Date Published: February 27, 2019
Publisher: Public Library of Science
Author(s): Ingo Rohlfing, Tobias Schafföner, María Carmen Díaz Roldán.
Does economic globalization influence the positioning of parties and, as a consequence, the ideological characteristics of party systems? Answering this question is important because we need to understand the constraints that parties face in formulating policies from which voters have to choose. In our paper, we take a systemic perspective and conceptualize a party system’s ideological center of gravity as the outcome of interest. We define the center of gravity as the weighted mean position of all parliamentary parties in a country that represents the position to which parties gravitate. We start by formulating static hypotheses on the effect of imports and exports on the center of gravity and derive their underlying mechanisms. We further derive dynamic hypotheses stipulating varying effects over time based on the premise that partisan attitudes toward globalization have undergone multiple changes over the last decades. A time-series cross-section analysis of 129 elections in 15 Western European countries from 1974 to 2015 finds evidence for opposite effects of exports and imports in the pooled data. Additionally, a moving-window analysis indicates that the relationship between globalization and the center of gravity varies over time. This is a significant finding because it suggests that economic globalization has an influence on party systems and that it is important to test for time-varying effects.
Political parties made economic globalization possible by reducing barriers to the free flow of goods and services . However, did its promotion backfire by influencing the ways in which parties position themselves on national and international economic issues and, as a consequence, the ideological characteristics of party systems? There is extensive research on how globalization is related to political, social and economic output and outcomes and party positioning [2–4], but there have been relatively few studies on its relationship with party systems. Research on globalization effects on party positioning yields contradictory evidence. There is evidence of an “efficiency effect” and rightward shifts of parties , probably moderated by the median voter , and for a “compensation effect” and leftward shifts of party positions [7, 8]. These studies present valuable insights, but they do not offer information on the systemic effects of globalization because of the fallacies of inferring effects on party systems from globalization effects on individual parties. The effects of economic globalization on party systems work through the positioning of individual parties but, to study these effects, the empirical analysis must center on party systems and cannot be inferred from a party-level analysis. An analysis of the relationship between globalization, party positioning and their implications for the characteristics of party systems is of societal importance. Economic globalization could affect party positioning and, by implication, the ideological characteristics of party systems and policies offered to voters on a systemic level. Depending on how globalization affects party systems, it might indirectly and negatively shape citizen satisfaction with those parties and with democracy as a whole .
Our analysis focuses on party systems and their ideological center of gravity (henceforth COG). Following conventional usage, we use ‘party ideology’ synonymously with a party’s position on a dimension of party competition, which is the economic dimension in our paper. That position depends on how the party positions itself on the issues or policies that constitute the dimension . Gross and Sigelman  introduced the COG as one element of party systems that has received scant attention compared to other measures such as fractionalization and polarization . We define the COG as the mean positioning of parties on a specific dimension, with party positions weighted by their vote shares. We use the vote share instead of the parliamentary seat share to avoid distorting effects deriving from the disproportionality of the electoral system. At the heart of our analysis is the economic dimension of party competition because this is where we expect to observe the consequences of economic globalization. The COG conveys information about the hypothetical point on the economic dimension to which parties gravitate. The COG is an important measure because it indicates whether parties take positions such that the center is located more on the left, right or middle of the economic dimension and the hypothetical point on the dimension from which voters can choose among parties. We are aware of the argument that economic globalization made parties more responsive on a non-economic or social dimension because of globalization constraints on their behavior on the economic dimension [6, 13]. If this argument was true, there might still be an effect of globalization on the COG on an economic dimension. The argument about the decreasing room of maneuver is tied to the claim that party positions converge in a specific range of the economic dimension and the COG should shift in response to globalization. Globalization effects on party positioning on a social dimension, which are not our theoretical interest, are therefore compatible with an effect of globalization on the COG.
We present the estimates for imports and exports in Fig 2 and present the full regression results in Table 5, column 1. The estimate for imports is negative and significant at 0.05 and allows us to reject the null hypothesis of no association. The estimate for exports is positive and significant at 0.01, unexpectedly leading us to reject hypothesis 2 stipulating no association between export levels and the COG. Both estimates have in common that the confidence interval is relatively wide and the uncertainty large.
We test hypotheses 3 to 7 with a moving-window analysis estimating the OLS regression with two-way clustered standard and a lagged DV for subsets of the total period of analysis (Fig 3). Each window covers 20 years and is moved forward by one year, taking 1974 as the starting point. The number of observations increases from 53 for the first window up to 68 for the last window because data availability improves for more recent periods. We do not control for multiple testing because the windows have been determined in an exploratory manner as their size and number do not follow from theory. The results are robust to changes in the size and number of windows.
The empirical results show that economic globalization is associated with the development of a party system’s center of gravity, but the link is different across dimensions of globalization and time. We summarize our findings in Table 6 and compare them with the hypotheses we formulated in Section 3. The two hypotheses that fail to be supported concern the static hypothesis for exports and the hypothesis on imports during the first of our three periods. Our hypothesis on the absence of an effect of exports on the COG was based on the premise that both the parties on the left and right value rising exports, but promote them with different policies, which results in no significant change of the COG on the systematic level. The positive effect of exports suggests that this premise is incorrect. Three non-exclusive reasons can account for this. First, left and right parties propose different policies to foster exports and right parties do this more strongly than left parties, leading to a rightward shift of the COG. A possible reason for this could be that exporting companies might be more influential and more often among the constituency of right parties, which therefore seek export-promoting polices more than left parties. Second, contrary to the compensation hypothesis, changing expectations of voters are also conceivable: Workers in the export sectors who benefit from rising exports by occupying high-skilled jobs and rising wages could demand policies fostering the reason for their economic success that are picked up by party platforms. Third, left parties and right parties might pursue similar, supply-side policies when it comes to the promotion of exports.
Political parties are involved in making decisions about the trajectory of globalization. We reverse the perspective and present results indicating that imports and exports, as the two most important facets of economic globalization, might influence the positions that parties take on an economic dimension of party competition. If correct (see previous section), this has important empirical consequences for voter representation because the average positioning of parties reflects the options from which voters can choose. On a more general level, our study indicates that, as two other salient concepts in the globalization literature, the COG and convergence/divergence are best considered in conjunction to formulate a complete picture.