The changing relevance and meaning of left and right in 34 party systems from 1945 to 2020

The extent to which the left–right dimension still structures party systems in highly developed, industrialized democracies is a contested field in comparative politics. Most studies in this area take the position that a stable and universal left–right dimension is either still the most important game in town or has become obsolete and replaced by other policy dimensions. Although country-specific studies focusing on voters’ left–right self-placement discover different meanings of left and right that vary between countries and change over time, few macro-comparative studies focusing on parties or governments take this aspect into account. Using a left–right concept for party politics from the PIP project on Parties, Institutions and Preferences that distinguishes an ideological core derived from political theory, as well as country- and time-specific issues uncovered through empirical analysis, the article demonstrates fundamental differences in the relevance and meaning of left and right by analyzing 34 party systems from 1945 to 2020. The article shows that the thesis of the decline of the left and right is premature. An important aspect for the continued high relevance of the left–right dimension is the fact that left and right changes their meaning by including controversial issues such as European integration, migration and environmental degradation.


Introduction
The left-right dimension structures political party systems and is a strong predictor for policy stances of political parties. As demonstrated in the seminal work of Lipset and Rokkan (1967), it is manifested in the cleavage structure between capital and labor. This conclusion has since been echoed in many studies during the last five decades (to mention just some : Hibbs 1977;Benoit and Laver 2006;McDonald and Budge 2005;Potrafke 2017). As a result, the classification of political parties on a left-right scale is a lively field of research in macro-comparative analysis (Castles and Mair 1984;Laver and Hunt 1992;Huber and Inglehart 1995;Benoit and Laver 2006;Warwick 2006;Budge et al. 2001;Klingemann et al. 2006;Imbeau et al. 2001). However, the dominance of the left-right dimension in party politics has been questioned. Some claim that parties are no longer clearly left or right and adapt their programs according to contextual or situational circumstances (Häusermann et al. 2013;Zohlnhöfer and Bandau 2020). Others claim that the left-right dimension is challenged by other policy dimensions (Inglehart 1977). Party research has shown that an ecological dimension emerged in the 1970s and 1980s and shaped a "new politics" (Hildebrandt and Dalton 1978;Kitschelt 1988;Müller-Rommel 1990). Others have demonstrated that the left-right dimension has been replaced by adjusting to new challenges. Kitschelt (1994) for instance suggests that the left-right dimension moved its axis due to the impact of structural changes and new upcoming demands. Most recently Hooghe and Marks (2018;Marks et al. 2021) refer to this hybrid dimension as the GAL/TAN dimension (Green-Alternative-Left versus Traditional-Authoritarian-Nationalism). They claim that through the impact of globalization, migration, and the financial crisis in the first decade of the new millennium the GAL/TAN forms a new cleavage which dominates in many industrialized societies and by doing this, replaces the left-right dimension (see also Kriesi et al. 2012).
In this article, I take a different perspective on changes in policy dimensions by starting from the left-right dimension. Rather than postulating that the left-right dimension is being replaced in party politics, I analyze the extent to which the relevance and meaning of left and right have changed in the course of the postwar period in 34 democracies. It is an accepted fact in political theory and the behavioral sciences that a political ideology changes over time or presents itself differently under specific regional circumstances (Freeden 2013;Caprara 2020). However, comparative party research seems to assume ultra-stable political dimensions. Although there are sometimes calls to apply more context-sensitive ideology indices (Bakker and Hobolt 2013), few studies in comparative politics use flexible left-right dimensions (Gabel and Huber 2000;Franzmann and Kaiser 2006;Jahn 2011;Bakker and Hobolt 2013). The focus of this study is therefore on the question of which particular left and right positions the parties adopt and how these differ regionally, country-specifically and/or change over time.
