of competitiveness but also according
to specialised knowledge has proven to be correct.
A cross-class alliance of voters (referring to an overrepresentation of self-
employed people without employees and manual workers) was not found to be
consistently supportive of the LPF. Small businessmen did turn out to be in favour of
the LPF, but manual workers were not found to deviate significantly from either
technocrats, routine non-manual workers or lower sales and service employees. Once
more, the basic divide in voting for the LPF lies between socio-cultural specialists
and all other occupational categories. We showed that subscribing the importance of
social attitudes hardly explained differences between social classes in their likeli-
hood to vote LPF. Only lower technocrats and self-employed with employees turned
out to be somewhat less in support of the social attitudes. Because social attitudes
affect other important determinants of LPF voting (ethnic threat, political distrust
and left/right placement), an indirect effect of these social attitudes exists.
The differences between cultural ethnic threat and economic ethnic threat turned
out to shed some light on the lower likelihood of the socio-cultural specialists to
vote LPF compared with the technocrats. Higher and lower technocrats perceive a
significantly stronger cultural ethnic threat than the socio-cultural specialists,
whereas they do not differ in their perceived economic ethnic threat. Where
economic ethnic threat also explains the larger likelihood of the unskilled manual
workers to vote LPF, cultural ethnic threat does so to some extent for the techno-
crats. Technocrats stay significantly more likely to vote LPF though, even after
controlling for the attitudes. Hence, there must be other explanation why they do so.
Previous research has not focused on the extent to which attitudes are equally
important for different social categories. We find that the answer to this question is
relevant. The interaction effects between social attitudes and education and class are
significant. People who perceive an economic ethnic threat or distrust politics are
more likely to vote LPF, regardless of their education or class. People disassociating
themselves from these attitudes are less likely to vote LPF – much more so when they
are higher-educated or socio-cultural specialists. We have to conclude that conse-
quential acting does not differ between lower- and higher-educated people when they
support LPF core programmatic aspects. For the lower-educated without support for
the core ideas of Fortuyn, we may speculate that they are either driven by sentiments
about the murder of Fortuyn or by possible group pressure to vote Fortuyn.
We are aware that our findings are based on the Dutch political and occupational
structure. However, almost all advanced economies have moved from an industrial to
a post-industrial employment structure and the criteria for distinguishing the socio-
Voting LPF 43
cultural specialists should apply to other countries as well. Furthermore, almost all
Western societies experience a rise of a (far right-wing) party that emphasises national
identity and assimilation of immigrants. Therefore, we encourage scholars to use the
adjusted EGP class schema, the details of which are reported in Appendix 2. Future
research should reveal whether our conclusions apply to other countries as well.
Notes
1. We are aware that sector of employment is an important factor to control for. We expect that sector
of employment explains a part of the differences between the socio-cultural specialists and the other
class members in voting for LPF. Unfortunately, we could not distinguish between the private and
public sector in the ESS data.
2. Calculated as the sum of squares of the deviation parameters of all social class categories, divided by
the product of the number of voting outcomes and the number of categories of social class.
3. We dropped the category of farmers and farm labourers from the computation of class voting index
because within this category no one voted D66 or GroenLinks (Green Left), making estimation of the
respective outcome impossible. As the category is small, we think distortions are minimal.
References
Akkerman, T. (2005) Anti-immigration parties and the defence of liberal values: the exceptional case of
the List Pim Fortuyn, Journal of Political Ideologies, 10(3), pp. 337–354.
Billiet, J. (1995) Church involvement, ethnocentrism, and voting for a radical right-wing party: diverging
behavioural outcomes of equal attitudinal dispositions, Sociology of Religion, 56(3), pp. 303–326.
Billiet, J. & Swyngedouw, M. (1995) De keuze tussen Groen en Extreem Rechts in Vlaanderen; sporen
van een breuklijn, Res Publica, 2, pp. 247–262.
Blalock, H. M. (1967) Toward a Theory of Minority Group Relations (New York: John Wiley).
Brooks, C., Nieuwbeerta, P. & Manza, J. (2006) Cleavage-based voting behaviour in cross-national
perspective: evidence from six postwar democracies, Social Science Research, 35(1), pp. 88–128.
Brug, W. van der (2003) How the LPF fuelled discontent: empirical tests of explanations of LPF-support,
Acta Politica, 38(1), pp. 89–106.
Clark, T. N., Lipset, S. & Rempel, M. (1993) The declining political significance of social class,
International Sociology, 8, pp. 293–316.
Erikson, R. & Goldthorpe, J. H. (1992) The Constant Flux: A Study of Class Mobility in Industrial
Societies (Oxford: Clarendon Press).
Erikson, R., Goldthorpe, J. H. & Portocarero, L. (1979) Intergenerational class mobility in three Western
European societies, British Journal of Sociology, 30, pp. 415–441.
European Social Survey (ESS) Netherlands (2002).
European Social Survey (ESS) (2004a) ESS Data File, Version 4.1, available online at: <http://
ess.nsd.uib.no> (accessed January 2004).
