Professional Documents
Culture Documents
47
Cmo citar
Leonisio, Rafael y Strijbis, Oliver (2014). Ms all de la autoubicacin: por qu el nacionalismo predice
mejor el comportamiento electoral en el Pas Vasco. Revista Espaola de Investigaciones Sociolgicas,
146: 47-68.
(http://dx.doi.org/10.5477/cis/reis.146.47)
* Rafael Leonisio agradece la financiacin del Gobierno Vasco a travs del Grupo consolidado IT-610-13. Ambos autores
agradecen a Edurne Bartolom por facilitarles la versin vasca de la Encuesta Europea de Valores de 1999.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
48 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 49
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
50 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 51
1 2 3 4 5
Impuestos y contaminacin medioambiental 0,8529
Contaminacin medioambiental sin costes 0,723
Los polticos y Dios 0,6391
Los lderes religiosos deberan influir en la poltica -0,7529
Religin y cargos pblicos 0,6904
Estado liberal espaol en el siglo XIX7. Aun- voreca (Montero, 1998). Los partidarios de
que oficialmente fueron motivos dinsticos Carlos, hermano de Fernando VII los de-
los que desencadenaron la guerra civil que nominados carlistas, defendan el absolu-
comenz en Espaa en 1833, la causa sub- tismo y la confesionalidad del Estado, es
yacente fue la competencia entre las partes decir, la sociedad tradicional. Por el contra-
en conflicto con el fin de promover los mo- rio, los seguidores de Isabel, la hija del rey,
delos de sociedad que cada una de ellas fa- eran partidarios del Estado liberal, del libre
comercio y de ciertas medidas de laicizacin
(si bien muy moderadas). Sin embargo, la
7La primera dimensin fue la oposicin campo-ciudad, victoria de esta ltima no resolvi el proble-
que vena de muy atrs, y que, en la edad moderna, ha ma; la oposicin carlista, que perdur duran-
tenido parte en los conflictos entre los habitantes del
mundo rural-agrcola y la burguesa comercial de las te todo el siglo XIX, provoc una nueva guerra
ciudades (Garca de Cortzar y Lorenzo, 1997). civil tras la proclamacin de la monarqua
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
52 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 53
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
54 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
merando preferencias relativas a diferentes Por lo tanto, resulta crucial para el debate
dimensiones temticas. sobre si la dimensin izquierda-derecha o la
Durante muchos aos, los politlogos dimensin nacionalista es ms importante
han deducido las posiciones ideolgicas de para las decisiones de los votantes en el Pas
los votantes pidindoles que se autoubica- Vasco entender lo que izquierda y dere-
sen en escalas ideolgicas. Es evidente que cha significan en este contexto particular.
la escala ms importante en este sentido ha Una manera de conseguirlo es mediante
sido la escala izquierda-derecha. Desde el la prediccin de la autoubicacin de los vo-
modelo espacial de competencia electoral de tantes en la escala izquierda-derecha a partir
Downs (1957), la escala izquierda-derecha ha del posicionamiento de los votantes en las
sido ampliamente utilizada para modelar el diversas dimensiones temticas. Hemos
comportamiento electoral, el comportamien- creado estas dimensiones temticas para la
to estratgico de los partidos y otros aspectos versin vasca de la Encuesta Europea de Va-
similares. No obstante, el modelo izquierda- lores de 1999, aplicando el anlisis factorial
derecha, como tal, presenta al menos dos que se describe en el apartado anterior (va-
inconvenientes muy graves (para una discu- se la tabla 1). Adems, es posible realizar un
sin ms completa vase Mair, 2007). En pri- anlisis similar haciendo uso del estudio del
mer lugar, los trminos izquierda y dere- CIS Modelos espaciales de preferencias
cha no describen necesariamente las polticas en su edicin de 2009 (CIS 2799).
distinciones econmicas clsicas entre inter- Si bien estos datos son menos tiles para la
vencin estatal y liberalismo econmico. En induccin de dimensiones temticas9, el es-
cambio, conglomeran a menudo un amplio tudio est bien diseado para probar las re-
conjunto de otras oposiciones, convirtiendo laciones entre las dimensiones temticas y la
la escala izquierda-derecha en un super- escala izquierda-derecha. La encuesta del
issue (Knutsen, 1995; Van der Eijk et al., CIS pidi a los encuestados ubicarse en es-
2005: 166). En segundo lugar, el significado calas que perfilaban las dimensiones laicis-
de la escala izquierda-derecha vara en fun- mo frente a confesionalismo, Estado de bien-
cin del contexto. En realidad, la nocin de estar frente a reduccin de impuestos,
izquierda y derecha no se limita ni siquie- proinmigracin frente a anti-inmigracin, ley
ra a las cuestiones polticas, sino que tam- y orden frente a libertades cvicas, y ecologa
bin puede hacer referencia a los valores y frente a crecimiento econmico. Adems, la
lealtades partidistas, as como a grupos so- encuesta sonde las preferencias de los en-
ciales tales como la clase (p. ej., Knutsen, cuestados con respecto a la secesin de una
1998; Freire, 2006; Milic, 2008). El significado comunidad autnoma frente a la integridad
de los trminos izquierda y derecha para
del Estado espaol. Dado que solo hemos
los ciudadanos vara con el tiempo dentro de
seleccionado a los encuestados del Pas
un pas, y probablemente an ms de unos
Vasco, interpretamos estas preferencias en
pases a otros. Incluso en un momento dado
referencia a la escala nacionalismo vasco
en el tiempo, los significados de izquierda
frente a espaolismo. Se utilizaron escalas
y derecha pueden variar considerablemen-
te entre los ciudadanos, en funcin de las
inclinaciones partidistas y del grado de sofis-
9 Las preguntas tienden a relacionarse con dimensiones
ticacin poltica (Zechmeister, 2006). En resu- temticas bastante generales, lo que significa que los
men, la dimensin izquierda-derecha puede diseadores de la encuesta dedujeron lo que podra
tener diferentes significados en diferentes considerarse como dimensiones temticas potencialmente
relevantes. El hecho de haber dado con unas dimensiones
puntos en el tiempo, e incluso en el mismo muy similares respalda nuestro anlisis inductivo (y las
punto en el tiempo de unos votantes a otros. deducciones de los diseadores de la encuesta).
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 55
N 343
de once puntos para todas las preguntas, Tal y como se muestra en la tabla 2 para
incluyendo una pregunta acerca de la autou- los datos de la versin vasca de la Encues-
bicacin izquierda-derecha, por lo que los ta Europea de Valores, nuestra hiptesis de
resultados presentan un alto grado de com- que la dimensin izquierda-derecha fun-
parabilidad entre s. ciona igualmente como un super-issue
Con el fin de predecir la autoubicacin en en el Pas Vasco recibe respaldo emprico.
la escala izquierda-derecha, hemos aplicado Aunque la distincin clsica entre interven-
regresiones de mnimos cuadrados ordina- cin estatal y liberalismo econmico se
rios utilizando los valores de las dimensiones relaciona significativamente con la autou-
temticas (las puntuaciones factoriales en el bicacin en la escala izquierda-derecha
caso de los datos de 1999) como variables (cuando se controlan otras dimensiones de
independientes, y los valores de la autoubi- temticas), est claro que no se trata de la
cacin izquierda-derecha como variables nica correlacin. Adems, las preferen-
dependientes10. Nuestra meta no es ni expli- cias sobre dimensiones temticas conside-
car la autoubicacin izquierda-derecha en su radas normalmente como liberal frente a
conjunto ni hacer afirmaciones causales so- conservador (p. ej., Conover y Feldmann,
bre la relacin entre izquierda y derecha 1981), libertario-cosmopolita frente a auto-
y las preferencias temticas. Ms modesta- ritario-particularista (Kitschelt, 1992), ma-
mente, nuestro objetivo es mostrar que, aun- terialista frente a postmaterialista (Ingle-
que la autoubicacin izquierda-derecha no hart, 1990), y verde/alternativo/libertario
constituye una medida particularmente bue- (en ingls: gal) frente a tradicional/autorita-
rio/nacionalista (en ingls: tan) (Hooghe et
na de las preferencias sobre intervencin
al., 2002) se hallan estadsticamente rela-
estatal y liberalismo econmico, se halla, no
cionadas de manera significativa con la
obstante, vinculada a las preferencias sobre
dimensin izquierda frente a derecha.
diferentes dimensiones temticas.
