BREAKING DOWN THE CHAIN OF RESPONSIVENESS: ASSESSING TO WHAT EXTENT INDIVIDUAL REDISTRIBUTIVE PREFERENCES TRANSLATE INTO SOCIAL POLICY OUTCOMES. A COMPARATIVE ANALYSIS.

This dissertation investigates how well political parties, especially those holding executive positions, achieve policy congruence by translating voters’ redistributive preferences into consistent social policy outcomes. It addresses this topic by conducting large-N studies with the aid of quantitative techniques. It is made up of six chapters. Chapter I provides the theoretical background. In detail, it defines the meaning of the elusive concept of political representation; it discusses the expressive and the instrumental functions that political parties are expected to perform in contemporary representative democracies and recalls party government theories formulated from the 1970s. This literature review allows reorganizing the core stipulations for party government to emerge in a unified theoretical framework. Chapter II begins the journey along the ideal chain of responsiveness from voters' redistributive preferences to actual social policy outcomes. In detail, it focuses on the first link of this ideal chain by verifying whether and to what extent the economic conditions people experience in their everyday life, which are largely given by the position they occupy in the labor market, determine their redistributive preferences and political behaviors. Combining individual level data from the European Social Surveys for 23 OECD countries with party level information from the Comparative Manifesto Project Database, this chapter demonstrates that individuals tend to feel closer to political parties which express in their electoral manifestos social policy supplies consistent with their redistributive preferences. These results are obtained trough discrete choice models, instrumental variables and matching statistical techniques. Chapter III deals with the second link of the chain of responsiveness, addressing a perennial question for students of parliamentary democracy, namely how do coalition governments build their policy proposals. In detail, chapter III explores the degree of correspondence between declared cabinet position and the weighted position of cabinet parties as expressed in their electoral manifestos on two separate issues: the traditionally employed left-right scale and a more policy based welfare scale. Results obtained through a time-series cross-section methodology suggest that “the owners of the agenda setting power” over the two scales are different. In particular, on the traditional left-right scale, declared cabinet position is strongly driven by the weighted position of cabinet parties and by that of the formateur party. Conversely, on the welfare scale, declared cabinet position is also affected by the position of the party holding the median legislator in Parliament and by those of the parties expressing the labor and social affairs ministers. In addition, declared cabinet position on the welfare dimension shows a marked tendency to drift rightward with adverse economic conditions. Chapter IV investigates the last link of the chain of responsiveness between governing parties' long lasting ideological preferences and actual social policy outcomes. In particular, it employs panel data on 19 OECD countries from 1985 to 2011 and tests alternative hypotheses on cabinets' ability to shape social policy outcomes through an Error Correction Model. Results demonstrate that the ideological position of the executive on the left-right scale is unable to affect social policy outcomes in the short run, when economic control variables prevail. However, governing parties acquire relevance on the long run. Specifically, when the government coalition moves to the right, there is a negative impact on social expenditure as a whole, on public spending in active (ALMP) and passive (PLMP) labor market policies and on the net unemployment replacement rate. Chapter V zooms in on the last link of the chain of responsiveness. In detail, it develops a single country study on the Italian case over le last 70 years to address two shortcomings affecting comparative works on this issue. The first limitation comes at the empirical level and concerns the ways in which governments’ partisanship uses to be operationalized. The second limitation, instead, is rooted at the theoretical level and refers to the over-simplified idea of political parties as unitary actors. In particular, chapter V takes advantage of two datasets: the first one provides data on the positions of Italian Prime Ministers and parties; the second one reports the policy positions expressed by factions inside Italian parties. Results demonstrates that declared cabinets’ positions on welfare state expansion as expressed in Prime Ministers’ investiture and confidence speeches are strong determinants of social expenditure. However, results prove also that the “agenda setting power” enjoyed by the cabinet is strongly weaken by party politics dynamics (the majority the governing coalition enjoys in the lower chamber and the ideological distance between the actual and the previous government) and, more interestingly, by intra-party dynamics. In particular, the ability to shape social expenditure according to the content of the coalition agreement proves to be strongly undermined by the degree of internal polarization inside the major party of the governing coalition. Finally, the last chapter reviews and discusses the main findings of the thesis. It underlines how the doomsday scenario according to which democracy at the national level has been hollowed out has yet to come. Of course, exogenous socio-economic processes and the complexities of politics weaken parties’ responsiveness toward their voters. However, this thesis reveals that parties still play a role, at least in the social policy domain. Moreover, this chapter has a last look on the overall chain of responsiveness, from voters’ redistributive preferences directly to governmental social policy outcomes, to understand whether citizens’ preferences manage to overcome all the obstacles identified in the previous chapters so that they can be translated into consistent spending choices. In particular, controlling for a complete set of potential external confounders and, more interestingly, for the position of the median legislator in Parliament and for that of the ruling coalition, this last chapter demonstrates that median voter’s preferences in the year t-1 are related to the change in social expenditure in the subsequent year. Finally, to contrast popular frustration with political parties, this chapter proposes a learning model of party government, which sees both voters and parties as able to alter, respectively, their policy preferences and their electoral promises according to the contingent constraints imposed by external reality.