You are on page 1of 20

622 European Journal of Political Research 48: 622–641, 2009

doi: 10.1111/j.1475-6765.2009.00845.x

Between ideology and social representations: Four theses plus


(a new) one on the relevance and the meaning of the political
left and right ejpr_845 622..641

PIERGIORGIO CORBETTA,1 NICOLETTA CAVAZZA2 &


MICHELE ROCCATO3
1
University of Bologna, Italy; 2University of Modena-Reggio Emilia, Italy; 3University of
Torino, Italy

Abstract. This study analyses the relevance and the meaning given by Italians to the politi-
cal labels ‘left’ and ‘right’ between 1975 and 2006. Based on responses to the open-ended
question ‘What do you mean by “left/right” in politics?’, the study compares five alternative
hypotheses on the meaning of the left-right axis and show that, despite the alleged end of
ideologies, the relevance of the axis has increased over time. A core of abstract meanings
persists throughout the thirty-year period considered. As the importance of abstract mean-
ings has increased over time, reference to more concrete contents (such as ‘parties’ and
‘leaders’) has decreased. The findings thus support the hypothesis that the left-right axis has
the functional characteristics of social representations.

The left-right distinction has been at the heart of the European political debate
for more than 200 years, since the participants of the Etats Généraux, gathered
together as a constituent assembly on 28 August 1789 at Versailles, divided
themselves spatially on the issue of the king’s veto power. Those in favour
positioned themselves to the right of the president, and those in opposition to
his left. Since then, the left-right distinction has gained considerable meaning,
representing over the course of the last two centuries in Western (and espe-
cially European) countries the conceptual template that best summarises
political differences (Inglehart & Klingemann 1976).
Nonetheless, much has changed in society and in European politics in
recent years, leading many people to ask if and to what extent this distinction
is still relevant in contemporary political discourse. These questions implicitly
pose two problems, respectively concerning the relevance of the left-right
dichotomy and its content. With regard to both issues, many points of view
have been advanced. After reviewing them, we will attempt to test them
empirically in the Italian context, where many rapid political changes took
place in the last decades, depicting a particularly complex judgemental envi-
ronment for citizens.
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Malden,
MA 02148, USA
between ideology and social representations 623

The relevance of the left-right dichotomy

The starting point of this debate is represented by the provocative statement


on ‘the end of ideology’ put forth by Bell (1965: 402), who argued that: ‘The
ideologies . . . which emerged from the nineteenth century . . . today are
exhausted.’ Several scholars agreed with this thesis (see Aron 1962;
Kornhauser 1960; Lipset 1960; Shils 1968; among others; for a summary of the
debate, see Waxman 1968). The ‘end of ideology’ should have brought about a
substantial attenuation of the harsh conflicts of the earlier period. Such attenu-
ation, in turn, should have produced a sharp decrease in the relevance of the
left-right axis for individuals’ political orientation. Nevertheless, attempts to
validate this prediction empirically have not been very successful. More than
thirty years ago, based on a study on nine European countries, Inglehart and
Klingemann (1976: 243) argued that: ‘The end of ideology did not materialize.
On the contrary, there are indications that presumably ideological concepts
such as left and right are taking new life and new meaning for Western politics.’
Over the last thirty years, empirical research has repeatedly confirmed this
thesis. Mair (2007: 208), for example, recently claimed that the left-right dis-
tinction still ‘appears to offer both sense and shape to an otherwise complex
political reality’ at three levels. First, in terms of voters, data from the Euro-
pean Social Surveys conducted in recent years show that more than 80 per cent
of voters define themselves as left- or right-wingers. Second, regarding observ-
ers and researchers ‘expert surveys’ systematically identify the left-right con-
flict as one of the most relevant issues in the competition among political
parties. Third, with party programmes and electoral platforms, content analysis
systematically shows that ‘some form of left-right dimension dominates com-
petition at the level of the parties’ (Mair 2007: 209–210). Jost (2006), after
reviewing the recent psychological research, reached the same conclusion.
To understand the reasons behind the persistence of the left-right dimen-
sion in Western political culture we need to explore the functions it performs.
At the collective level it serves the purpose of social orientation, functioning as
a symbolic frame of reference used by political groups to place themselves and
communicate their positions.As such, this dimension represents a fundamental
organising principle of the political space, used to define conflicts and com-
monalities among parties and leaders, like the cardinal points on a map allow-
ing one to locate a town as geographically close to or far from another. At the
individual level, the dimension simplifies individual decision making and
judgements in the political sphere. Indeed, socio-cognitive psychology showed
that people cope with complex decision tasks using ‘heuristics’ – that is, simple
decision-making rules, also defined as ‘mental shortcuts’, which are used to
bypass deeper processing of the information (Tversky & Kahneman 1974).
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
624 piergiorgio corbetta, nicoletta cavazza & michele roccato

Heuristics are particularly useful when individuals face complex decision


tasks (such as making judgements and decisions in the political sphere) while
simultaneously lacking the motivation to acquire more accurate information
(often the case for many citizens). Given this scenario, the positioning of
parties, leaders and stances along the left-right axis greatly facilitates voters’
selection processes by reducing the complexity of the political content (Fuchs
& Klingemann 1990; Inglehart & Klingemann 1976).
The function of orienting voters is particularly useful in two circumstances.
First, when the political system is complex. Sartori (1976; see also Inglehart &
Sidjanski 1976), for example, emphasises the indispensability of the one-
dimensional left-right simplification in the case of multiparty systems. Second,
when the political system undergoes a radical transformation the axis may be
conceived as an instrument capable of channeling ‘voters from their former
party to the nearest equivalent among the new ones’ (Inglehart & Klingemann
1976: 271).
To summarise, despite profound changes in socio-political contexts over
time, the two necessities met by the left-right dimension – the ‘individual’ need
to understand, simplify, represent and make decisions, and the ‘social’ need to
communicate, define reciprocal positions and establish trust – have remained
intact.