Esto quiere decir que los ciudadanos vas-
cos ms laicistas, ecologistas y poltica-
mente liberales tienden a ubicarse todos
10 Dado que nuestras variables independientes se
componen de factores procedentes de un mismo anlisis ellos ms a la izquierda en la escala que los
factorial, hemos evitado (como era de esperar) los ciudadanos ideolgicamente opuestos.
problemas de multicolinealidad (correlacin alta entre Ms an, las preferencias de los ciudada-
variables independientes: 0,186; p<0,01). Verificamos
tambin la distribucin normal de los residuos (media: nos relativas a la dimensin nacionalismo
7,93E-15; desviacin estndar: 0,992). vasco-espaolismo tambin se hallan rela-
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
56 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
N 526
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 57
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
58 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
TABLA 4.Modelos de regresin logstica multinomial de intencin de voto para las elecciones generales
espaolas del 2000
Variable
dependiente: Variables independientes Modelo 1 Modelo 2 Modelo 3 Modelo 4
voto
Nacionalismo espaol/vasco -1,397*** -1,380***
Liberalismo econmico/interven-
cin estatal 0,192 0,187 0,0923
PP Religioso/laicista 0,0732 -0,0222
Crecimiento econmico/ecologa -0,0429 -0,0794
Autoritarismo/liberalismo poltico -0,197 -0,191
Nacionalismo espaol/vasco -1,214*** -1,227***
Liberalismo econmico/interven-
cin estatal 0,172 0,187 0,0964
PSOE Religioso/laicista 0,461 0,397
Crecimiento econmico/ecologa 0,0900 0,0499
Autoritarismo/liberalismo poltico -0,174 -0,190
Nacionalismo espaol/vasco -0,0359 0,186
Liberalismo econmico/interven-
cin estatal 1,049*** 1,197*** 1,177***
HB/EH Religioso/laicista 0,829*** 0,825***
Crecimiento econmico/ecologa 0,403 0,380
Autoritarismo/liberalismo poltico 0,889*** 1,034***
Nacionalismo espaol/vasco 0,370 0,409
Liberalismo econmico/interven-
cin estatal 0,715** 0,738** 0,678*
EA Religioso/laicista 0,416 0,459
Crecimiento econmico/ecologa 0,104 0,0608
Autoritarismo/liberalismo poltico -0,264 -0,258
Nacionalismo espaol/vasco -0,689* -0,663*
Liberalismo econmico/interven-
cin estatal 0,599* 0,633* 0,611*
IU Religioso/laicista 0,101 0,0166
Crecimiento econmico/ecologa 0,0964 0,0816
Autoritarismo/liberalismo poltico 0,106 0,0764
Categora de referencia:
PNV
R2 de Nagelkerke 0,0548 0,0338 0,0840 0,1412
N 301 301 301 301
*p<0,1, **p<0,05, ***p<0,01. No se muestran los resultados de los Verdes, los votos en blanco ni la abstencin. Los mo-
delos incluyen constantes.
Fuente: Versin vasca de la Encuesta Europea de Valores (1999).
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 59
Dow y Endersby aplicaron con xito las re- PSOE en relacin con el PNV. La posicin del
gresiones MNL)11. votante en la dimensin nacionalismo vasco-
Hemos llevado a cabo nuestro anlisis espaolismo, por el contrario, no tiene nin-
con y sin un conjunto de variables de control. gn efecto significativo sobre la probabilidad
Los modelos sin variables de control no pue- de su voto a EH o EA en lugar del PNV. Este
den excluir la omisin de terceras variables. ltimo caso no es sorprendente, teniendo en
Esto solo es problemtico, sin embargo, cuenta que EA ocupa una posicin muy si-
cuando se trata de la omisin de variables milar en la dimensin nacionalismo vasco-
que tienen un efecto sobre los coeficientes espaolismo (Llera, 1988). Una explicacin
de nuestras variables independientes, ya que para la creciente importancia de la dimen-
nuestro objetivo no es explicar una variacin sin econmica izquierda-derecha con res-
mxima en las preferencias electorales, sino pecto a las preferencias electorales entre EH
y PNV podra encontrarse en el carcter con-
comparar los efectos de nuestras dos varia-
siderablemente heterogneo en ese momen-
bles independientes. Esto significa, adems,
to de la coalicin EH, que deriv de la tregua
que la inclusin de variables de control solo
de ETA. En dichas elecciones, HB se presen-
sera adecuada en los casos en que tales va-
t junto con otros pequeos partidos de iz-
riables no fueran causalmente anteriores a la
quierda en la coalicin EH. Adems, reuni
formacin de las preferencias temticas.
votos prestados de votantes que queran
La tabla 4 muestra los modelos MLN de
fortalecer EH con el fin de consolidar el pro-
intencin de voto para las elecciones gene-
ceso de paz. Otra explicacin para la menor
rales de 2000. Los valores indican su impac-
importancia de la dimensin nacionalismo
to en la probabilidad de voto a cada partido
vasco-espaolismo para el voto a EH con
con respecto a los votos al Partido Naciona-
respecto al voto al PNV podra ser atribuida
lista Vasco (PNV), que (debido a su posicin
a la posicin del PNV, el cual se encontraba
central en el sistema de partidos dada su
ms cerca de los nacionalistas ms radicales
ideologa y magnitud) ha sido elegido como
de lo que lo estaba antes, debido a la alinea-
la categora de referencia. Tal y como mues-
cin de las fuerzas polticas nacionalistas
tran los resultados del modelo 1, la dimen-
tras la firma de los Acuerdos de Lizarra en
sin nacionalismo vasco-espaolismo est 1998.
fuertemente relacionada con la probabilidad
En contraste con la dimensin nacionalis-
de voto a los conservadores espaoles (PP)
mo vasco-espaolismo, la dimensin econ-
o a los socialistas espaoles (PSOE). Podra-
mica izquierda-derecha tuvo una importan-
mos interpretar estos resultados de la si-
cia considerablemente menor para las
guiente manera: cuanto menor es la tenden-
preferencias electorales en 2000. Esto se
cia nacionalista vasca de un votante, lo ms
puede ver en el hecho de que el R2 de Na-
probable es que este vote por el PP o el
gelkerke es tres veces menor en el segundo
modelo que en el primero. Estas relaciones
no son sorprendentes en la medida en que
11Todos los modelos se calcularon como regresin
probabilstica multinomial. No se encontr ninguna se hallan vinculadas con las preferencias
desviacin sustancial con respecto a los modelos electorales: los votantes de EH, EA e IU se
logsticos. Por otra parte, la multicolinealidad puede ser hallan ms a favor de la intervencin estatal
rechazada. No se incluyen las variables de control, ya
que nuestro objetivo no es explicar una variacin mxima que los votantes del PNV. Esta imagen no
en las preferencias electorales, sino comparar los efectos cambia cuando modificamos la categora de
de nuestras variables independientes. En tal caso, solo referencia. Si, por ejemplo, tomamos el Par-
sera adecuada la inclusin de variables de control que
no sean causalmente anteriores a la formacin de tido Socialista como categora de referencia
preferencias temticas. (datos no presentados), nos encontramos
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
60 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
con que un mayor grado de nacionalismo podemos discernir una relacin estadstica-
vasco aumenta la probabilidad de votar a mente significativa entre la probabilidad de
uno de los partidos nacionalistas vascos voto al PSE-EE en lugar de al PNV, y su po-
(PNV, HB/EH, EA), y que un grado menor sicin en la dimensin econmica izquierda-
aumenta la probabilidad de votar por los derecha. No obstante, la distancia entre los
conservadores espaoles (PP), mientras que dos grupos de votantes es claramente me-
la posicin de los votantes en la dimensin nor en este caso que en la dimensin nacio-
econmica izquierda-derecha no tiene nin- nalista. En contraste con las elecciones de
gn efecto sobre la probabilidad de que vo- 2000, nos encontramos con que la dimen-
ten por un partido distinto a los socialistas. sin econmica no se relaciona con la pro-
En los modelos 3 y 4 se introducen las babilidad de voto al nacionalismo vasco de
otras dimensiones temticas. El modelo 3 izquierdas (EA y Aralar) en lugar del PNV. Una
solo es interesante en comparacin con el vez ms, no encontramos ningn resultado
modelo 1, ya que nos permite comprender la significativo en la dimensin nacionalista.
capacidad explicativa de la dimensin nacio- La comparacin de los modelos 1 y 2
nalismo vasco-espaolismo con respecto a deja claro que la dimensin nacionalista po-
todas las dems dimensiones juntas. Un va- see un poder explicativo de la variacin en
lor R2 de Nagelkerke de alrededor de 0,05 las preferencias electorales de 2009 cinco
muestra que, en 1999, la dimensin naciona- veces mayor que la dimensin econmica.
lismo vasco-espaolismo por s sola posee Por otra parte, la comparacin entre los mo-
una capacidad explicativa relativamente si- delos 1 y 3 muestra que, en 2009, la dimen-
milar a la de todas las dems dimensiones sin nacionalista fue ms importante para las
temticas juntas (en este caso, el valor R2 es preferencias electorales que el resto de di-
0,08). El modelo 4 muestra la relacin entre mensiones juntas. El modelo 4 reitera la rela-
cada dimensin temtica y las preferencias cin entre las preferencias temticas y las
electorales a la vez que se controlan todas preferencias electorales para todas las di-
las dems dimensiones temticas. La rela- mensiones. Adems, es interesante observar
cin entre nacionalismo y preferencia electo- que los votantes de Aralar y de IU se distin-
ral interpretada anteriormente no hace frente guen sustancialmente de los votantes del
a objeciones sustanciales, y lo mismo vale PNV en sus preferencias ms libertarias,
para la dimensin econmica. Sin embargo, mientras que no existe ningn indicio de que
hallamos tambin que la intencin de voto a el laicismo siga desempeando un papel im-
Euskal Herritarrok (en lugar del PNV) se rela- portante. Sin embargo, no cabe excluir este
ciona con fuertes preferencias laicistas y an- ltimo caso, ya que a Batasuna no se le per-
tiautoritaristas. Por lo tanto, este modelo re- miti participar en estas elecciones y, por lo
vela que la distancia ideolgica entre los tanto, es posible que los votantes ms clara-
votantes que apoyaron a EH y todos los de- mente distinguibles no se concentraran en-
ms es muy grande, independientemente de tonces en un solo partido.