A related topic that has been hotly debated is whether the left-right dimension has diminished over time and whether it is equally important in every country or region. The first three decades of the postwar period are often assumed to be the golden age of the left-right dimension. From the 1970s, structural change and the signs of the limits of growth posed a new challenge to the left-right dimension. Furthermore, there is certainly variation in how relevant the left-right dimension is across various countries. Some countries have a particular history that can make the left-right dimension less relevant. Werner Sombart's (1976Sombart's ( [1906) challenging question "Why is there no socialism in the USA?" represents these aspects. The second research question is therefore to analyze whether the relevance of the left-right dimension has diminished over time.
To date, no study has systematically examined the relevance and meaning of the left-right dimension in the party systems of democratic states throughout the entire postwar period. The assumption that the left-right dimension dominated in the first postwar years is pure speculation and has not yet been thoroughly investigated. The claim that the left-right dimension has lost importance therefore lacks any empirical basis. In this article, I give some initial indications as to whether these speculations and conjectures have any empirical basis. I address the above research questions by first conducting an analytical-descriptive study of 34 democratic countries from 1945 to 2020 to determine whether the left-right dimension has become less relevant over the past eight decades. This is followed by five country observations that illustrate the evolution of country-specific meanings of left and right. The analysis shows that the left-right dimension does not lose relevance when political parties incorporate relevant societal issues into the left-right semantic. Before turning to these empirical questions, however, I would like to outline the theoretical basis for the empirical analysis.

Defining left and right positions in party systems for macro-comparative studies
Although left and right is frequently used in various fields in political science as well as in politics, clearly theoretically embedded definitions for empirical studies are rare (Cochrane 2015). There are numerous conceptual and empirical studies that address this issue. Many of these studies base their concept of parties' left and right positions on voters' self-assessments (Vries et al. 2013;Freire 2015) or on expert judgments (Laver and Hunt 1992;Warwick 2006;Benoit and Laver 2006;Jolly et al. 2022). These studies conclude that the left-right dichotomy is multi-dimensional and often distinguish between a socioeconomic and a cultural dimension. Moreover, these studies show that the meaning of left and right varies greatly across geographic areas (Markowski 1997;Colomer 2005;Jou 2010;Freire and Kivistik 2016). The major drawbacks of these studies are that they have no theoretical basis for what is left and right (they are based on assessments of individuals, sometimes supported by briefly outlined statements) and that they refer to parties from assumptions of voters or experts, which risks ecological fallacy (for a comparison of these approaches see: Bakker and Hobolt 2013). Analyses of party election programs address the latter point by using material produced by the political parties themselves. 1 The most commonly used index is the right-left index (RILE) of the MARPOR project (Budge et al. 2001;Klingemann et al. 2006;Volkens et al. 2013). However, the RILE also struggles with a theoretical underpinning of its left-right concepts. 2 To know theoretically what we mean by left and right, we must start from political theory. Most thorough is Norberto Bobbio (1996) who explores this elusive distinction and argues that left and right are ultimately divided by different attitudes toward equality. He points out that the left strives for greater equality, whereas the right legitimizes inequality (Bobbio 1996; see also Lukes 2005). That means that attitudes toward equality constitute the core of the left-right ideology. The left aims to mitigate or even eliminate social inequality through redistributive measures. The right legitimizes inequality; either by referring to tradition, arguing that inequality is caused by nature (conservatism), or by pointing out that inequality is caused by differences in individual achievement (liberalism).