Content of the left-right dichotomy

Any analysis of the relevance and function of the left-right dimension should
focus on its content and how this has changed over time. From a review of the
scientific literature, four theses emerge concerning the ‘meaning’ voters
attribute to this conceptual dichotomy. The first thesis assigns an intrinsic and
fairly stable meaning to the left-right distinction, independent of space, time
and other factors. The most convincing formulation of this has been advanced
by Bobbio (1996), according to which the most widely used criterion to distin-
guish between left and right is attitude toward equality: left-wingers tend to
consider people more equal than different, while the contrary holds true for
right-wingers.
A second thesis similarly assigns an intrinsic and autonomous content to
the categories of left and right, although its supporters argue that such content
is not set once and for all, but tends to change over time (Fuchs & Klingemann
1990). For example, in European societies the principal conflict is no longer
defined in terms of class cleavages, but new lines of conflict have emerged,
primarily between materialist and post-materialist values (Inglehart 1977,
1990; with regard to other lines of cleavage, see Knutsen 1998). According to
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 625

Fuchs and Klingemann (1990), however, this change in the structure of the
dominant social conflict will not result in the overcoming of the distinction
between left and right, nor will it deprive the terms of their meaning: while the
content of this distinction may change over time, it will continue to carry out
its specific function of simplifying political reality and making sense of it.
These first two theses describe the left-right dimension content as operating
top-down on voters’ opinions. On the contrary, the third and fourth theses do
not attribute an intrinsic meaning to the left-right dichotomy, but rather con-
ceive its meanings to be elaborated by citizens in a bottom-up process. Accord-
ing to the third thesis (put forward by Butler & Stokes 1969), voters consider
left and right simply as ‘conventional labels’ derived from the relative position
of the political parties along the axis and partisan identification should precede
individual self-placement on the left-right scale, often representing its main
component (Inglehart & Klingemann 1976).
The fourth thesis, sustained primarily by Sani (1974) and Sartori (1976), is
an extreme version of the second. Left and right are stripped of their own
autonomous meaning, taking on any meaning depending on the circumstances.
In other words, they are considered empty vessels ready to be filled or
decanted with any and all contents, located along the one-dimensional left-
right spectrum, in response more to the practical needs of order and intelligi-
bility in politics than to any law of internal coherence.They are also in constant
change, depending on the society and the time period (e.g., during the 1980s
issues related to the Cold War were a central concern, while current topics
include attitudes towards immigrants, taxes, crime, reproductive rights, the
family, Europe and economic globalisation).
This study compares the empirical validity of the four theses presented
above, along with a fifth one that we propose building upon them. Our under-
lying idea is that ‘left’ and ‘right’ are categories ensuring both the individual
and the social functions mentioned above, as well as the flexibility of meaning
necessary to adapt to the historical and political context. Borrowing a well-
known concept from social psychology, we argue that the ideas of what left and
right represent in politics today have the characteristics of ‘social representa-
tions’ (Moscovici 1984). As Moscovici (1988: 214) put it, social representations
‘concern the contents of everyday thinking and the stock of ideas that gives
coherence to our religious beliefs, political ideas, etc. and the connections we
create as spontaneously as we breathe’. They are forms of social knowledge:
systems of values, beliefs, opinions, semantic repertoires and theories of
common sense resulting from a process of reconstruction of reality into a
symbolic system elaborated in relation to socially relevant objects through
communicative exchanges between people in groups and communities (Doise
1989). Social representations, therefore, are a sort of map of the semantic field
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
626 piergiorgio corbetta, nicoletta cavazza & michele roccato

relative to an object (in our case, the ‘left’ or the ‘right’ in politics), used by
people to move at ease and in a coordinated way with other individuals. Within
this common field of reference, individuals and social groups adopt different
positions, but they are able to meet on common ground (e.g., everybody has an
idea of what a ‘left-wing party’ is, regardless of his or her personal positive or
negative judgement about it).
Social representations are both stable and flexible. Stability is guaranteed
by a central core of essential meanings, while flexibility is due to their periph-
eral system (Flament 1987). The central core represents the point of reference
for all members of a certain culture. It includes the elements that determine
the overall meaning of the representation. Surrounding the core, we find the
peripheral elements, which are substantially context-dependent and subject to
variation in relation to the social groups expressing the representation. Indeed,
groups occupy different social positions and define the perspective from which
individuals perceive and think about an object. For example, it is quite obvious
that old activists of a political party with a long history develop a representa-
tion of left and right that is somewhat different to that of young students voting
for the first time. Thus, the peripheral components guarantee the evolution of
the representation since they may change in relation to the context and to
transformations of the object, leaving the core mostly untouched.
In the case under analysis, we can thus hypothesise that some intrinsic
meanings compose the central core of the social representations of left and
right. In particular, the most abstract symbolic references are likely to be part
of it – mainly equality, as proposed by Bobbio (1996), but also other distinctive
values. On the contrary, the more concrete and contingent elements (e.g.,
references to specific topics, political parties and political actors) will be con-
tinuously re-examined in relation to the context, the historical phase and the
social groups that express them.
During specific historical phases, people and social groups may experience
difficulties in incorporating concrete references to the social representations of
left and right. This occurs, for example, when political parties traditionally
emblematic of the left and right disappear or experience severe crisis; when
renowned or charismatic leaders leave politics; when traditional international
points of reference (such as the USA–USSR opposition at the end of the last
century) or national institutions (e.g., the direct involvement of the Catholic
Church in the Italian political debate following the collapse of the Christian
Democratic Party) weaken; and, finally, we should also consider – with par-
ticular regard to self-identification with the left – the loss of political identity
and political agency on the part of the working classes in Western society.
In these situations, social representations of left and right tend to be
deprived of concrete references and limited to more abstract content. For
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 627