si se trata de nacionalistas vascos o no. En resumen, nos encontramos con que la
Hallamos resultados similares en la inten- divisin nacionalismo vasco-espaolismo es
cin de voto con respecto a las elecciones en general ms importante para las prefe-
autonmicas de 2009 (vase la tabla 5). Una rencias electorales de los votantes del Pas
vez ms, nos encontramos con una diferen- Vasco que la dimensin econmica izquierda-
cia significativa en las preferencias de los derecha. La dimensin econmica izquier-
votantes del PP y del PSOE frente a las de da-derecha tuvo cierta importancia en 1999
los votantes del PNV con respecto a la di- dentro del bloque de los partidos nacionalistas
mensin nacionalista. En este caso, tambin vascos. Esto sugiere que, en el caso de una
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 61
TABLA 5.Modelos de regresin logstica multinomial de recuerdo de voto para las elecciones autnomas vascas de 2009
Variable
dependiente: Variables independientes Modelo 1 Modelo 2 Modelo 3 Modelo 4
voto
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
62 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 63
te de competicin poltica. Sin embargo, los Bartolini, Stefano (2000). The Political Mobilisation of
resultados de este anlisis sugieren lo contra- the European Left, 1860-1980: The Class Clea-
rio: las cuestiones econmicas parecen ser vage. Cambridge: Cambridge University Press.
de poca relevancia para la poltica electoral Benoit, Kenneth y Laver, Michael (2005). Mapping
vasca. Por lo tanto, el espacio poltico vasco the Irish Policy Space: Voter and Party Spaces
actual podra entenderse como unidimensio- in Preferential Elections. Economic and Social
Review, 36 (2): 83-108.
nal antes que como bidimensional. Una lectu-
ra alternativa de los resultados sera que otras Budge, Ian et al. (eds.) (2001). Mapping Policy Pre-
dimensiones temticas tales como las relacio- ferences: Estimates for Parties, Electors and Go-
vernments 1945-1998. Oxford: Oxford University
nadas con la divisin religioso-laicista o las
Press.
relacionadas con la importancia del creci-
miento econmico frente a la prioridad de la Calle, Luis de la (2005). Cuando la proximidad deja
de ser importante. Revista Espaola de Ciencia
proteccin del medio ambiente son tan im-
Poltica, 12: 21-52.
portantes como la dimensin econmica iz-
quierda-derecha. Esta interpretacin de nues- Conover, Pamela J. y Feldman, Stanley (1981). The
Origins and Meaning of Liberal/Conservative
tros resultados pondra claramente de relieve
Self-Identifications. American Journal of Political
la multidimensionalidad del espacio poltico
Science, 25 (4): 617-645.
en el Pas Vasco y planteara importantes pre-
Dinas, Elias (2012). Left and Right in the Basque
guntas acerca de qu dimensiones polticas
Country and Catalonia: The Meaning of Ideology
estn ganando en importancia en el contexto
in a Nationalist Context. South European Soci-
de pacificacin del conflicto vasco. ety and Politics, 17(3): 467-485.
Sin embargo, nuestros resultados no solo Dow, Jay K. (2001). A Comparative Spatial Analysis
tienen que ver con un debate sobre la dimen- of Majoritarian and Proportional Elections. Elec-
sionalidad del espacio poltico vasco. Por el toral Studies, 20 (1): 109-125.
contrario, nuestro anlisis apoya la argumen- y Endersby, James W. (2004). Multinomial Pro-
tacin de politlogos como Knutsen (1995) o bit and Multinomial Logit: A Comparison of
Van der Eijk et al. (2005) segn los cuales la Choice Models for Voting Research. Electoral
escala izquierda-derecha no es unidimensio- Studies, 23 (1): 107-122.
nal, sino un super-issue que contiene una Downs, Anthony (1957). An Economic Theory of De-
diversidad de dimensiones que pueden variar mocracy. New York: Harper and Row.
adems en el tiempo y el espacio. Esto nos Feire, Andr (2006). Bringing Social Identities Back
lleva a la pregunta de si nuestras observacio- In: The Social Anchors of Left-Right Orientation
nes concernientes a la multidimensionalidad in Western Europe. International Political Scien-
de la escala izquierda-derecha son aplicables ce Review, 27 (4): 359-378.
tambin al conjunto de Espaa y, de ser as, Fernndez Albertos, Jos (2002). Votar en dos di-
cmo cambiara esto nuestra comprensin de mensiones: el peso del nacionalismo y la ideo-
la poltica electoral espaola y la manera en loga en el comportamiento electoral vasco.
que debemos aproximarnos a ella. 1993-2001. Revista Espaola de Ciencia Polti-
ca, 6: 153-181.
Garca de Cortzar, Fernando y Lorenzo, Jos Mara
Bibliografa (1997). Historia del Pas Vasco. San Sebastin:
Txertoa.
Alonso, Sonia; Volkens, Andrea y Gmez, Braulio
(2012). Anlisis de contenido de textos polticos. Gillespie, Richard (2000). Political Polarization in the
Un enfoque cuantitativo. Madrid: CIS. Basque Country. Regional & Federal Studies, 10
(1): 112-124.
Ansolabehere, Stephen y Puy Segura, Mara Socorro
(2013). Ideology, Nationalism, and Identity in Basque Granja, Jos Luis de la (1995). El nacionalismo vasco:
Regional Elections. Working Paper no publicada. un siglo de historia. Madrid: Tecnos.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
64 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
(2002). Paz entre dos guerras civiles. En: Bazan, Lipset, Seymur M. (2001). Cleavages, Parties and
I. (ed.). De Tbal a Aitor. Historia de Vasconia. Ma- Democracy. En: Karvonen, L. y Kuhnle, S.
drid: La Esfera de los Libros. (eds.). Party Systems and Voter Alignments Re-
visited. London: Routledge.
Gunther, Richard; Sani, Giacomo y Shabad, Goldie
(1986). El sistema de partidos en Espaa. Gne- y Rokkan, Stein (1967). Cleavage structures,
sis y evolucin. Madrid: CIS. party system, and voter alignments: An introduc-
tion. En: Lipset, S. M. y Rokkan, S. (eds). Party
Hooghe, Liesbet; Marks, Gary y Wilson, Carol J.
systems and voter alignments. Nueva York: Free
(2002). Does Left/Right Structure Party Positions
Press.
on European Integration?. Comparative Political
Studies, 35 (8): 965-989. Llera, Francisco Jos (1986). Las segundas elec-
ciones autonmicas vascas. Revista de Derecho
Inglehart, Ronald (1977). The Silent Revolution: Chan-
Poltico, 23: 135-165.
ging Values and Political Styles among Western
Publics. Princeton: Princeton University Press. (1988). Continuidad y cambio en el sistema de
partidos vascos: 1977-1987. Revista de Estu-
(1990). Culture Shift in Advanced Industrial So-
dios Polticos, 59: 277-375.
cieties. Princeton: Princeton University Press.
(1989). Continuidad y cambio en la poltica
(2008). Changing Values among Western Publics
vasca: notas sobre identidades sociales y cultu-
from 1970 to 2006. West European Politics, 31
ra poltica. Revista Espaola de Investigaciones
(1): 130-146.
Sociolgicas, 47: 107-135.
Kennedy, Peter (2008). A Guide to Econometrics (6 (1999). Basque Polarization: between Autonomy
ed.). Oxford: Blackwell. and Independence. Nationalism and Ethnic Po-
Kitschelt, Herbert (1992). The Formation of Party litics, 5 (3-4): 101-120.
Systems in East Central Europe. Politics and (2000). Los gobiernos de coalicin en el Pas
Society, 20 (1): 7-50. Vasco. En: Matas, J. (ed.). Coaliciones polticas
Klingemann, Hans-Dieter et al. (2006). Mapping Po- y gobernabilidad. Barcelona: ICPS.
licy Preferences II: Estimates for Parties, Electors (2001). La encrucijada vasca. Revista de Oc-
and Governments in the OECD, EU and CEE cidente, 241: 89-113.
1990-2003. Oxford: Oxford University Press.
Mair, Peter (2007). LeftRight Orientations. En: Dal-
Knutsen, Oddbjrn(1995).Value Orientations, Poli- ton, R. J. y Klingemann, H.-D. (eds.). The Oxford
tical Conflicts and Left-right Identification: A Handbook of Political Behaviour. Oxford: Oxford
Comparative Study. European Journal of Politi- University Press.
cal Research, 28 (1):63-93.
Mata, Jos Manuel (1998). Nationalism and Political
(1998). The Strength of Partisan Competition of Parties in the Autonomous Community of the
Left-Right Identity: A Comparative Longitudinal Basque Country (ACBC). Strategies and Tensions.