The core of an ideology is what makes an ideology. It is universal across time and space. This leads to the question of whether left and right can have different meanings in different regions and over time. The answer is yes and no. The core is universal, but an ideology does not consist only of its core. Freeden distinguishes between the ideological core, the adjacencies and peripheries. He explains that ideologies "… are not set in stone and-albeit at a variable speed-will fluctuate over 1 There are also studies that capture party positions by analyzing materials indirectly related to parties, such as newspapers (Kriesi et al. 2012). However, this example only covers a handful of countries over a short time period. 2 It must be mentioned that the coding of the party manifestos was originally not aimed to construct a left-right index. Early explanations made clear that the index is based on empirical observations, i.e., it is inductive. In describing the construction of the RILE, Laver and Budge (1992, 23) stated: "…we did a large number of exploratory factor analyses to search for combinations of variables that persistently loaded together over a wide range of systems." It was later claimed that the RILE was based on deductive considerations drawn from the writings of Marx and Burke (Klingemann et al. 2006: 6). However, there are some inconsistencies because in still another source it was proclaimed that the right issues have not deduced from Burke but by the policy statements of the British Prime Minister Margaret Thatcher and the President of the United States Ronald Reagan (Budge et al. 2001, 22). The latter are certainly no source for a theoretically deductive analysis. Referring to their left items, Klingemann et al. (2006, 5-6) states: "There is after all no logical or inherent reason why support for peace should be associated with government interventionism …." This is only justified because these statements are mentioned together by political parties: "The fact remains however that party ideologies do put them together." In this view it sounds rather inappropriate to call the RILE a "… theoretically driven general Left/Right scale …" (Budge and McDonald 2007). The theoretical justification took place a posteriori and reached its climax in publication, in which further classical authors were listed, which were supposed to have led to the selection of the ideological statements of the RILE (Budge and Meyer 2013). Using 13 statements for the left and 13 for the right means to include almost half of the party manifesto coding. This has also been criticized on the empirical side. Hans Keman (2007, 79) concludes that the inclusion of aspects like "… military, constitutionalism, morality, law and order, democracy, and so on …" in a Left-Right scale "… is not only confusing, but also wrong. … Including other matters of serious dispute that divide political parties and their constituencies is certainly relevant for understanding electoral politics and party systems, as well as policy-making by government. Yet reducing these complex differences into one dimension or using simple dichotomies is bad for comparability and conceptual clarity.". time and space". (Freeden 2013, 124) Core concepts are the most stable and "… indispensable to hold the ideology together …" (Freeden 2013, 125). Without the core concepts, the ideology cannot be described as a specific ideology. Adjacent and peripheral concepts refine and specify the core concept and change more frequently and differ according to the environments of the ideology. Adjacent concepts are second-ranking in pervasiveness and breadth of meaning and they "… are crucial in finessing the core and anchoring it-at least temporarily-into a more determinate and de-contested semantic field" (Freeden 2013, 124). Peripheral concepts are most volatile and align with an ideology in specific situations for a short time.
For an empirical analysis, these considerations require us to examine the extent to which the ideological core of the left and the right is able to align itself with other issues over time and space. We may assume that the left-right dimension is still relevant in countries where political parties are able to align a substantial amount of adjacent and peripheral ideological issues to the left-right core. Adjacent concepts are assumed to stabilize the left-right dimension when they encompass other dimensions of political conflict, such as opposing positions on European integration or the opposition between the environment and growth. In the following, I outline how these theoretical considerations can be translated into a systematic macro-comparative analysis.

Methodological considerations
In order to conduct an empirical analysis of 34 countries over eight decades, I use a left-right index which distinguishes between the ideological core which is the same over all the countries and periods, on the one hand, and those issues which align temporally and country-specifically, on the other. Such an index has been developed by Jahn (2011) using data from the MARPOR project, which analyzes party manifestos in more than 50 countries over the entire post-World War II period (Budge et al. 2001;Klingemann et al. 2006). 3 Jahn uses Bobbio's key terms to describe the core of left and right and identifies ten statements from the 56 coded statements in the MARPOR project that correspond to these terms. 4 In a next step, Jahn uses regression analysis to identify adjacent and peripheral concepts (he calls them extra left-right issues or LR_plus). If the remaining 46 statements correlate significantly with the index of core statements, they are identified as belonging to the left-right dimension in a particular country or time period. Finally, the core index and the LR_plus index are summed to obtain a country-and period-specific left-right (LR) index (for further details see : Jahn 2011;Jahn et al. 2018).
Although this data is readily available, it has never been thoroughly analyzed. I will do this for the first time in two steps. To examine the extent to which the left-right dimension is relevant, I consider the proportion of left and right statements relative to total statements made in an election program. I do this for the core issues and for the overall index. This analysis provides information on whether the left-right dimension as a whole or the core left-right statements have lost relevance over time. In addition, I can determine whether the relevance of the left-right dimension has been achieved by integrating (new) adjacent and peripheral issues.