example, when asked ‘What does the left mean to you?’ the working class
individual of the 1970s likely answered ‘the USSR’ or ‘the Communist Party’.
In the twenty-first century, after the fall of the Berlin Wall and the disappear-
ance of the Communist Party, individuals from the same group have plausibly
shifted their response from the concrete to the abstract, answering ‘less
inequality among people’ or ‘social solidarity’.
This study focuses on the situation in Italy during the approximately thirty-
year span between 1975 and 2006. This time frame is of particular interest since
Italian politics throughout the postwar period, and up until the 1980s, was
always extremely ideological. Italy was the European country with the largest
Communist Party and the largest Catholic Party, and featured strong party
identification, especially among Communist and Christian-Democratic voters.
At the beginning of the 1990s, however, the Italian party system quite suddenly
experienced a process of complete disintegration. In fact, almost all of the
traditional parties that had given stability to the political system for half a
century disappeared from the political scene (they either simply dissolved or
took on new names). The process was triggered by the legal proceedings of
‘Tangentopoli’ (‘Bribetown’), which led to the indictment and prosecution for
corruption of a significant portion of the two main governing parties (Christian
Democrats and Socialist Party). The system’s disintegration was greatly accel-
erated by the fall of the Berlin Wall, which neutralised the ideological foun-
dation of the communism/anti-communism opposition. Moreover, Christian
Democrats, who had governed the country without interruption for the entire
postwar period, experienced a fatal crisis. The party system finally came to a
formal end in 1993 when the electoral law transformed the electoral system
from proportional to majoritarian, generating new political formations and
eliminating old ones. All these events taken together allow us to consider Italy
as a sort of ‘exemplary case’, ideally suited to study the ways in which the
left-right axis changes in relevance and content when party identification
weakens traditional political objects disappear and politics suddenly becomes
more complicated.

Research design and hypotheses

Among the reasons that empirical research has not yet been able to provide
conclusive evidence about the meanings ascribed by voters to the concepts of
left and right, we should include the difficulty of fruitfully exploring the
‘meaning’ attached by individuals to such a complex concept using closed-
ended questions and the fact that empirical studies of change must be carried
out over a long period of time. To circumvent these problems we tried to
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
628 piergiorgio corbetta, nicoletta cavazza & michele roccato

answer the questions outlined above by comparing the results from two
surveys conducted thirty years apart, both of which include the same open-
ended questions about the meaning of left and right. We have formulated the
following hypotheses concerning the two aspects of the left-right dimension
highlighted above: relevance and content.
In terms of the former, we hypothesised that in Italy the relevance of the
left-right dimension has risen over time due to the increasing need for simpli-
fication of the political scene where traditional referential anchors have dis-
appeared (HP1.1). If this is the case, we should observe a significant increase
in the relevance of the left-right dimension among the elderly, the less-
educated individuals and the working class – that is, the least culturally
equipped social groups, those who have lost concrete points of reference (such
as political parties) through which they used to define left and right (HP1.2).
With regard to content, in addition to the theses presented above, we have
formulated five alternative hypotheses differentiated in terms of: autonomy/
non-autonomy of meaning, one-dimensionality/multidimensionality, and
invariability/variability over time.

HP2.1: The left-right dimension has autonomous content, basically one-


dimensional and invariant over time (primarily organised along the
equality/inequality axis).
HP2.2: The left-right dimension has autonomous content, basically one-
dimensional, derived from the society’s principal concerns; it therefore
varies over time, dependent on the characteristics of society.
HP2.3: The left-right dimension has no autonomous meaning; its content
can be rather inferred from the political parties with which it is associ-
ated; it is thus multi-dimensional and relatively stable in so far as the
party system is also stable.
HP2.4: The left-right dimension has no autonomous content; it is an
‘empty vessel’ taking on multi-dimensional and variable meanings
depending on the time period.
HP2.5: The left-right dimension has a central core of autonomous and
abstract contents that do not change over time, and a set of concrete
contents that tend to vary over time according to contextual characteris-
tics and historical circumstances (peripheral elements).

The crisis of the Italian political system in the 1990s should have brought
about a greater need to refer to abstract and stable contents. This should be
especially true among social groups for whom concrete elements in the social
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 629

representations of right and left were particularly salient. Examples of such


social groups include the working class with regard to other social classes (due
to the disappearance of the Communist Party), less educated people with
regard to more educated groups (less equipped to manage a crisis in tradi-
tional points of reference), and elderly people with regard to younger people
(due to the reduced efficacy of concrete points of reference in the context of a
rapid transformation of political parties and historic political blocs).