Study of Left-Right Party Polarization in Eight West Working Paper 137. Barcelona: ICPS.
European Countries. Party Politics, 4 (1): 5-31.
Milic, Thomas (2008). Links, Mitte, Rechts: Worauf
Kriesi, Hanspeter et al. (2008). West European Politics grndet die ideologische Selbstidentifikation der
in the Age of Globalization. Cambridge: Cambrid- Schweizer und Schweizerinnen?. Swiss Political
ge University Press. Science Review, 14 (2): 245-285.
Leonisio, Rafael (2012). Parliament on the Center- Molina, Fernando (2010). The Historical Dynamics
right, Government on the Left: Explaining the of Ethnic Conflicts: Confrontational Nationa-
Basque Exceptionalism. Regional and Federal lisms, Democracy and the Basques in Contem-
Studies, 22 (1): 45-60. porary Spain. Nations and Nationalism, 16 (2):
y Strijbis, Oliver (2011). Izquierda-Derecha vs 240-260.
Centro-Periferia: una aproximacin al discurso de Montero, Manuel (1998). Historia del Pas Vasco. San
los partidos polticos vascos (1977-2009). Re- Sebastin: Txertoa.
vista Espaola de Ciencia Poltica, 26: 63-85.
Pallars, Francesc; Muoz, Jordi y Retortillo, Alfredo
Linz, Juan Jos et al. (1986). Conflicto en Euskadi. (2006). Depolarization in the 2005 Autonomous
Madrid: Espasa-Calpe. Elections in the Basque Country: Towards a New
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 65
Scenario for Peace?. Regional and Federal Stu- Strijbis, Oliver y Leonisio, Rafael (2012). Political
dies, 16 (4): 465-479. Cleavages in the Basque Country: Meaning and
Salience. Regional and Federal Studies, 22 (5):
Prez-Nievas, Santiago y Bonet, Eduardo (2006).
595-611.
Identidades regionales y reivindicacin de au-
togobierno. El etnorregionalismo en el voto a Torcal, Mariano y Chhibber, Pradeep (1995). Elites,
partidos nacionalistas de Blgica, Espaa y Rei- cleavages y sistema de partidos en una democra-
cia consolidada: Espaa (1986-1992). Revista
no Unido. Revista Espaola de Ciencia Poltica,
Espaola de Investigaciones Sociolgicas, 69: 7-38.
15: 123-161.
Van der Eijk, Cees; Schmitt, Hermann y Binder,
Sani, Giacomo y Sartori, Giovanni (2006). Polariza-
Tanja (2005). Left-right Orientations and Party
cin, fragmentacin y competicin en las demo-
Choice. En: Thomasen, J. (ed.). The European
cracias occidentales. En: Sartori, G. Partidos y Voter: A Comparative Study of Modern Demo-
sistemas de partidos. Madrid: Alianza. cracies. Oxford: Oxford University Press.
Stoll, Heather (2010). Elite-level Conflict Salience Zechmeister, Elizabeth (2006). Whats Left and Whos
and Dimensionality in Western Europe: Concepts Right? A Q-method Study of Individual and Con-
and Empirical Findings. West European Politics, textual Influences on the Meaning of Ideological
33 (3): 445-473. Labels. Political Behavior, 28 (2): 151-173.
RECEPCIN: 30/07/2012
REVISIN: 25/07/2013
APROBACIN: 25/10/2013
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
66 Ms all de la autoubicacin: por qu el nacionalismo predice mejor el comportamiento electoral en el Pas Vasco
Apndice
Para cada una de las [afirmaciones] ledas, me puede decir si est Vd. muy de acuerdo, de acuerdo, en
desacuerdo o muy en desacuerdo?1
Estara de acuerdo con un aumento de los impuestos si el dinero extra fuese usado para prevenir la con-
taminacin del medio ambiente.
El Gobierno debe reducir la contaminacin medioambiental sin que ello me cueste dinero.
En qu medida est Vd. de acuerdo o en desacuerdo con cada una de los siguientes [afirmaciones] (muy
de acuerdo, de acuerdo, ni de acuerdo ni en desacuerdo, en desacuerdo, muy en desacuerdo)?
Los polticos que no creen en Dios no son aptos para ejercer cargos pblicos.
Los lderes religiosos no deberan influir en cmo vota la gente en las elecciones.
Sera mejor que hubiera ms gente con fuertes creencias religiosas en los cargos pblicos.
De acuerdo con sus opiniones, dnde se colocara Vd. en esta escala (escala de 1 a 10)?
Los parados deberan aceptar cualquier trabajo disponible o bien perder el subsidio de paro./Los parados
deberan tener derecho a no aceptar un trabajo que no desean.
La competencia es buena: estimula a las personas a trabajar duro y a desarrollar nuevas ideas./La com-
petencia es perjudicial: saca a flote lo peor de las personas.
El Estado debera dar ms libertad a las empresas./El Estado debera controlar a las empresas ms efec-
tivamente.
Para cada uno [tipo de sistema poltico], dira Vd. que es una manera muy buena, bastante buena, bastante
mala o muy mala de gobernar este pas?
Tener un lder fuerte que no tenga que preocuparse por el parlamento ni las elecciones.
Que sean expertos, y no el Gobierno, quienes tomen decisiones de acuerdo con lo que piensan que es
mejor para el pas.
De acuerdo con sus opiniones, dnde se colocara Vd. en esta escala (escala de 1 a 10), en lo que es extre-
ma izquierda y la extrema derecha, hablando en trminos generales?
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
Rafael Leonisio y Oliver Strijbis 67
Algunas personas piensan que deberan mejorarse los servicios pblicos y las prestaciones sociales
aunque para ello haya que pagar ms impuestos (en una escala de 0 a 10, estas personas se situa-
ran en la posicin 10). Otras piensan que es ms importante pagar menos impuestos aunque ello
signifique reducir los servicios pblicos y las prestaciones sociales (estas personas se situaran en la
posicin 10 de la escala). Hay otras personas que se sitan en posiciones intermedias. En qu
posicin se situara Vd.?
Refirindonos ahora a la inmigracin, piense en una escala en la que el punto 0 representa la libre
entrada de inmigrantes y el punto 10 la restriccin absoluta a la entrada de inmigrantes, en qu
posicin cree Vd. que se encuentra Espaa?
Algunas personas consideran que los poderes pblicos deben perseguir la mxima eficacia en la
lucha contra la delincuencia, incluso si ello supusiera recortar derechos y libertades individuales
(estas personas se situaran en la posicin 0 de la escala). Otras, en cambio, creen que es preciso
dar prioridad a los derechos y libertades individuales, incluso si ello supusiera una menor eficacia en
la lucha contra la delincuencia (estas personas se situaran en la posicin 10). Por ltimo, algunas
personas se sitan en posiciones intermedias. En qu posicin se situara Vd.?
Hablando ahora del medio ambiente, si, en una escala, 0 representa que la proteccin del medio
ambiente debe ser la prioridad, incluso si ello supone que la economa crezca menos, y 10 repre-
senta que el crecimiento econmico debe ser la prioridad, incluso si ello supone una menor protec-
cin del medio ambiente, en qu posicin se colocara Vd.?
Notas: 1La cursiva indica que hay ms de una subpregunta. 2Las categoras de respuesta originales (Ayuntamiento, Dipu-
tacin, Parlamento vasco, Gobierno vasco, Parlamento espaol, Gobierno espaol, instituciones europeas) se han reco-
dificado en Pas Vasco, nivel local/europeo, Estado espaol.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, Abril - Junio 2014, pp. 47-68
doi:10.5477/cis/reis.146.47
Citation
Leonisio, Rafael and Strijbis, Oliver (2014). Beyond self-placement: Why Nationalism is a Better
Predictor of Electoral Behaviour in the Basque Country. Revista Espaola de Investigaciones
Sociolgicas, 146: 47-68.
(http://dx.doi.org/10.5477/cis/reis.146.47)
* Rafael Leonisio thanks the Basque Government for the financing through the Grupo consolidado IT-610-3. Both authors
thank Edurne Bartolom for providing them European Values Study (1999).