European Social Survey (ESS) (2004b) ESS Survey Documentation and Fieldwork Documents, available
online at: <http://ess.nsd.uib.no> (accessed January 2004).
Evans, J. A. J. (2005) The dynamics of social change in radical right-wing populist party support,
Comparative European Politics, 3, pp. 76–101.
Fishbein, M. & Ajzen, I. (1975) Belief, Attitude, Intention and Behaviour: An Introduction to Theory and
Research (Reading, MA: Addison–Wesley).
Ganzeboom, H. B. G. & Treinman, D. J. (1996) Internationally comparable measures of occupational
status for the 1988 International Standard Classification of occupations, Social Science Research,
25, pp. 201–239.
Gijsberts, M. & Nieuwbeerta, P. (2000) Class cleavages in party preferences in the new democracies in
Eastern Europe. A comparison with Western democracies, European Societies, 2(4), pp. 397–430.
44 M. Lubbers & A. Güveli
Givens, T. E. (2004) The radical right gender gap, Comparative Political Studies, 37(1), pp. 30–54.
Goldthorpe, J. H. (2000) On Sociology: Numbers, Narratives and the Integration of Research and Theory
(Oxford: Oxford University Press).
Graaf, N. D. de, Heath, A. & Need, A. (2001) Declining cleavages and political choices: the interplay of
social and political factors in the Netherlands, Electoral Studies, 20, pp. 1–15.
Güveli, A., Need, A. & Graaf, N. D. de (2005) De opkomst van de “nieuwe” sociale klassen binnen de
dienstenklasse in Nederland. Politieke oriëntatie van de “nieuwe” sociale klassen tussen 1970 en
2000, Mens & Maatschappij, 80(3), pp. 257–279.
Güveli, A., Need, A. & Graaf, N. D. de (2006) Socio-political, cultural and economic preferences and
behaviour of the social and cultural specialists versus the technocrats: social class or education?,
Social Indicator Research, 79, 1–35.
Güveli, A., Need, A. & Graaf, N. D. de (2007) The rise of the ‘new’ social classes within the service class
in the Netherlands. Political orientation of social and cultural specialists and technocrats between
1970 and 2003, Acta Sociologica, 50(2) (forthcoming).
Holsteyn, J. van & Irwin, G. A. (2003) Never a dull moment: Pim Fortuyn and the Dutch Parliamentary
election of 2002, West European Politics, 26(2), pp. 41–66.
Inglehart, R. (1990) Culture Shift in Advanced Industrial Society (Princeton: Princeton University Press).
Ignazi, P. (2003) Extreme Right Parties in Western Europe (Oxford: Oxford University Press).
Ivarsflaten, E. (2005a) The vulnerable populist right parties: no economic realignment fuelling their
success, European Journal of Political Research, 44, pp. 465–492.
Iversflaten, E. (2005b) Threatened by diversity: why restrictive asylum and immigration policies appeal
to Western European, Journal of Elections, Public Opinion and Parties, 15(1), pp. 21–45.
Jowell, R. & the Central Coordinating Team (2003) European Social Survey: Technical Report (London:
Centre for Comparative Social Surveys, City University).
Kiehl, M. & Werner, H. (1999) The labour market situation of EU and of third country nationals in the
European Union, Institut für Arbeitsmarkt- und Berufsforchung, Labour Market Research Topics,
32, pp. 1–25.
Kitschelt, H. (1995) The Radical Right in Western Europe. A Comparative Analysis (Ann Arbor:
University of Michigan Press).
Kriesi, H. (1989) New social movement and the new class in the Netherlands, American Journal of
Sociology, 94, pp. 1078–1116.
Lubbers, M., Gijsberts, M. & Scheepers, P. (2002) Extreme right-wing voting in Western Europe,
European Journal of Political Research, 41(3), pp. 345–378.
Lubbers, M., Scheepers, P. & Billiet, J. (2000) Multilevel modelling of Vlaams Blok voting: individual
and contextual characteristics of the Vlaams Blok vote, Acta Politica, 35(4), pp. 363–398.
Mayer, N. (1996) The National Front vote and right-wing extremism, 1988–1995, Research on
Democracy and Society, 3, pp. 197–221.
McGann, A. & Kitschelt, H. (2005) The radical right in the Alps: evolution of support for the Swiss SVP
and Austrian FPÖ, Party Politics, 11(2), pp. 147–172.
Olzak, S. (1992) Dynamics of Ethnic Competition and Conflict (Stanford, CA: Stanford University Press).
Quillian, L. (1995) Prejudice as a response to perceived group threat: population composition and anti-
immigrant and racial prejudice in Europe, American Sociological Review, 60, pp. 586–611.
Rippl, S. & Seipel, C. (1999) Gender differences in right wing extremism: intergroup validity of a
second-order construct, Social Psychology Quarterly, 62(4), pp. 381–393.