Method

The core of our study was based on the comparison of two Italian surveys,
conducted thirty years apart, in which the same open-ended questions about
the meaning of ‘left’ and ‘right’ in politics were asked. The first (N = 1,657) was
carried out in 1975 as part of the comparative project Eight nations – Political
action.1 The second (N = 1,377) was conducted by the Itanes2 research group in
2006. In order to explore the changes in the ‘relevance’ of the left-right dimen-
sion, as well as the 1975 and 2006 surveys we also used the following databases:
Italian Mass Election Survey (1968), Italian Mass Election Survey (1972); and
Itanes Survey (2001). The phrasing of the questions used was always identical.
The operationalisation of the concepts of left-right ‘relevance’ and ‘content’
was based on the following criteria.
In terms of the relevance of the left-right dimension, we used the same
conceptual framework and, in part, the same measurement instruments
employed for the Eight nations – Political action research project in both the
first (Barnes & Kaase 1979, esp. Chapter 8 by Klingemann) and second pub-
lications (Jennings at al. 1990, esp. Chapter 7 by Fuchs & Klingemann). More
specifically, we operationalised the relevance of the left-right dimension
through the recognition and the understanding of the two terms. In order to
measure these two components, we used a combination of closed and open-
ended questions.3
With reference to ‘recognition’, we considered the respondents to be able
to recognise the terms ‘left’ and ‘right’ if they agreed to place both themselves
and the political parties on the left-right scale. On the one hand, we recoded
our variable expressing respondents’ self-placement on the left-right axis into
a dummy variable, assigning 1 to the respondents who actually placed them-
selves on the axis and 0 to those who did not. On the other hand, we computed
a new dummy variable, expressing whether our respondents placed at least five
of the main political parties existing at the time of the interview along the
left-right axis. Thus, we derived the following indicators of recognition: per-
centage of people placing themselves on the left-right scale; and percentage of
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
630 piergiorgio corbetta, nicoletta cavazza & michele roccato

people placing along the same scale at least five of the seven main political
parties existing at the time of the interview. As this question was not asked in
the survey performed in 1975, we used the 1968 and the 1972 surveys. Hence,
when analysing recognition we were not interested in how interviewees placed
themselves and the parties, but rather in understanding if respondents recog-
nised the left-right axis (any answer is acceptable) or not (answering ‘I don’t
know’ or refusing to answer).
With reference to ‘understanding’, following Baldassarri’s (2003) proce-
dure, we considered two types of understanding: inter-subjective and subjec-
tive. Inter-subjective understanding (corresponding to correct knowledge of
parties’ positioning on the left-right axis, consistent with the average4 evalua-
tion given by interviewees) is operationalised as the number of correct place-
ments on the left-right sequence according to the average sequence given by
the total sample of respondents (overall 21 pair-wise comparisons among
seven main parties). As this question was not asked in the survey performed in
1975, we used the 1968 and the 1972 surveys.5 Subjective understanding (cor-
responding to the ability to provide any answer to the open-ended question on
the meaning of left and right, regardless of the answer’s actual content) opera-
tionalised as the percentage of people who answer both questions: ‘What do
you mean by right in politics?’ and ‘What do you mean by left in politics?’.
As regards content of the left-right dimension, in order to analyse the
meanings voters attribute to the categories of ‘left’ and ‘right’ we recoded the
answers given by respondents in 1976 and 2006 to the open-ended question
‘What do you mean by left/right in politics?’. The recoding process was carried
out in two steps. First, we built a coding grid including all the semantic areas
needed to classify the answers into homogeneous macro-categories (see
Table 1 for the coding grid and examples of the most commonly used expres-
sions). Then a person blind in respect to our hypotheses proceeded to the
content analysis, coding all the answers following the grid.6
In order to compare variations in abstract and concrete content, we reclas-
sified the semantic areas shown in Table 1 into a condensed grid based on the
level of abstraction used to distinguish between left and right. These levels can
be described as follows. At the first level (maximum concreteness) left and
right are recognised on the basis of which actors are currently involved in
politics and what they do. We placed here responses that mention concrete
political actors, such as a party or a leader, as well as contingent political
arrangements such as government/opposition, majority/minority and so on. At
the second level, left and right are recognised for their support of the interests
of particular social groups, interests that represent guiding principles for their
political action. We placed here concepts that refer to class divisions or reli-
gious dimensions. At the third level, left and right are recognised in terms of
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 631

Table 1. Classification of the responses to the open-ended questions on the meaning of ‘left’
and ‘right’

Semantic areas Terms referred to in the responses

General ideologies Communism, fascism, socialism, Marxism, capitalism,


liberalism, etc.
Equality/inequality Democratic participation, elimination of the social
hierarchy, conservation of differences, social
discrimination, solidarity, ordinary people, etc.
Other values Tolerance, common good, freedom, justice, family,
meritocracy, etc.
Economic dimension Welfare state, state intervention, entrepreneurship,
economic freedom, private property, etc.
Social change/maintenance Conservatives, progressives, innovation, reform,
of status quo revolution, moderates, extremists, law and order,
social disarray, etc.
Democracy/anti-democracy Democracy, constitution, dictatorship, suppression of
individual freedom, anti-liberal system,
authoritarianism, etc.
Class cleavages The rich, the well-to-do, the poor, the blue-collar
workers, the bosses, etc.
Religion Clericalism, religious hypocrisy, fundamentalists,
anti-clericals, secular people, etc.
Political parties Names of parties and coalitions
Political leaders Names of leaders
General assessments I like it/I don’t like it, friends/enemies, honest/dishonest,
reliable/unreliable, competent/incompetent, etc.
Contingent political roles Majority, minority, government, opposition, etc.

their position with regard to the organisation of the state as a guiding principle
of political action. We placed here concepts that evoke an economic dimen-
sion, social change, democracy and/or the constitution. And at the fourth level
(maximum abstractness), left and right are recognised by the fundamental
goals guiding political action. We placed here references to general principles
such as values (e.g., equality) or ideological principles.