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
48 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 49
The line of argument in this paper is de- we will anchor these interpretations by rela-
veloped in three stages. In the first section, ting them to the political history of cleavage
we will induce various social divisions in con- formation in the Basque country.
temporary Basque society using survey data, In order to induce the most relevant is-
and interpret them in the light of Basque po- sue dimensions in the Basque Country, we
litical history. In the second chapter, we will used the Basque version of the European
relate these social divisions to self-percep- Value Survey. Conducted in 1999, it con-
tion on the left-right dimension (this enables tains a number of questions uniquely suited
us to prove whether left-right, as ordinary to our purposes, since they concern very
citizens understand it, can be considered a different political issues. In order to identify
super issue that encompasses a number of the latent issue dimensions in Basque con-
different social divisions). In the third chapter, temporary society, the questions on policy
we will show that, once the left-right dimen- preferences were extracted. These contain
sion has been reduced to the classic unders- questions on preferences in terms of the
tanding of state intervention versus econo- protection of the environment, the role of
mic liberalism, it is clearly less important for the church in politics, the generosity of the
voting decisions than the Basque-Spanish welfare state, the value of competition, who
division. Furthermore, we will show that, should have the power in the state, and the
even when the left-right dimension might be institutions in which power should be con-
associated with other issues such as conser- centrated (for the exact wording of the
vatism vs. liberalism, the Basque-Spanish questions, please see the Appendix). Unfor-
division remains markedly more salient for tunately, there were no questions that used
electoral choices (we do this by running mul- an ordinal scale to address the nationalism
tinomial logistic regressions on party votes). conflict. So for this dimension, we recoded
Finally, we conclude by relating our results to two questions on devolution and centralisa-
actual issues in the spatial theory of electoral tion into ordinal categories, with 1 meaning
behaviour. that power should be given to/taken from
the Basque Country, 2 that it should be gi-
ven to the European /local authorities, and
Social divisions in the Basque 3 to the Spanish state (see also the Appen-
Country dix). In order to make sure that these reco-
In order to understand the substantial ded questions accurately capture preferen-
meaning of left and right in the Basque ces regarding the Basque vs. Spanish
Country, it is necessary to link these omni- nationalism dimension, we added a ques-
present labels to important political divisions tion on national identity. If the preferences
within the society. As a starting point, there- regarding devolution vs. decentralisation, as
fore, it is helpful to present a broad picture of we coded them, correlated strongly with the
the issue dimensions along which Basque question on national identity, we could then
citizens can be distinguished. In order to do be sure that we were indeed measuring pre-
so, we proceed inductively: applying factor ferences on the nationalist dimension.
analysis to survey question responses for a We ran a varimax principal component
broad range of issue preferences a well- factor analyses with orthogonal rotation.
established technique in electoral studies The typically-made assumption that ordinal
(e.g. Benoit and Laver, 2005; Kriesi et al., scales can be used as ratio scales was
2008). To begin, then, we must first describe applied. Table 1 presents the results of the
the results of the factor analyses and label factor analysis. It shows that the voters in
them by ad hoc interpretations. Thereafter, the Basque Country can be distinguished in
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
50 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 51
and rural towns pitted against liberalism and nal society. Thus, the Carlist insurrection
the city (Montero, 1998). was a tragic struggle between Basque ur-
While the Carlists received significant ban liberals and traditionalist peasants
support in the Basque Country, it was far (Molina 2010: 244).
from unified (De la Granja 2002: 492). The Following the Carlist defeat, the division
religious conflict was superimposed onto between Basque political forces continued to
the existing rural-urban conflict (Montero, be the dividing line between the Carlist tradi-
1998) and, in general, the majority of the ci- tionalists and the liberals, although the ad-
ties positioned themselves in favour of libe- vent of industrialisation would completely
ralism, whereas the rural towns offered their change the Basque political system, not only
support to those who defended the traditio- with the rise of labour parties and the emer-
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
52 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
gence of an economic incarnation of the left- were thus in favour of Christian reform. On
right cleavage, but also with the advent of a national issues, PNV were completely sepa-
new economic system, which brought about rate from the other two blocks, since they
the emergence of Basque nationalism. The defended Basque nation building (De la
vast majority of labourers came from other Granja, 1995).
regions of Spain, and thus ethnic distinctions Francos dictatorship aborted all political
compounded existing differences of class pluralism within Spanish society. But even 40
and values. Subsequently, at some point du- years of authoritarian rule could not hinder
ring the 19th and 20th centuries, a pluralistic the survival of these socio-political divisions.
political system was formed, which De la As in other Western European societies, they
Granja (2002) called the Basque political have persisted until today, and subsequently,
triangle7, since it can be said to have three they can be linked by using the factor analy-
primary cleavages: the religious issue, the sis we have developed here. Following this
social issue, and the nationalist issue. This interpretation, the 19th century religious clea-
triangle explains the characteristics of the vage is related to the first factor of our analy-
left-leaning, secular (or even anticlerical), sis, which we have called the religious-secu-
pro-Spanish and republican: strong in the lar dimension. And the fourth and fifth factors,
cities, and in favour of social achievements, by contrast, refer to the economic left-right
both in its labour version (PSOE - Spanish cleavage and centre-periphery cleavages,
Socialist Workers Party) and later communist which arose at the end of the 19th century
incarnation (PCE - Spanish Communist Par- with the industrialisation process.
ty), as well as its bourgeois versions (the re- Political liberalism might be seen most
publican parties). In contrast, we can identify clearly, perhaps, against the background of
the archetypical pro-Spanish right: monar- the authoritarian past under Franco. Although
chists, loyal to the Spanish nation, clerical it is true that today there is broad consensus
(conservatives and Carlists) and sometimes regarding the merits of democracy, society
a bit more secular (liberals). These rightists remains divided on the issue of amnesty for
typically represented the interests of the big Francos followers and the role of the monar-
capitalist industrialists that emerged with the chy during the transition to democracy. On
economic development (the dynastic par- the one hand, political liberalism might mainly
ties), or alternatively, the interests of the rural be seen as a being related to the memory of
population (Carlists). Finally, there is Basque the dictatorship and the transition process.
nationalism, represented by PNV (Basque On the other hand, in the context of the Bas-
Nationalist Party). Catholic fundamentalist in que country, it could viewed as being related
the early stages, they drifted towards Chris- to policies against terrorism, such as the si-
tian democracy in the 1930s. With respect to tuation of the imprisoned members of ETA,
social issues, the Basque Nationalist Party the dialogue between the Government and
occupied a position near the centre, and the terrorist organisation, or, in more general
terms, the discussion about how to end te-
rrorism in the Basque Country by means of
7 Initially,
the triangle took place only in Bilbao and the dialogue and negotiation or by making use of
surrounding areas, where industrialisation first took the law and order motto.
place. It went on to extend throughout the Basque
Country, and only during the Second Republic (the By contrast, the last factor to be discus-
1930s) was it characteristic of the entire Basque Country. sed is related to a part of modern Basque
Basque pluralism continued to spread along with history that coincides with a more general
industrialisation, with the conflict between Carlists
(tradition) and liberals (modernity) persisting in the more phenomenon in industrialised societies, and
rural areas. can be described by using Ingleharts theory
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 53
of value change (Inglehart, 1977, 1990). Ac- as a super issue (Knutsen, 1995; Van der
cording to this theory, existential security Eijk et al., 2005: 166), aggregating preferen-
changes a societys value priorities from ma- ces on various issue dimensions.
terialist ones, emphasising economic and For many years, political scientists have
physical security, to post-materialist ones, deduced the ideological positions of voters
emphasising autonomy and self-expression by asking them to provide self-placements
(Inglehart, 2008: 130). Regarding this well- on ideological scales. Clearly, the most im-
documented general trend in Western demo- portant scale in this regard has been the left-
cracies, it is no surprise that the rapidly right scale. Since Downs (1957) spatial mo-
growing levels of wealth in recent decades del of electoral competition, the left-right
have also brought about the advent of new scale has been used extensively to model
generations with increasingly strong post- electoral behaviour, strategic party beha-
materialist values in the Basque Country. It is viour, and the like. The left-right model as
in this light that we interpret the economic such, however, has at least two very serious
growth vs. ecology dimension. Although no drawbacks (for a more complete discussion,
green party has ever achieved success in the see Mair, 2007). First, left and right do not
Basque Country, IU (United Left) has tried to necessarily describe the classic economic
gain the vote of those supporting environ- distinctions between state intervention and
mental values or, more generally, those with economic liberalism. Instead, they often en-
a post-materialist view of political life. In compass a broad set of other oppositions,
2004 IU even changed its name by adding turning the left-right scale into a super issue
berdeak (the greens), transforming itself into (Knutsen, 1995; Van der Eijk et al. 2005: 166).
Ezker Batua Berdeak (United Left The Second, the meaning of the left-right scale
Greens). varies according to the context. In fact, the
By interpreting the corresponding issue notion of a left and right is not even restric-
dimension in the light of Basque history, we ted to political issues; it might also refer to
have tried to avoid ad hoc interpretations or values and partisan loyalties, as well as to
the assumption that the data speak for them- social groups such as classes (e.g. Knutsen,
selves. It is clear, however, that our interpre- 1998; Freire, 2006; Milic, 2008). What left
tation remains open to debate. and right mean to citizens varies over time
within a country, and probably even more
between countries. Even at a given point in
The meaning of left and
time, the meanings of left and right might
right in the Basque Country
vary considerably between citizens, depen-
The results of the factor analysis yield a much ding on partisan leanings and political so-
more complex picture of the ideological spa- phistication (Zechmeister, 2006). To sum-
ce in the Basque Country, in terms of left- marise, the left-right dimension may have
right and Basque nationalism-Spanish natio- different meanings at different points in time,
nalism divisions, than the common and even at the same point in time among
description does. This raises the question of voters. For the debate over whether the left-
how these issue dimensions are related to right dimension or the nationalist dimension
these two more commonly used concepts. is more important for voters decisions in the
With regard to our research question, we Basque country, it is therefore crucial to un-
were particularly interested in finding whether derstand what left and right mean in this
the left-right cleavage refers to the classic particular context.