Scheepers, P., Gijsberts, M. & Coenders, M. (2002) Ethnic exclusionism in European countries, public
opposition to grant civil rights to legal migrants as a response to perceived ethnic threat, European
Sociological Review, 18(1), pp. 17–34.
SCP (2004) Rapportage Minderheden (The Hague: SCP).
Sniderman, P., Hagendoorn, L. & Prior, M. (2004) Predisposing factors and situational triggers:
exclusionary reactions to immigrant minorities, American Political Science Review, 98(1), pp. 35–49.
Tillie, J. (1995) Party Utility and Voting Behaviour (Amsterdam: Het Spinhuis).
Voting LPF 45
Social
attitude Satisfaction
economy Contra economy
intervention Political
distrust Economic
ethnic threat Cultural
ethnic threat Left/right
Social class:
Socio-cultural specialists (ref)
Higher technocrats −0.02 −0.00 0.08 −0.02 0.08 0.16* 0.19*
Lower technocrats −0.15* 0.05 0.07 0.06 0.11 0.17* 0.24**
Routine non-manual workers −0.28** −0.06 0.09 0.12 0.14 0.16∼0.06
Lower service workers −0.17 0.04 0.12 0.09 −0.03 0.08 0.05
Lower sales workers −0.12 0.03 0.19 0.07 0.14 0.08 0.21
Self-employed with employees −0.33* −0.11 0.23 0.31* 0.19 0.32* 0.26*
Self-employed without employees −0.09 0.12 0.36* 0.28* 0.30* 0.28* 0.21
Manual supervisors −0.10 0.01 0.17 0.19∼0.06 0.10 0.18
Skilled manual workers −0.03 −0.16 0.12 0.32** 0.34** 0.30* 0.11
Unskilled manual workers −0.07 0.04 0.19∼0.25** 0.25* 0.12 0.02
Farmers and labourers −0.45* 0.30 0.21 −0.13 0.31∼0.14 0.01
Never had a job/not classified −0.16 0.14 0.38** 0.00 0.30* 0.37** 0.29*
Education 0.02 0.07** −0.05** −0.09** −0.10** −0.16** −0.04*
Gender: male −0.28** 0.19** 0.08 −0.05 −0.17** 0.07 0.15**
Age 0.01** 0.02** 0.00 0.03** −0.01 −0.01 −0.01
Age-quadratic term 0.00 0.00** 0.00 0.00** 0.00 0.00** 0.00
Appendix 1. Regression of Socio-political Attitudes
46 M. Lubbers
Social
attitude Satisfaction
economy Contra economy
intervention Political
distrust Economic
ethnic threat Cultural
ethnic threat Left/right
Church attendance:
Once a week 0.05 0.16* −0.06 −0.20** 0.03 0.13* 0.53**
Once a month 0.22** 0.16* −0.17* −0.19** −0.06 0.09 0.21**
Seldom 0.02 0.09∼−0.02 −0.13* 0.07 0.04 0.15**
Never (ref.)
Importance of social attitude (z-score) −0.04* −0.10** −0.13** −0.17**
Satisfaction economy (z-score) −0.27** −0.18** −0.09** 0.04
Contra government intervention economy
(z-score) 0.17** 0.04 0.04 0.05*
R20.03 0.03 0.04 0.21 0.12 0.19 0.08
**p < 0.01; *p < 0.05; ∼p < 0.10.
Source: ESS Netherlands 2002, Jowell et al., 2003, own computations; N = 1888.
Appendix 1. Continued
Voting LPF 47
Appendix 2. ISCO88 Codes of Classes of Technocrats and Socio-cultural
Specialists (Güveli et al., 2005, 2006)
Higher Technocrats (1000 1100 1110 1120 1200 1210 1220 1222 1223 1224 1225
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1239 1250 1251
2000 2100 2110 2111 2112 2113 2114 2120 2121 2122 2130 2131 2140 2142 2143
2144 2145 2146 2147 2411 2420 2443 3143 3144).
Socio-cultural Specialists (2141 2213 2220 2221 2222 2223 2224 2229 2230 2300
2310 2320 2321 2322 2323 2330 2331 2332 2340 2350 2351 2352 2359 2400 2412
2419 2421 2422 2429 2430 2431 2432 2440 2441 2442 2444 2445 2446 2450 2451
2452 2453 2454 2455 2460 2470 3131 3200 3210 3229 3240 3241 3242 3470 3471
3472 3473 3474 5150 5151 5152).
Lower Technocrats (1130 1140 1141 1142 1143 1240 1252 1300 1310 1312 1313
1314 1315 1316 1317 1318 1319 2132 2139 2148 2410 3000 3100 3110 3111 3112
3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3130 3132 3133 3139
3140 3141 3142 3145 3150 3151 3152 3211 3212 3213 3220 3221 3222 3223 3224
3225 3226 3227 3228 3400 3410 3411 3412 3413 3414 3415 3416 3417 3419 3420
3421 3422 3423 3429 3431 3432 3434 3440 3441 3442 3443 3444 3449 3450 3451
3475).