Results

Relevance of the left-right dimension

Table 2 presents the results of the analysis of trends in the relevance indicators
between the early 1970s and the first years of the twenty-first century.7 All four
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
632 piergiorgio corbetta, nicoletta cavazza & michele roccato

Table 2. Relevance of the left-right dimension: variations over time

1968 1972 1975 2001 2006

Recognition
Self-placement: percentage of respondents who 75.8 61.4 69.6 82.3 79.3
placed themselves on the left-right scale
Placement of the parties: percentage of 73.0 62.1 – 85.4 89.9
respondents who placed 4–7 parties on the
left-right axis
Understanding
Inter-subjective understanding: average number 11.9 9.4 – 15.6 15.2
of correct responses in the placement of the
parties (range 1–21)
Subjective understanding: percentage of – – 48.7 – 58.5
respondents who gave a definition of ‘left’
and ‘right’

indicators show that over this thirty-year period the relevance of the left-right
dimension increased.With regard to recognition, the percentage of respondents
who placed themselves on the left-right axis rose from around 70 per cent to
about 80 per cent, while the percentage of individuals who could place the
majority of the parties increased from 60–70 per cent to 80–90 per cent. With
regard to understanding, the average number of correct responses about the
placement of the parties on the left-right scale (inter-subjective understanding)
roughly increased from 11 to 15 (out of 21), while the ability to provide an
answer to both questions on the meaning of left and right (subjective under-
standing) rose from about 50 per cent in 1975 to about 60 per cent in 2006.
In terms of both recognition and understanding, in Italy during the period
under consideration the relevance of the left-right dimension increased, thus
supporting HP1.1. It remains to be explained if this increase was due to the
disappearance of traditional political points of reference (HP1.2). We there-
fore investigated whether the relevance of the left-right dimension has par-
ticularly increased among the social categories we have hypothesised as being
more exposed to political disorientation. To perform this analysis on the basis
of sufficiently reliable trends, we analysed the four key indicators of relevance
in two historical periods (the first constituted by the 1968, 1972 and 1975
databases, and the second by the 2001 and 2006 databases). We measured
change between the two periods by examining the index number of the second
period (normalising the average value of indexes for the first period to 100).
Table 3 shows the increases in each indicator by social category. For the overall
sample, the four indicators were greater than 100; this finding is in line with the
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 633

Table 3. Increase in relevance of the left-right axis between 1968–1975 and 2001–2006 by
social category (index no. in 2001–2006 when 1968–1975 = 100)

Recognition Recognition Inter-subjective Subjective


1 2 understanding understanding

Age
Up to 24 104 117 123 121
25–34 120* 127 138 94
35–44 101 127 144 123
45–54 121* 131* 145* 133*
55–64 121* 134* 149* 145*
65+ 120* 146* 164* 130*
Education
High 113 103 104 104
Middle 103 106 108 111
Low 113 125* 139* 117*
Social class
Bourgeoisie 116 108 110 104
White-collar worker 107 115 118 107
Self-employed 111 116 128 114
Blue-collar worker 117* 130* 146* 123*

Notes: * Increase in relevance greater than the average increase of all categories. Recogni-
tion 1: people who placed themselves on the left-right axis. Recognition 2: people who placed
at least 4 out of 7 parties on the left-right axis. Inter-subjective understanding: rate of
accuracy in placing the 21 pairs of parties on the left-right axis. Subjective understanding:
people who answered the question on the meaning of both left and right.

general increase in the relevance of the concepts of left and right discussed
above. However, relevance does not increase in the same way for all social
categories. The percentage of people that are able to place themselves on the
left-right axis (Recognition 1) does not increase in a significantly different way
for respondents of different ages, levels of education or social class. The
remaining three indicators, however, move in the predicted direction. The
percentage of people who are able to place all or the majority of political
parties (Recognition 2) increased during the period under study, especially
among the elderly, people with lower levels of education and working-class
respondents. Among these three groups the capacity of accurately placing
political parties along the left-right axis (inter-subjective understanding) also
increased. Finally, relative to the other social categories, we observe a greater
increase among these three groups in the percentage of people who believe
they understand what is meant by left and right and therefore answer the
open-ended question (subjective understanding).8
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
634 piergiorgio corbetta, nicoletta cavazza & michele roccato

Content of the left-right dimension

The content analysis of the meanings attributed to left and right allowed us to
compare the five hypotheses presented above (HP2.1–HP2.5). In order to
confirm the first hypothesis, we should have observed a clear prevalence of
references to equality-inequality over other kinds of references both in 1975
and 2006. To confirm the second hypothesis, we should have observed a preva-
lence of contents related to class cleavage and economic and religious dimen-
sions in 1975, and a prevalence of references to values in 2006. The third
hypothesis would have implied the prevalence of references to political objects
(leaders and parties) both in 1975 and 2006. The fourth hypothesis would have
entailed the absence of prevalent references both in 1975 and 2006. To confirm
the fifth hypothesis, we should have observed a generalised increase in refer-
ences to more abstract elements relative to more concrete ones between 1975
and 2006, as well as a lack of dominant references in both periods. Moreover,
the increase in the abstraction of the responses should have been greater for
more disoriented social categories, which lose concrete political reference
points.
Table 4 shows the meanings respondents attributed to ‘left’ and ‘right’ in
1975 and in 2006. We compared the frequency of references to the various
semantic areas, regardless of the specific target label (left versus right); the

Table 4. Distribution of the semantic areas used in 1975 and 2006 to describe left and right
(for left and right together)

1975 2006 2006–1975

General ideology 15.4 13.9 –1.5


Equality/inequality 5.9 18.9 13.0
Other values 1.8 6.4 4.6
Economic dimension 1.2 11.5 10.3
Social change/maintenance of status quo 9.3 15.2 5.9
Democracy/anti-democracy 7.3 3.9 –3.3
Class cleavages 17.1 13.5 –3.5
Religion 0.6 0.2 –0.4
Political parties 31.0 4.1 –26.9
Political leaders 1.4 3.7 2.3
General assessments 7.6 7.8 0.2
Contingent political roles 1.4 0.8 –0.6
Total 100 100
Number of responses 1,765 1,737