distinction between state intervention and One way of accomplishing this is by pre-
economic liberalism, or whether it functions dicting voters self-positioning on the left-
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
54 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
right scale by using the positions of voters on as a whole, nor to make causal statements
the various issue dimensions. We created about the relationship between left and
these issue dimensions for the 1999 Basque right and issue preferences. More modestly,
Version of the European Value Survey, by our aim is to show that, while left-right self-
using the factor analysis described in the pre- placement is not a particularly good measure
vious section (see Table 1). In addition, we for preferences on state intervention and
can conduct a similar analysis by making use economic liberalism, it is nevertheless related
of the CIS study on Spatial Models of Politi- to preferences on various issue dimensions.
cal Preferences from 2009 (CIS 2799). While As Table 2 shows for the Basque Euro-
this data is less useful for the inducement of pean Value Survey data, our hypothesis that
issue dimensions8, it is well designed for tes- left-right also works as a super issue in the
ting the relationships between issue dimen- Basque Country receives empirical support.
sions and the left-right scale. The CIS survey Although the classic distinction between sta-
asked respondents to place themselves on te intervention and economic liberalism is sig-
scales which represent secular vs. religious, nificantly related to self-placement on the left-
welfare state vs. low taxes, pro-immigration right scale (when other issue dimensions are
vs. anti-immigration, law and order vs. civic controlled), it is clearly not the only correla-
liberties, and ecology vs. economic growth tion. Additionally, preferences on issue di-
dimension. Additionally, it asked about res- mensions that are usually accounted for as
pondents preferences regarding the seces- liberal vs. conservative (e.g. Conover and
sion of an autonomous community (regional Feldmann, 1981), libertarian-cosmopolitan
territory) vs. the integrity of the Spanish state. vs. authoritarian-particularist (Kitschelt, 1992),
Since we have only selected the respondents materialist vs. post-materialist (Inglehart,
from the Basque Country, we interpret this as 1990), and Green/alternative/libertarian (gal)
referring to the Basque vs. Spanish nationa- to traditional/authoritarian/nationalist (tan)
lism scale. For all questions, including a (Hooghe et al., 2002) are statistically signifi-
question about left-right self-placement, ele- cantly related to left and right. That is to
ven point scales were used, making the re- say, more secularist, ecologist, and politically
sults highly comparable. liberal Basque citizens tend to place themsel-
In order to predict self-placement on the ves further left on the scale than their coun-
left-right scale, we ran OLS regressions with terparts. Furthermore, citizens preferences
the values on the issue dimensions (i.e. the on the Spanish vs. Basque nationalism di-
factor scores in the case of the 1999 data) as mension were also related to self-placement
independent variables, and left-right self- on the left-right scale: Basque nationalists
placement as dependent variables.9 Our goal tended to place themselves further to the left
is neither to explain left-right self-placement than Spanish nationalists. While controlling
for other issue dimensions, however, the
coefficient is not significant. This suggests
8 Thequestions tended to relate to rather general issue that the correlation between nationalist prefe-
dimensions, which means that the designers of the
survey deduced what might be considered potentially rences and left-right placement is either very
relevant issue dimensions. It is supportive of our inductive small or not causally related.
analysis (and the deductions from the survey designers)
that we end up with very similar dimensions. Table 3 shows the same analysis for the
9 Since our independent variables consist of factors from 2009 data. We found results very similar to
the same factor analysis, we are (unsurprisingly) spared those obtained from the 1999 data: left
the problems of multicollinearity (highest correlation self-placement is significantly related to a
among independent variables: 0.186; p<0.01). We also
checked for normal distribution of residuals (mean: preference for a secular state and civic libe-
7.93E-15; sd: 0.992). ralism. We also found that Basque nationa-
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 55
lists placed themselves further to the left one vis-a-vis voters electoral choices. In the
than Spanish nationalists did. While the as- next section, therefore, a new attempt is pre-
sociation is not very strong an increase on sented to empirically investigate whether the
the nationalist dimension (towards Basque economic left-right dimension or the Bas-
nationalism) corresponds to an average de- que-Spanish dimension is more important in
crease of 0.1 points along the left-right di- Basque elections.
mension it appears to be significant. Unlike
the results obtained from the 1999 data,
however, we did not find a significant rela- What is more important for
tionship between preferences on the econo- elections: the economic left-
mic liberalism vs. state intervention dimen- right or the Basque-Spanish
sion. divide?
To summarise our argument so far: we In order to compare the importance of the
have shown that left and right in the Bas- economic left-right and the Basque-Spanish
que Country constitutes a super issue, for divisions during elections, we predicted par-
which it is difficult to provide a single, conci- ty choices by using voters positions on the
se interpretation. And it is clear from our two corresponding dimensions. The availa-
analysis of the empirical relationship bet- ble data allowed us to do so for the two elec-
ween self-placement on the left-right and the tions to the Basque Parliament in 2009, and
Basque-Spanish scale that we cannot con- also permitted making a prediction on voting
clude that voters preferences on issues con- intention for the national election in 2000. In
cerning left and right or Basque and Spanish 1999 there were seven political forces in the
nationalism are more or less relevant in a Basque Parliament. On the nationalist front
substantive way. The results of Fernndez there was the Basque Nationalist Party
Albertos (2002) and De la Calle (2005) might (PNV), a centrist party, as we noted above;
therefore be understood as evidence that, in Basque Solidarity (EA), a pro-independence
the Basque Country, the left-right dimension and social democratic party; and We the
provides a highly efficient heuristic for voters Basques (EH), an extremist Basque nationa-
to simplify a complex political space. This list and leftist party. On the other side of the
does not mean, however, that the Basque- spectrum there were four parties in the non-
Spanish dimension is not the most important nationalist field, representing the branches of
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
56 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Adjusted R2 0.286
N 526
Source: CIS 2799 (2009); Note: *p<0.1, **p<0.05, ***p<0.01; unstandardised Beta-coefficients; contains a constant
major political parties present throughout all used the values on these dimensions to pre-
of Spain: the Spanish Socialist Workers Par- dict voters choices. In order to compare the
ty (PSOE), social democratic in nature; the relevance of actors preferences along the
Peoples Alliance (AP), conservative; United different issue dimensions for electoral choi-
Left (IU), representing the new left; and fina- ces, we ran multinomial logistic (MNL) re-
lly, Union of lava (UA), a small party that gressions. Because of the probable violation
defended the separation of lava (a Basque of the independence of irrelevant alternatives
province) from the Autonomous Community hypothesis, classic statistical guides are of-
of the Basque Country. ten eager to propose multinomial probit
Ten years later the Basque party system (MNP) (e.g. Kennedy 2008: 244-245). But as
had changed only superficially. UA had di- Dow and Endersby (2004) have shown,
sappeared and a centralist and centrist party within the limits of normal data (i.e., with a
had emerged: UPyD (Union Progress and small sample and few parties to choose
Democracy). The self-nominated Basque Pa- from), MNP might be susceptible to a num-
triotic Left had split into two branches, one ber of estimation problems, while the viola-
that condemned ETAs terrorism, Aralar, and tion of the independence of irrelevant alter-
one that did not, Batasuna (Unity). The latter natives might be neither particularly
was banned from the 2009 Basque elections restrictive nor relevant. This is especially true,
but asked to have its vote recorded as a null according to their argument, if electoral choi-
vote. Therefore, we can consider that as a ces are relatively fixed and stable (Dow and
proxy for its support of Batasuna. PNV, EA, Endersby, 2004: 112), which is certainly the
PSE-EE (Basque equivalent of PSOE), PP case in the Basque Country (and is even
(formerly, AP) and IU remained in the Basque more true here than it was in the cases to
Parliament. which Dow and Endersby successfully
applied MNL).10
With the above explanation completed,
we can once again make use of the data re-
garding the different issue dimensions. As 10All models were also calculated as a multinomial probit
before, instead of attempting to predict self- regression. We could not find any substantial deviation
placement along the left-right dimension, we from the logit models. Furthermore, multicollinearity can
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 57
Our analyses were run both with and rather heterogeneous nature of the EH coali-
without a set of control variables. The models tion at that time, which resulted from ETAs
without control variables cannot exclude the truce. In that election, HB ran for election to-
omission of third variables. This is only pro- gether with other small leftist parties in the
blematic, however, when it comes to omitted EH coalition. Furthermore, it gathered bo-
variables that have an effect on the coeffi- rrowed votes from voters who wanted to
cients of our independent variables, since strengthen the EH coalition in order to con-
our goal is not to explain a maximum varia- solidate the peace process. Another expla-
tion in party choice, but to compare the nation for the reduced importance of the
effects of our two independent variables. Basque-Spanish dimension for EH vote rela-
This also means that the inclusion of control tive to PNV vote might further be ascribed to
variables would only be adequate in cases the position of PNV, which was closer to the
where such variables were not causally prior more radical nationalists than it was pre-
to the formation of issue preferences. viously, due to the alignment of nationalist
Table 4 displays the MLN models for vote political forces after the signing of the Lizarra
intentions for the 2000 general election. The Agreement in 1998.