In order to identify trends I consider four periods. First, the years after World War II from 1945 to 1973. The year 1973 marked the end of sustained growth and a solidification of the reorientation of the left in the 1960s. It is assumed that 1945-1973 are the heydays of the left-right dimension, but we lack robust empirical analysis of whether this is indeed the case. The focus on this period addresses this assumption for the first time in a systematic comparative analysis. The second period starts after the oil crisis and lasts until 1989, the year before the breakdown of the Communist regimes in Eastern Europe. During this period, green movements and parties emerged, and social democratic parties were challenged by an increasing number of post-industrial voters and the emergence of environmental concerns (Kitschelt 1994;Fagerholm 2016 Explanation: The number of the "per" refers to the statements as mentioned in Budge et al. (2001) and Klingemann et al. (2006). The table is taken from Jahn (2011, 753) In fact, the empirical result of a multidimensional analysis shows a triangle that confirms that the left is one-dimensional and the right is split between liberalism and conservatism (Jahn 2011, 754). There is the matter of coding "social equality" in the manifesto coding (per503). This statement focuses on the core of left-right, but unfortunately also mixes this with liberal statements about "fair treatment in the tax system" (whatever that means, DJ). Therefore, I did not consider this statement to be a core left-right statement in this article.
its dynamism during this period in countries that have been challenged by the environmental requirements of large sections of the population. However, it is also possible that in some countries the environmental issue has been integrated as an adjacent element in the left-right dimension, increasing the relevance of the left-right dimension in these countries. Globalization took off in the 1990s, which is why this period is considered a separate period until the financial crisis in 2007. The impact of globalization on the relevance of the left-right dimension in party politics is difficult to predict. On the one hand, globalization may have had depoliticizing effects since domestic politics has been more and more influenced by international factors. In addition, the end of the Cold War may have depoliticized the class conflict. All these factors may have weakened the relevance of the left-right dimension (Kwon and Pontusson 2010;Potrafke 2017). On the other hand, social inequality increased in most countries during globalization, which may have strengthened the relevance of the left-right dimension in this period. The years following the financial crisis at the end of the first decade of the new millennium constitute the last period analyzed in this article. It can be assumed that the crisis has further increased social inequality and poverty and that the left-right dimension is reviving, as some studies suggest (Bremer 2018;McManus 2019).
In a second step, I analyze the adjacent and peripheral issues in further detail and ask which statements were integrated into the left-right dimension. Since this analysis is more specific and explorative, I will focus on the countries where the left-right dimension is still most predominant and on the countries where it is weakest; once for the established Western democracies and once for the Central European democracies. This diverse case design allows for the identification of whether there are systematic differences between these groups of countries (Gerring 2008;Seawright and Gerring 2008). The presentation of the data is based on heatmaps. The values in the heatmaps are scores from a multi-dimensional analysis of all statements which turned out to be significant in a regression with the left-right core index (Cox and Cox 2001;Jahn et al. 2018). The replication data or uploaded in Harvard Dataverse.

The relevance of the left-right ideology in party systems over time and space
In order to estimate the relevance of the left-right dimension, I first conduct an analysis of the established OECD countries during the post-World War II period. There are data for only these countries over the whole period from 1945 to the most recent election in the dataset. 5 This long time period makes it possible to assess if there have been essential changes in the relevance of the left-right dimension. In a second step, I focus on the Central Eastern European (CEE) countries and analyze the relevance of the left-right dimension in these countries from 1993 to 2020.
Both these analyses give us a detailed picture about the changing relevance of the left-right dimension in various countries and periods. The results of the analysis are presented, on the one hand, for all countries individually and, on the other, according to families of nations (Castles 1993). 6 Considering the entire period and all the 23 established democracies from 1945 to 2020 shows that more than half of all statements raised in party manifestos are left-right issues. Around one-third of them refer to the core left-right ideology. However, the data over time reveal some surprising findings. It is true that in the first period, the left-right dimension had a high relevance, but it has not declined substantially during the following decades. In fact, after the financial crisis at the end of the first decade of the new millennium, the relevance of left and right actually increased by more than ten percentage points. Interesting is the fact that this trend cannot be confirmed for the core ideology. Here the relevance has steadily declined over the years. That means that the left-right dimension remained relevant but that its meaning had obviously changed. This aspect will be discussed in more detail in the next section, but first let us look at the geographical differences in the relevance of the left-right dimension.