© 2009 The Author(s)


Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 635

percentages were thus calculated on the basis of the actual responses, rather
than on the number of respondents.
In 1975 there was a clear prevalence of direct references to political parties.
Almost one-third of the respondents traced the meaning of ‘left’ and ‘right’ to
political parties competing for voter support during that period. Some 31 years
later, the dominance of parties was no longer noticeable: Italians explained
meanings of ‘left’ and ‘right’ by drawing in a more balanced way from all
semantic areas. Moreover, between 1975 and 2006 an increase in references to
values (in large part concentrated in ‘equality’: +13 percentage points; but
‘other values’ increased as well: +4.6) took place. Although the conceptual pair
equality/inequality was mentioned as an aspect of the left-right distinction by
almost 19 per cent of responses in 2006, it was not possible to claim that this
was a dominant area of meaning. This pair (in the same year) represented 30
per cent of the meanings attributed to the left, but remained extremely weak
among the meanings attributed to the right (8 per cent).9 Again, in the diach-
ronic comparison we observed an increase in references to economic dimen-
sions (+10.3 percentage points) and to social change (+5.9).
Consistent with H2.5, these changes translated into an overall reversal in
the concreteness-abstraction proportions in the responses between 1975 and
2006 (see Table 5).The strong increase in references to abstract elements could
not be attributed entirely to the development of more sophisticated cognitive
abilities due to mass education, rather than to the evolution of social repre-
sentations of left and right in politics. Indeed, a linear regression analysis using
the abstraction level of the meaning given to left and right as the dependent
variable and the year of data collection (1975 = 0; 2006 = 1) and respondents’

Table 5. Distribution of the level of abstraction of the meanings attributed to left and right
in 1975 and 2006

1975 2006 2006–1975

1.00 Maximum concreteness 28.5 9.3 –19.2


1.5 8.2 2.0 –6.2
2 13.7 11.6 –2.1
2.5 14.9 12.9 –2.0
3 16.6 23.5 6.9
3.5 8.2 17.6 9.4
4.00 Maximum abstraction 9.8 23.0 13.2
Total 100 100
N 958 931
Mean 2.234 2.920

© 2009 The Author(s)


Journal compilation © 2009 (European Consortium for Political Research)
636 piergiorgio corbetta, nicoletta cavazza & michele roccato

Table 6. Regression coefficients between year (1975/2006) and abstraction

Independent variables B Beta Significance Adjusted R2

Model 1: year 0.686 0.334 0.000 0.111


Model 2: year, level of education 0.572 0.279 0.000 0.127

Table 7. Average levels of abstraction in the meaning of left and right by social category
(range 1–4); variations between 1975 and 2006

Index no. in 2006


1975 2006 (1975 = 100)

Age
21(18)–24 2.19 2.87 131
25–34 2.28 2.79 122
35–44 2.37 3.01 127
45–54 2.15 2.93 136*
55–64 2.17 2.88 133*
65+ 2.20 2.99 136*
Education
High 2.84 3.03 107
Middle 2.52 2.97 118
Low 2.14 2.81 131*
Social class
Bourgeoisie 2.54 3.20 126
White-collar worker 2.33 2.96 127
Self-employed 2.24 2.88 129
Blue-collar worker 2.10 2.85 136*

Note: * Increase in abstraction greater than the average increase of all categories.

educational level (elementary, middle or high) as independent variables,


entered in steps 1 and 2 respectively, shows that the regression coefficient for
the year of data collection was highly significant even when controlling for
education (Table 6).
The fifth hypothesis was further confirmed by the differences we observed
among social groups in the level of abstraction and by their change over time.
As shown in Table 7, in 1975 the most disoriented groups (made up by older
and less educated respondents and by people belonging to the working class)
expressed the most concrete level of content; yet these same groups also
displayed the greatest increase in abstraction in 2006.
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 637

Conclusions

Our findings consistently show that the left-right axis is alive and well. Over
the course of the past thirty years, citizens have ascribed increasing relevance
to the left-right dimension, as testified by the growing proportion of individu-
als who recognised and were capable of using this guideline to orient them-
selves in a complex political environment. Although distinct from one another,
our four indicators of recognition and understanding of the left-right dimen-
sion show a coherent trend over time. Furthermore, three out of four compari-
sons provide compelling evidence of the trend for more than two points in
time. One could argue that the improved performance of Italians in locating
parties on the left-right axis cannot entirely be attributed to an increased
relevance of the left-right dimension, but rather might be due to the increasing
salience of the left-right opposition, resulting from the change in the electoral
system from proportional to majoritarian. Nevertheless, this alternative inter-
pretation is ruled out by the evidence showing that such change is most
pronounced among the most disoriented and less sophisticated social
categories.
As far as content is concerned, the results obtained from the analysis of
Italians’ responses to open-ended questions are consistent only with our fifth
hypothesis. The evidence can be summarised as follows:
We find no support for the first hypothesis (HP2.1), which anchors the
left-right meanings to the concept of ‘equality’ and predicted no change over
time. First, as in Klingemann (1979; Fuchs & Klingemann 1990), the multiplic-
ity of the interpretations given by respondents renders implausible the idea
that, in the eyes of voters, the left-right axis is characterised by a single
meaning. Second, the frequency with which Italians refer to the value of
equality is too weak to justify Bobbio’s (1996) view that the notion of equality
is the main basis of opposition between left and right.
Our results do not confirm the second hypothesis (HP2.2), asserting that
there has been a shift from a materialist, economic axis, largely based on class
divisions, to a ‘post-materialist’ axis centred on values. Indeed, class divisions,
which should have been the principal focus of left-right meanings during the
1970s, received only 17 per cent of the responses in 1975.Their decline between
1975 and 2006 is modest and compensated for by an increase in the relevance
of economic issues (state intervention in the economy, private propriety,
welfare state, taxes and so on). When the two areas of largely ‘materialist’
concerns are combined, we observe an overall rise in these meanings – con-
trary to the ‘post-materialist’ thesis. Moreover, the increase of the importance
of values between 1975 and 2006 we found is not sufficient to confirm the
dominance of values in the meanings attributed to left and right.
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
638 piergiorgio corbetta, nicoletta cavazza & michele roccato