values indicate their impact on the probabili- In contrast to the Basque-Spanish issue
ty of a partys vote, relative to the vote for the dimension, the economic left-right dimen-
Basque Nationalist Party (PNV), which (be- sion had noticeably less importance for elec-
cause of its central place in the party system, toral choices in 2000. This can be seen in the
vis-a-vis its ideology and magnitude) has fact that the Nagelkerke R2 is three times
been chosen as the reference category. As smaller in the second model than in the first.
the results of model 1 show, the Basque na- To the degree that they are related to vote
tionalism dimension is strongly related to the choice, these relationships are not surpri-
probability of voting for the Spanish Conser- sing: those voting for EH, EA and IU are all
vatives (PP) or for the Spanish Socialists more in favour of state intervention than PNV
(PSOE). We might interpret this result as fo- voters. This picture does not change when
llows: the less Basque nationalist-leaning a the reference category is changed. Taking,
voter is, the more likely it is that they voted for example, the Socialist party as a referen-
for PP or PSOE relative to PNV. The voters ce category (data not shown), we find that
position along the Basque-Spanish nationa- increasingly pro-Basque preferences impro-
lism dimension, by contrast, has no signifi- ve the probability of voting for one of the
cant effect on the probability of them voting Basque nationalist parties (PNV, HB/EH, EA),
for EH or EA instead of PNV. In the latter case and that lower values increase an actors
this is hardly surprising, considering the very probability for voting for the Spanish conser-
similar position of EA on the Basque nationa- vatives (PP), while voters position on the
list dimension (Llera, 1988). An explanation economic left-right dimension has no effect
for the increased importance of the econo- on the probability that they will vote for a par-
mic left-right dimension vis-a-vis the choice ty other than the Socialists.
between EH and PNV might be found in the In models 3 and 4 the other issue dimen-
sions were introduced. Model 3 is only inter-
esting in comparison to model 1, since it
be rejected. We did not include control variables, because allows us to form an understanding of the
our goal is not to explain a maximum variation in party explanatory power of the nationalist dimen-
choice, but to compare the effects of our independent sion relative to all the other dimensions com-
variables. In such a case, only the inclusion of control
variables which are not causally prior to the formation of bined. The Nagelkerke R2 of around 0.05
issue preferences would be adequate. shows that in 1999, the nationalist dimension
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
58 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Table 4. Multinomial logit regression models for vote intention for the 2000 Spanish general election
Notes: *p<0.1, **p<0.05, ***p<0.01. Results for Greens, blank votes and abstention not shown. Models include constants.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 59
alone comes relatively close in its explana- choice for all dimensions. And it is interesting
tory power to all of the other issue dimen- to note that Aralar and IU voters distinguis-
sions combined (there, the R2 is 0.08). Model hed themselves from PNV voters substantially
4 displays the relationship between each is- by their more libertarian preferences, while
sue dimension and vote choice while con- there is no sign that secularism still plays an
trolling for all other issue dimensions. The important role. However, this could also be
relationship between nationalism and vote the case because Batasuna was not allowed
choice interpreted above is not challenged in to participate in this electionand hence, the
any substantive way, and the same is true for most clearly distinguishable voters may not
the economic dimension. We also find, have been concentrated on one party in this
however, that the intention to vote for Euskal election.
Herritarrok (instead of PNV) is related to To summarise: we found that the Basque-
strong secularist and anti-authoritarian pre- Spanish division is generally more important
ferences. Hence, this model reveals that the to voters electoral choices in the Basque
ideological distance between voters who Country than the economic left-right dimen-
supported EH and all others is very large, in- sion. The economic left-right dimension was
dependent of whether they were Basque na- of some relevance in 1999, within the bloc of
tionalists or not. Basque nationalist parties, suggesting that,
We found similar results for voting inten- in the case of a similar position on the issue
tion relative to the 2009 regional elections concerning Basque autonomy and indepen-
(see Table 5). Again, we found a significant dence, the economic left-right dimension
difference in the preferences of PP and PSOE might be of importance when actors were
voters compared to PNV voters on the natio- choosing between different nationalist par-
nalism dimension. In this case, a statistically ties. Furthermore, in the 2009 election PSE-
significant relationship was identified bet- EE voters were significantly more pro-state
ween voters likelihood of supporting PSE- intervention than PNV-voters. Overall, howe-
EE instead of PNV, and their position on the ver, there can be no doubt that issue prefe-
economic left-right dimension. However, the rences on the nationalist dimension predict
distance between the two groups of voters is vote choice much more accurately than pre-
clearly smaller here than it is on the nationa- ferences on economic issues. The fact that
list dimension. In contrast to the elections in we found the same basic results for two di-
2000, we found that the economic dimension fferent elections, using very different measu-
was not related to voters likelihood of voting rements for the independent and dependent
for the leftist Basque nationalist parties (EA variables (in one case, vote intention for a
and Aralar) instead of voting for PNV. And national election and in the other case, vote
once again, we find no significant results on recall for a regional election), increases our
the nationalism dimension. confidence in the claim that the electoral de-
Our comparison of models 1 and 2 makes cisions of voters in the Basque Country are
it clear that nationalism can explain more driven more by preferences regarding the
than five times as much variance in vote Basque-Spanish dimension than the econo-
choice in 2009 than the economic dimension mic left-right one.
can. Furthermore, the comparison between Additionally, our analysis has shown that
models 1 and 3 shows that in 2009, the na- the nationalist dimension is far more relevant
tionalist dimension was more important for to electoral choices than any other issue di-
vote choice than all the other dimensions mensions. This means that, even when left
combined. Model 4 reiterates the relations- and right are not understood in economic
hips between issue preferences and vote terms, but (for example) in cultural terms, dis-
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
60 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Table 5. Multinomial logit regression models for vote recall for the 2009 election to the Basque Parliament
Dependent variable:
Independent variables Model 1 Model 2 Model 3 Model 4
vote
Spanish-Basque nationalism -0.774*** -0.732***
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 61
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
62 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Ansolabehere, Stephen and Puy Segura, Mara Socorro Granja, Jos Luis de la (1995). El nacionalismo vasco:
(2013). Ideology, Nationalism, and Identity in Basque un siglo de historia. Madrid: Tecnos.
Regional Elections. Working Paper no publicada.
(2002). Paz entre dos guerras civiles. In: Bazan,
Bartolini, Stefano (2000). The Political Mobilisation of I. (ed.). De Tbal a Aitor. Historia de Vasconia. Ma-
the European Left, 1860-1980: The Class Clea- drid: La Esfera de los Libros.
vage. Cambridge: Cambridge University Press.
Gunther, Richard; Sani, Giacomo and Shabad, Goldie
Benoit, Kenneth and Laver, Michael (2005). Mapping (1986). El sistema de partidos en Espaa. Gne-
the Irish Policy Space: Voter and Party Spaces sis y evolucin. Madrid: CIS.
in Preferential Elections. Economic and Social
Hooghe, Liesbet; Marks, Gary and Wilson, Carol J.
Review, 36 (2): 83-108.
(2002). Does Left/Right Structure Party Positions
Budge, Ian et al. (eds.) (2001). Mapping Policy Pre- on European Integration?. Comparative Political
ferences: Estimates for Parties, Electors and Go- Studies, 35 (8): 965-989.
vernments 1945-1998. Oxford: Oxford University
Inglehart, Ronald (1977). The Silent Revolution: Chan-
Press.
ging Values and Political Styles among Western
Calle, Luis de la (2005). Cuando la proximidad deja Publics. Princeton: Princeton University Press.
de ser importante. Revista Espaola de Ciencia
(1990). Culture Shift in Advanced Industrial So-
Poltica, 12: 21-52.
cieties. Princeton: Princeton University Press.
Conover, Pamela J. and Feldman, Stanley (1981).
(2008). Changing Values among Western Publics
The Origins and Meaning of Liberal/Conservati-
ve Self-Identifications. American Journal of Po- from 1970 to 2006. West European Politics, 31
litical Science, 25 (4): 617-645. (1): 130-146.
Dinas, Elias (2012). Left and Right in the Basque Kennedy, Peter (2008). A Guide to Econometrics (6
Country and Catalonia: The Meaning of Ideology ed.). Oxford: Blackwell.
in a Nationalist Context. South European Soci- Kitschelt, Herbert (1992). The Formation of Party
ety and Politics, 17(3): 467-485. Systems in East Central Europe. Politics and
Dow, Jay K. (2001). A Comparative Spatial Analysis Society, 20 (1): 7-50.
of Majoritarian and Proportional Elections. Elec- Klingemann, Hans-Dieter et al. (2006). Mapping Po-
toral Studies, 20 (1): 109-125. licy Preferences II: Estimates for Parties, Electors
and Endersby, James W. (2004). Multinomial and Governments in the OECD, EU and CEE
Probit and Multinomial Logit: A Comparison of 1990-2003. Oxford: Oxford University Press.
Choice Models for Voting Research. Electoral Knutsen, Oddbjrn(1995).Value Orientations, Poli-
Studies, 23 (1): 107-122. tical Conflicts and Left-right Identification: A
Downs, Anthony (1957). An Economic Theory of De- Comparative Study. European Journal of Politi-
mocracy. New York: Harper and Row. cal Research, 28 (1):63-93.