The strongholds of the left-right dimension are the Nordic and continental countries. In the Nordic countries, the core statements are most frequently mentioned: they comprise almost one-fifth of all statements. However, there is a clear linear trend of decline in which the Nordic countries had the highest amount of core statements in the early period and fell dramatically in the following decades. This means that the "new" left-right dimension must be formed by integrating adjacent issues.
The decline in the left-right index between 1974 and 1989 in continental European countries may be due to the emergence of green or libertarian-left ideology in that decade. Green parties appeared in many party systems, including Germany, Austria, Belgium, France, Luxembourg and Italy. In contrast to the Nordic countries, the core left-right statements remained at similar levels from the 1970 to 2020. In continental Europe, the countries with the highest relevance of the left-right dimension are the Netherlands and France. In the Netherlands, however, the relevance of the left-right dimension declined rapidly in the following decades in the course of de-pillarization, i.e., the separation of ideological camps in Dutch society. In France, the left-right dimension was strong in the first five decades until the 1990s. This is especially true for the early years. Looking at France only in the period of the Fourth Republic, we find that the left-right dimension was strongest in this country between 1946 and 1958 (not shown in Table 1). Here 70 percent referred to left and right statements, in comparison with 64 percent from October 1958 to October 1973 in the first years of the 5th Republic. Between 1974 and 1989 it remained on this level and then dropped dramatically in the period of globalization. Also, the post-crisis period in the new millennium did not return the left-right dimension back to its former level of importance in the two first periods and in particular in the Fourth Republic. The Anglo-Saxon countries are a bit of a surprise. The Antipodes frequently refer to the left-right dimension. During the 1945-1973 period, New Zealand used nearly 70 percent left-right statements, tied with Sweden, the leading country during that period. In this context, the UK is a deviant case. Despite being the motherland of class struggle and early industrialization, only around 41 percent of the statements refer to the left and right in the early period. In the post-2008 period, however, there has been a huge increase in the relevance of the left-right dimension in all Englishspeaking countries except Canada. This rise has been particularly intense in the UK.
Mediterranean countries assumed above-average levels of left-right statements in the period during their first democratic years, but the left-right dimension became less important in subsequent decades. Even if the saliency of the left-right dimension increased in the aftermath of the financial crisis, the Mediterranean countries continue to have the lowest score. This is surprising because the crisis hit these countries severely. In particular, Greece has a very low left-right score. That means that despite the severe crisis in this country, the left-right dimension cannot align with many other issues. However, the core ideology is more relevant in Greece than in most other countries, which shows a return to the traditional left-right issues in the aftermath of the crisis (see also Hutter et al. 2018).
The data on Japan show that Japan is a special case. The left-right value is significantly below the mean value of the other countries. However, the financial crisis has increased the proportion of left-right statements even in Japan.
The results so far show that the left-right dimension is not decreasing. All in all, it is very stable and in the time after the financial crisis the left-right ideology even reached its preliminary climax. However, it is also evident that the core left-right statements have decreased significantly in the postwar period. In what follows, I will focus on countries that do not have as long a history of free elections as the countries examined so far. This is true above all for the CEE countries where some claim that left and right is not as relevant as in Western democracies (Myant and Drahokoupil 2014;Tavits and Letki 2009). I distinguish two time periods: first, the years from the first democratic election to the financial crisis in 2007 and then the period from 2008 to 2018, which is identical to the last period of established democracies and therefore comparable to these countries. The countries of CEE are divided into four groups: the Visegrád countries, the Baltic countries, the countries of former Yugoslavia and the Eastern Balkans. Few indications exist to assess different meanings and relevance of the left-right dimension in this region. Table 2 shows the results. Table 2 shows that the CEE countries make significantly fewer references to left-right statements than Western democracies. This is especially true for the overall index. The core index hardly differs from the Western democracies. Another finding is that the left-right reference has hardly changed over time. This is in stark contrast to the Western democracies, where the period after 2008 led to a significant increase in left-right statements. However, CEE countries show very opposite trends.