Based on our previous analyses on relevance, we refute the hypothesis that


the left-right axis is a proxy for party identification (HP2.3), having confirmed
the persistence of the left-right axis even in a situation characterised by a
substantial decline of political parties as points of reference. The results of the
content analysis are also consistent with this interpretation. In 1975 the most
common meaning assigned to the left-right dimension referred to political
parties (e.g., ‘left = Communist Party’). In those years, the hypothesis that the
left-right opposition was a proxy for party identification was to a large extent
valid. However, things have changed over time: today very few Italians, when
asked about the meaning of left-right, mention political parties.Yet, despite the
general tendency toward a more personalised political context, leaders have
not replaced parties; in fact leaders are mentioned only in 4.9 per cent of the
responses.
The fourth hypothesis (HP2.4), conceptualising a decline in the area of
abstract principles in favour of more concrete references over time, finds no
support. On the contrary, we observe a dramatic decrease of concrete refer-
ences (primarily those related to parties) and an increase of abstract core
values (primarily equality), plausibly due to the complex and transitory char-
acter of the political situation in 2006, as compared to that of 1975.
Our fifth hypothesis (HP2.5) anticipated that the complexity and transitory
nature of the thirty years under examination would have resulted in changes in
the more peripheral system of representations of left and right – that contain-
ing the most concrete references. In addition, we hypothesised that these
changes would have pushed voters particularly disoriented by changes in the
political scene to adopt even more abstract and fundamental criteria to dis-
tinguish between left and right. Our results show an increasing recourse to
abstract guiding principles, even by those segments of population that in the
past tended to express concrete contents; this tendency is valid despite the
weakening of ideology as an orienting framework. In addition, we show that
this effect was not simply due to a general improvement in the cognitive
abilities of Italians resulting from a rise in the level of education. Hence, more
than in the past, the criteria according to which individuals ascribe meanings to
left and right are to be found in a solid core, assuring the intelligibility of
politics above and beyond its contingent changes.
In conclusion, our results show that the left-right dimension continues to be
very meaningful for citizens and to perform decisive functions of political
orientation. In the current context, the social representations of left and right
mainly perform two functions previously fulfilled by ideologies: social and
individual orientation. The core/periphery structure of social representations
enables voters to activate both top-down processes (e.g., to determine what
can be categorised as left-wing or right-wing based on some fundamental value
© 2009 The Author(s)
Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 639

criteria) and bottom-up processes (e.g., attach additional meanings to right and
left in relation to new objects that appear on the political space).
What then is the difference between ideologies and social representations?
Why are ideologies no longer able to render political conflict intelligible,
whereas social representations can? The principal reason concerns the mode
of elaboration of these two systems of ideas. Ideologies are a product of
intellectual and political elites, who defend orthodoxy, thereby making ideolo-
gies inflexible. They thus resemble what Moscovici (1984: 44) has called ‘reified
universes’ or ‘a map of forces, objects and events that are independent of our
desires and outside of our awareness, and to which we can react impartially
and submissively’. On the contrary, social representations draw on consensual
universes – that is, those that ‘reestablish collective awareness and shape it,
explaining objects and events in a way that makes them accessible to every-
body and makes them coincide with our most immediate interests’. Empirical
evidence supporting this genetic process is to be found precisely in the differ-
ent dynamics that we have identified for distinct social categories. During the
thirty years under examination, regardless of their level of education, Italians
have developed the ability of using symbolic tools to link what is new to what
is already familiar.
We believe that our results could be extended beyond Italy. Indeed we
speculate that in other contexts, under conditions of political uncertainty,
change in traditional points of reference and structural transformations, citi-
zens collectively endeavour to restructure meanings. The starting point of this
process would be the most solid and reliable core of past experience, ensuring
continuity, maintaining an identity-sustaining function (telling us who we are
and who we have been) and facilitating the incorporation of novelty into
familiar schema. Future research might directly test this hypothesis within the
framework of a comparative research project.

Notes

1. The results from five of these nations, not including Italy, were reported in Barnes and
Kaase (1979).
2. Itanes stands for ITAlian National Election Studies. Readers interested in this research
programme should visit the website at: www.itanes.org/.
3. We used the following questions. Closed-ended questions:‘In political matters people talk
of “the left” and “the right”. In this card there is a row of cells going from the left to the
right. Thinking about your political opinions, where would you place yourself?’ (card with
10 cells); ‘Where would you place the following parties on the same scale? If you don’t
know the party or the answer, please tell me you don’t know.’ As usual in this kind of
questions, the scale consisted of ten horizontally ordered boxes, without an explicit
midpoint and without any number or label in the boxes, except for the words ‘left’ and

© 2009 The Author(s)


Journal compilation © 2009 (European Consortium for Political Research)
640 piergiorgio corbetta, nicoletta cavazza & michele roccato