Feire, Andr (2006). Bringing Social Identities Back (1998). The Strength of Partisan Competition of
In: The Social Anchors of Left-Right Orientation Left-Right Identity: A Comparative Longitudinal
in Western Europe. International Political Scien- Study of Left-Right Party Polarization in Eight West
ce Review, 27 (4): 359-378. European Countries. Party Politics, 4 (1): 5-31.
Fernndez Albertos, Jos (2002). Votar en dos di- Kriesi, Hanspeter et al. (2008). West European Politics
mensiones: el peso del nacionalismo y la ideo- in the Age of Globalization. Cambridge: Cambrid-
loga en el comportamiento electoral vasco. ge University Press.
1993-2001. Revista Espaola de Ciencia Polti- Leonisio, Rafael (2012). Parliament on the Center-
ca, 6: 153-181. right, Government on the Left: Explaining the
Garca de Cortzar, Fernando and Lorenzo, Jos Ma- Basque Exceptionalism. Regional and Federal
ra (1997). Historia del Pas Vasco. San Sebastin: Studies, 22 (1): 45-60.
Txertoa. and Strijbis, Oliver (2011). Izquierda-Derecha vs
Gillespie, Richard (2000). Political Polarization in the Centro-Periferia: una aproximacin al discurso de
Basque Country. Regional & Federal Studies, 10 los partidos polticos vascos (1977-2009). Re-
(1): 112-124. vista Espaola de Ciencia Poltica, 26: 63-85.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 63
Linz, Juan Jos et al. (1986). Conflicto en Euskadi. Molina, Fernando (2010). The Historical Dynamics
Madrid: Espasa-Calpe. of Ethnic Conflicts: Confrontational Nationalisms,
Democracy and the Basques in Contemporary
Lipset, Seymur M. (2001). Cleavages, Parties and
Spain. Nations and Nationalism, 16 (2): 240-260.
Democracy. In: Karvonen, L. and Kuhnle, S.
(eds.). Party Systems and Voter Alignments Re- Montero, Manuel (1998). Historia del Pas Vasco. San
visited. London: Routledge. Sebastin: Txertoa.
and Rokkan, Stein (1967). Cleavage structures, Pallars, Francesc; Jordi Muoz, Jordi and Alfredo
party system, and voter alignments: An introduc- Retortillo, Alfredo (2006). Depolarization in the
tion. In: Lipset, S. M. and Rokkan, S. (eds). 2005 Autonomous Elections in the Basque Coun-
Party systems and voter alignments. New York: try: Towards a New Scenario for Peace?. Regio-
Free Press. nal and Federal Studies, 16 (4): 465-479.
Llera, Francisco Jos (1986). Las segundas elec- Prez-Nievas, Santiago and Bonet, Eduardo (2006).
ciones autonmicas vascas. Revista de Derecho Identidades regionales y reivindicacin de auto-
Poltico, 23: 135-165. gobierno. El etnorregionalismo en el voto a partidos
nacionalistas de Blgica, Espaa y Reino Unido.
(1988). Continuidad y cambio en el sistema de
Revista Espaola de Ciencia Poltica, 15: 123-161.
partidos vascos: 1977-1987. Revista de Estu-
dios Polticos, 59: 277-375. Sani, Giacomo and Sartori, Giovanni (2006). Polari-
zacin, fragmentacin y competicin en las de-
(1989). Continuidad y cambio en la poltica
mocracias occidentales. In: Sartori, G. Partidos
vasca: notas sobre identidades sociales y cultu-
y sistemas de partidos. Madrid: Alianza.
ra poltica. Revista Espaola de Investigaciones
Sociolgicas, 47: 107-135. Stoll, Heather (2010). Elite-level Conflict Salience
and Dimensionality in Western Europe: Concepts
(1999). Basque Polarization: between Autonomy
and Empirical Findings. West European Politics,
and Independence. Nationalism and Ethnic Po-
33 (3): 445-473.
litics, 5 (3-4): 101-120.
Strijbis, Oliver and Leonisio, Rafael (2012). Political
(2000). Los gobiernos de coalicin en el Pas
Cleavages in the Basque Country: Meaning and
Vasco. In: Matas, J. (ed.). Coaliciones polticas
Salience. Regional and Federal Studies, 22 (5):
y gobernabilidad. Barcelona: ICPS.
595-611.
(2001). La encrucijada vasca. Revista de Oc-
Torcal, Mariano and Chhibber, Pradeep (1995).
cidente, 241: 89-113.
Elites, cleavages y sistema de partidos en una
Mair, Peter (2007). LeftRight Orientations. In: Dal- democracia consolidada: Espaa (1986-1992).
ton, R. J. and Klingemann, H.-D. (eds.). The Revista Espaola de Investigaciones Sociolgi-
Oxford Handbook of Political Behaviour. Oxford: cas, 69: 7-38.
Oxford University Press.
Van der Eijk, Cees; Schmitt, Hermann and Binder,
Mata, Jos Manuel (1998). Nationalism and Political Tanja (2005). Left-right Orientations and Party
Parties in the Autonomous Community of the Choice. In: Thomasen, J. (ed.). The European
Basque Country (ACBC). Strategies and Tensions. Voter: A Comparative Study of Modern Demo-
Working Paper 137. Barcelona: ICPS. cracies. Oxford: Oxford University Press.
Milic, Thomas (2008). Links, Mitte, Rechts: Worauf Zechmeister, Elizabeth (2006). Whats Left and Whos
grndet die ideologische Selbstidentifikation der Right? A Q-method Study of Individual and Con-
Schweizer und Schweizerinnen?. Swiss Political textual Influences on the Meaning of Ideological
Science Review, 14 (2): 245-285. Labels. Political Behavior, 28 (2): 151-173.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
Rafael Leonisio and Oliver Strijbis 65
Appendix
Table A: Original survey questions (translated)
For each statement read out, can you tell me whether you agree strongly, agree, disagree or strongly disa-
gree?1
I would agree to an increase in taxes if the extra money were used to prevent environmental pollution
The Government has to reduce environmental pollution but it should not cost me any money
To what extent do you agree or disagree with each of the following (Agree strongly, Agree, Neither agree nor
disagree, Disagree, Disagree strongly)?
Politicians who do not believe in God are unfit for public office
Religious leaders should not influence how people vote in elections
It would be better if more people with strong religious beliefs held public office
How would you place your views on this scale (1-10 point scale)?
Individuals should take more responsibility for providing for themselves vs. The state should take more
responsibility to ensure that everyone is provided for
People who are unemployed should be required to take any job available or lose their unemployment ben-
efits vs. People who are unemployed should have the right to refuse a job that they do not want
Competition is good. It stimulates people to work hard and develop new ideas vs. Competition is harmful,
it brings out the worst in people
The state should give more freedom to companies vs. The state should control companies more effec-
tively
For each of the following statements, would you say it is a very good, fairly good, fairly bad or very bad way of
governing this country?
Having a strong leader who does not have to bother with parliament and elections
Having experts, not government make decisions according to what they think is best for the country
From which institution should political power be taken away (recoded into 1-3 point scale)?2
Which institution should be given more political power (recoded into 1-3 point scale)?2
How would you place your position on this scale, generally speaking (1-10 point scale)?
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68
66 Beyond Self-Placement: Why Nationalism is a Better Predictor of Electoral Behaviour in the Basque Country
Thinking about the presence of the Catholic religion in Spanish politics, if 0 means a completely secular (non-
religious) state and 10 a confessional (religious) state, in which position of the scale would you place yourself
(scale from 0 to 10 shown)?
Some people think that public services and social benefits should be improved, although this would mean hav-
ing to pay more taxes (on a scale from 0 to 10, these people would place themselves in position 0). Others think
that it is more important to pay lower taxes, although this would significantly reduce public services and social
benefits (these people would position themselves at position 10 in the scale). There are others who position
themselves in between. Where would you place yourself?
Let us talk now about immigration. On a scale where point 0 represents free entry of immigrants and point 10,
absolute restriction on the entry of immigrants. In which position would you like to see Spain?
Some people think that the state should pursue maximal efficiency in the fight against crime, even if that cuts
into individual rights and freedom (these people place themselves on point 0 of the scale). Others, in contrast,
believe that it is better to give priority to individual rights and freedoms, even if that means lower efficiency in the
fight against crime (these people place themselves at point 10 on the scale). Finally, some people place them-
selves in between. Where would you place yourself?
Speaking now about the environment, if on a scale 0 represents that the protection of the environment should
have priority, even if this would mean a diminished growth of the economy and 10 represents that economic
growth should have priority, even if this means less environmental protection, in which position would you place
yourself?
If tomorrow an autonomous community of Spain would secede from Spain, to what degree would you like that?
Use the following scale, where 0 means I would not like that at all and 10 I would like that a lot.
When people speak about politics they normally use the expressions left and right. In which position would you
place yourself (scale shown)?
Notes: 1Italics mean that there is more than one sub-question. 2The original answer categories (city council, county council,
Basque parliament, Basque government, Spanish parliament, Spanish government, European institutions) have been recoded
into Basque Country, local/European authorities, Spanish state.
Reis. Rev.Esp.Investig.Sociol. ISSN-L: 0210-5233. N 146, April - June 2014, pp. 47-68