In the Visegrád countries and in the countries of former Yugoslavia, the relevance of left-right statements increased significantly after 2008. This is especially true for the Czech Republic and Slovakia.
In Baltic States, except of Estonia the relevance of the left-right dimension is clearly declining. In the two Eastern Balkan states, the influence of the left-right axis also declines after 2008. This decline is very moderate in Bulgaria, but radical in Romania.
Core statements from the left and right are relatively stable in CEE. Exceptions are the Eastern Balkans, where these statements have declined sharply after the financial crisis of 2007. In contrast, we see an increase in Slovakia, Latvia and Slovenia. Overall, the binding power of the left-right dimension is less pronounced in the CEE countries than in many Western countries. To explore this in more detail below, I conduct an analysis of the adjacent and peripheral ideological elements in the leading and lagging countries in both Western and Eastern Europe.

The changing meaning of left and right in party systems in modern democracies
As previously mentioned, the core of left and right is stable in time and space, so the change in the relevance of left and right is based on the adjacent and peripheral ideological elements. The integration of these ideological elements determines the relevance of the left and the right, as we saw in the previous part of the article. To illustrate which aspects are integrated into the left-right dimension and how they  1993-2007 2008-2020 1993-2020 1993-2007 2008-2020 1993-2020  change the meaning of left and right, I present the adjacent and peripheral ideological elements of the countries with the highest and lowest left-right scores in the established and new democracies, respectively, over the entire period. In addition, the selected countries have high and low percentages of adjacent and peripheral ideological elements, respectively. These are Sweden and the Czech Republic on the one hand, and the USA and Poland 7 on the other. In addition, I have included the UK here because the number of left-right statements there has doubled over the last period. So the data from the UK can show us how adjacent and peripheral ideological elements in a country have changed over a short period of time. Figures 1, 2, 3, 4, and 5 show the intensity with which the adjacent and peripheral statements of the party manifesto statements correlate with the core LR index. 8 The red entries show left-aligned statements and the blue entries right-aligned statements. The deeper the color, the more clearly the statements belong either to the left or to the right. Figure 1 shows a very consistent pattern for Sweden. Not surprisingly, equality is a left issue during all time periods. The same is true for citizens' participation rights and democracy as well as the favorable mentioning of trade unions. With a brief interruption between 1979 and 1988 this is also true for environmental protection. This is exactly the period when the Swedish Green party entered the political scene in Sweden and moved the environmental issue from the left-right dimension to a policy dimension of its own. However, since the 1990s the environmental dimension has again realigned with the left-right dimension. Furthermore, there is an alignment with the right over the whole period concerning tax incentives for enterprises, and freedom and human rights as well as administrative efficiency. From 1988 to the last election analyzed, European integration is a left-right issue in Sweden where the right is for and the left is against it. This shows that many highly salient issues and even new conflict dimensions are aligned to the core left-right ideology and make the LR-score so extremely relevant in Sweden.
The Czech Republic (Fig. 2) also shows a very stable pattern concerning the issues which align with the left-right dimension as well. The favorable mention of the need for collaboration of employers and trade union organizations in overall economic planning and direction through the medium of tripartite bodies of government (corporatism) is a left issue over the whole period. The same is true for favoring trade unions. The expansion of the welfare state and the support of equality have always been left issues in the Czech Republic until the election in 2017. The populist discourse of the ANO has shifted both issues from left to right (Hanley and Vachudova 2018). Fostering economic growth has been a left issue for a long time as well. Over more than a decade, military issues have become a conflict point framed in the left-right dimension. Interesting is the fact that since the 2010 election a skeptical position toward European integration or a negative standpoint on internationalization have been right issues. In sum, for the Czech Republic, we can conclude that the left-right dimension aligns with many important issues and that national security is an issue that divides left and right. However, in contrast to Sweden there is no other policy dimension integrated into the left-right dimension and recent developments have redefined important left-right issues such as equality and welfare. Consequently, the left-right index is considerably weaker in the Czech Republic than in Sweden.