‘right’ at the end points. Open-ended questions: ‘What do you mean by “left” in politics?’
(probe fully); ‘What do you mean by “right” in politics?’ (probe fully).
4. The ordering of political parties along the left-right axis, on the basis of an average score
obtained on a 1–10 scale, corresponds to the shared representation of the Italian political
system in all elections (Ricolfi 1999).
5. All of the possible 21 pairings of the seven parties were examined. If in the standard
sequence of comparisons based on the average placement of parties by respondents, party
A was further to the left of party B, the voter who placed it further to the right or gave it the
same score on a ten-point scale made a mistake. We did not count as erroneous cases in
which an equal score was given to parties separated by less than one point in the average
score. Since there were 21 possible comparisons, the range of the variable was 0–21.
6. The question was posed only to those who responded to the question of self-placement on
the left-right axis.
7. The bulk of our analysis consisted of a comparison between 1975 and 2006 – the years for
which open-ended questions on the meaning of left and right were available. When using
other indicators (when available) we have been able to extend this comparison to other
elections, with a better understanding of the trends under analysis. The question about the
party placement on the left-right axis was not available in 1975.
8. These results may not be attributed to a general increase in the willingness or readiness
of Italians to answer survey questions. Indeed, a linear regression analysis performed on
a database merging the 1975 and the 2006 data showed that, net of respondents’ educa-
tion, the number of non-responses on the common items was not significantly influenced
by the year of data collection. Readers interested in examining the results of the analysis
may contact the first-named author of this article.
9. Data not reported in the table for the sake of economising on space. Readers interested
in examining them may contact the first-named author of this article.

References

Aron, R. (1962). The opium of the intellectuals. New York: Norton.


Baldassarri, D. (2003). Il voto ideologico esiste? L’utilizzo delle categorie di sinistra e destra
nell’elettorato italiano. Quaderni dell’osservatorio elettorale 49: 7–35.
Barnes, S.H. & Kaase, M. (eds) (1979). Political action: Mass participation in five Western
democracies. Beverly Hills, CA: Sage.
Bell, D. (1965). The end of ideology: On the exhaustion of political ideas. Glencoe, IL: Free
Press.
Bobbio, N. (1996). Left and right: The significance of a political distinction. Chicago, IL:
University of Chicago Press.
Butler, D. & Stokes, D. (1969). Political change in Britain: Forces shaping electoral choice.
London: Macmillan.
Doise, W. (1989). Attitudes et représentations sociales. In D. Jodelet (ed.), Les représenta-
tions sociales. Paris: PUF.
Flament, C. (1987). Pratiques et représentations sociales. In J.-L. Beauvois & J.-M. Monteil
(eds), Perspectives cognitives et conduits sociales. Cousset: DelVal.
Fuchs, D. & Klingemann, H.D. (1990). The left-right schema. In M.K. Jennings & J. van Deth
(eds), Continuities in political action: A longitudinal study of political orientations in three
Western democracies. Berlin: de Gruyter.

© 2009 The Author(s)


Journal compilation © 2009 (European Consortium for Political Research)
between ideology and social representations 641

Inglehart, R. (1977). The silent revolution: Changing values and political styles among Western
publics. Princeton, NJ: Princeton University Press.
Inglehart, R. (1990). Cultural shift in advanced industrial society. Princeton, NJ: Princeton
University Press.
Inglehart, R. & Klingemann, H.-D. (1976). Party identification, ideological preference and
the left-right dimension among Western publics. In I. Budge, I. Crewe & D. Farlie (eds),
Party identification and beyond. London: Wiley.
Inglehart, R. & Sidjanski, D. (1976). The left, the right, the establishment and the Swiss
electorate. In I. Budge, I. Crewe & D. Farlie (eds), Party identification and beyond.
London: Wiley.
Jennings, M.K. et al. (1990). Continuities in political action: A longitudinal study of political
orientations in three Western democracies. Berlin: de Gruyter.
Jost, J. (2006). The end of the end of ideology. American Psychologist 61: 651–670.
Klingemann, H.D. (1979). Measuring ideological conceptualizations. In S.H. Barnes &
M. Kaase (eds), Political action: Mass participation in five Western democracies.
Beverly Hills, CA: Sage.
Knutsen, O. (1998). Expert judgements of the left-right location of political parties:
A comparative longitudinal study. West European Politics 21: 63–94.
Kornhauser, W. (1960). The politics of mass society. London: Routledge & Kegan Paul.
Lipset, S.M. (1960). Political men: The social bases of politics. Garden City, NY: Doubleday.
Mair, P. (2007), Left-right orientations. In R.J. Dalton & H.D. Klingemann (eds), The Oxford
handbook of political behaviour. Oxford: Oxford University Press.
Moscovici, S. (1984). The phenomenon of social representations. In R.M. Farr & S. Moscovici
(eds), Social representations. Cambridge: Cambridge University Press.
Moscovici, S. (1988). Notes towards a description of social representation. European Journal
of Social Psychology 18: 211–250.
Ricolfi, L. (1999). Destra e sinistra? Torino: Omega.
Sani, G. (1974). A test for the least-distance model of voting choice: Italy, 1972. Comparative
Political Studies 7(2): 193–208.
Sartori, G. (1976). Parties and party systems: A framework for analysis. Cambridge:
Cambridge University Press.
Shils, E. (1968). The concept and function of ideology. In D.L. Sills (ed.), International
encyclopedia of the social sciences. New York: Macmillan/Free Press.
Tversky, A. & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases.
Science 185: 1124–1131.
Waxman, C.I. (ed.) (1968). The end of ideology debate. New York: Funk & Wagnalls.

Address for correspondence: Piergiorgio Corbetta, Dipartimento di Scienze dell’educazione


‘Giovanni Maria Bertin’, Università di Bologna, Via Filippo Re 6, I-40126 Bologna, Italy.
Tel.: +39 0512091503; Fax: +39 0512091489. Email: piergiorgio.corbetta@unibo.it

© 2009 The Author(s)


Journal compilation © 2009 (European Consortium for Political Research)

You might also like