In the USA, the decentralization of political decision-making has constantly been a right issue, and, only with some exceptions, social equality is a left issue. 9 The need for military spending was for the most of the postwar period a right issue and a critical stance or emphasis on peace was a left issue. This means that US security policy is set in the left and right dimensions. From the late 1970s onwards, welfare state expansion and the improvement of educational provisions have become left issues. Since 1992 and with the election of President Clinton, more and more issues have been integrated into the political semantics of the left in the USA, most recently in the three last elections analyzed here, including the environmental issue. As a result, the left-right axis has grown in importance in the USA over the past few decades (Fig. 3). Since the Trump election campaigns, the emphasis on the "constitutional way of doing things" has become a right-wing issue and internationalization, i.e., support for international cooperation, is seen as negative (see also Lewis 2019).
In Poland (Fig. 4), only a few issues are left-right-aligned. This could be due to the fact that cultural issues (religion) dominate in Poland and override the left-right dimension (Markowski 1997). The left-right dimension lost importance in the 2007 elections, which were overshadowed by the corruption scandal involving the junior coalition party of the ruling PiS party. There is, however, a struggle over support for or rejection of the constitution, which was fought within the left-right dimension in the 2011 elections. In the early years, left-wing statements were accompanied by the expansion of the welfare state and the improvement of educational opportunities. However, this stopped after the 2001 election and even became a right-wing issue. The same is true of the call for centralized political decision-making. The only statement that was a left-wing theme, with few exceptions, was the demand for economic growth. Since 2007, a skeptical attitude toward European integration or internationalization has been part of right-wing ideology in Poland, a trend that can also be observed in other European countries. All in all, the integrated issues tended to have the character of peripheral issues instead of long-lasting adjacent issues. Moreover, the themes frequently switched from left to right or vice versa. This makes the left-right dimension in Poland rather fragmented and explains the low relevance of the left-right dimension in this country. Figure 4 also shows that significantly fewer peripheral statements were assigned to the left-right dimension in the 2007 election, leading to a dramatic decline in the importance of the left-right dimension in Poland. This trend reversed in the 2011 and especially the 2017 elections. In the 2019 election, however, there were again fewer adjacent and peripheral statements, so that Poland as a whole is at roughly the same level of left-right relevance before and after the financial crisis.
In the UK, the expansion of the welfare state has been a left issue most of the time. European integration was debated in terms of left (anti) and right (pro) in 1970s. Since the 1980s, the right integrated issues such as law and order, government efficiency and the reduction of civil servants, military policy and economic incentives in their discourse. The position taken on trade unions is an issue of left versus right in British politics, and the opposing views on this issue are much clearer than in any other country in this study. This shows that the traditional class struggle in which labor organizations play an important role is more alive in British politics than in other countries. Figure 5 shows that the integrative power of the left-right dimension changed in the UK. It gained importance in the 1970s and 1980s, but in the 1990s many topics lost their relevance, such as European integration and questions of equality and the welfare state. However, the recent elections have framed many conflicting issues as left and right: security policy, European politics and issues of centralization. In addition, new issues related to equality, welfare and the environment are highlighted by the left. The data also clearly show that European integration has been a conflicting issue for left and right in recent elections. However, now the right takes an anti-EU position, and the left emphasizes European integration in a positive light. All of this explains the dramatic increase in the left-right dimension in recent years.

Conclusion
The analysis of this article could not confirm an overall decline of the left-right dimension in structuring party competition in most of the 34 democracies. This finding can be revealed if one conceptualizes left and right according to political theory by separating, on the one hand, the ideological core and, on the other hand, the issues that are temporally and spatially aligned with the ideological core. Only such a conceptualization is able to analyze the changing relevance and meaning of the left-right dimension. The findings of such an analysis are highly valuable for a better understanding of the left-right dimension in party politics. Focusing on the questions raised in this article, we have reached some intriguing conclusions.
The analysis shows that the left-right dimension has hardly lost its relevance in the postwar period. However, we see opposing trends when we look at the core ideological elements and adjacent and peripheral elements. The importance of the core ideology has continuously decreased, while the adjacent and peripheral elements have increased, resulting in the relevance of the left-right dimension remaining rather stable. Most intriguing is the fact that in the years following the post-2008 crisis, the left-right dimension has been revived in most of the countries. In fact, the left-right dimension has never been as relevant as it is in the last period. The article also shows that the left-right dimension was strong in the immediate postwar decades, but that it is a myth that this dimension had its heyday during that time. This finding calls into question the starting point of many studies, which postulate a decline in left-right dominance that actually never existed to the extent assumed in the first place. However, there are regional and country-specific differences. In particular, I was able to confirm that the left-right dimension is much less relevant in CEE than in established democracies.
Another finding that I could only touch on is the changing meaning of left and right. Adjacent issues are increasingly shaping left and right. In many countries, new political issues such as environmental concerns have been integrated into the left-right dimension. The ideological position of European integration, which has undergone a fundamental transformation in recent years, must also be seen in this context. For a long time, supporting European integration has been an integral part of the right ideology. However, in the new millennium this changed fundamentally. Either European integration has no ideological leaning in terms of left and right anymore or, even more common, right ideology is now associated with a skeptical view on European integration. The Brexit strategy of the Conservative Party may be a case in point, but this trend is also visible in many other countries. An anti-European integration position is associated with the right ideology in the Czech Republic, Poland, Germany, the Netherlands, Slovakia, Iceland and Austria.
The analysis in this article sheds new light on the relevance and changing meaning of the left-right dimension. However, it could not address the reasons for these changes (structural changes, populism, new issues, etc.). For an elaboration of these points, macrocomparative studies focusing on specific causal mechanisms on ideological changes (Adams et al. 2004;Düpont 2017;Lacewell 2017;Zohlnhöfer and Bandau 2020;Schmitt and Franzmann 2020) should integrate the changing meaning of left and right in their analyses. Furthermore, country-specific case studies are needed which trace the ideological developments in particular countries (see for instance Lewis 2019). In this article, I could only use the statements given by the party manifesto project, and it would be desirable to broaden the range of topics by recoding the empirical material (Düpont and Rachuj 2021). Nevertheless, I was able to provide evidence that the left-right ideology has been able to integrate new issues. In this respect, the analysis in this article compliments studies which depart from the left-right dimension by defining a new policy space (Hooghe and Marks 2018;Kitschelt 1994;Kriesi et al. 2012). In this context, it is important to note that none of the studies that identify a new or specific policy space have ever tested whether the policy space has actually changed or how long the newly identified policy space will last. The data from such studies cover only about a decade, making it impossible to draw conclusions from an assumed decline in the left-right dimension. In this regard, the analysis in this article is more apt to draw such conclusions, as it analyzes 34 countries over eight decades. Nevertheless, the analysis of this article does not challenge the findings of those studies but focuses on the other side of the coin and warns that we should not throw out the baby (the left-right dimension) with the bathwater (with temporally emerging new political issues). 10 The initial step should be to analyze the capability of integrating new issues by well-established societal conflict dimensions. As political theory tells us, the 10 At the end of the day, it is a matter of taste to continue to refer to a modified left-right dimension as a left-right dimension or to find another name for it. However, the empirical results in this article show that it is difficult to identify a clearly new political dimension for all countries. With respect to the explanations of political theory on the development of ideologies, the approach proposed here seems to me to be at least as legitimate, if not more appropriate, than talking about new political dimensions. essence of an ideology is that it changes over time and adjusts to specific circumstances. To this aim, the analysis in this article is the beginning of a more sophisticated empirical analysis of political ideology rather than the last word.