Close window
 MPIfG Discussion Paper 02/1

Bernhard Kittel and Herbert Obinger


Political Parties, Institutions, and the Dynamics of Social Expenditure in Times of Austerity


 

Abstract


 
The containment of social expenditure growth has been (and still is) a core issue of public policy in advanced industrial countries since the 1980s and has received much academic attention during that period. Among the most extensively discussed explanatory factors of social expenditure are partisan politics and political institutions, as well as the dependency of the real impact of the former on the latter. The paper distinguishes five competing theoretical perspectives and explores their power to explain the empirical variation in the period 1982-1997 in 21 OECD countries. The empirical analysis of short-term dynamics is performed in a time-series cross-section framework while long-term level effects are explored in a cross-sectional setting. By using an interactive model specification the authors show that there is empirical evidence for this conditional effect, albeit it is neither thoroughly convincing nor leading to lasting long-term level effects. Extensive specification tests show that the 1990s witnessed a weakening of partisan effects which were still present in the 1980s. In total, the evidence tends to give most support to the "growth-to-limits" and the "new politics" perspectives.
 

Zusammenfassung


 
Die Beschränkung des Wachstums der Sozialausgaben war (und ist) ein zentrales Anliegen staatlicher Politik in fortgeschrittenen Industrieländern seit den 1980er Jahren und hat in dieser Periode in der akademischen Debatte große Aufmerksamkeit erhalten. Zu den meistdiskutierten Erklärungsfaktoren zählen parteipolitische Differenzen und politische Institutionen sowie die Abhängigkeit ersterer von letzteren. Das Discussion Paper unterscheidet fünf verschiedene Theorieansätze und untersucht ihren Beitrag zur Erklärung der empirischen Varianz im Zeitraum von 1982 bis 1997 in 21 OECD-Ländern. Die Analyse der kurzfristigen Veränderung erfolgt im Rahmen einer kombinierten Zeitreihen-Querschnittanalyse, während die langfristigen Niveauunterschiede mittels einer Querschnittanalyse untersucht werden. Mit Hilfe einer interaktiven Modellspezifikation zeigen die Autoren, dass empirische Belege für diesen Konditionaleffekt vorhanden sind, die allerdings weder vollkommen überzeugen noch langfristige Niveaueffekte zeitigen. Ausführliche Spezifikationstests deuten darauf hin, dass die Parteieneffekte, die in den 1980er Jahren vorhanden waren, in den 1990er Jahren deutlich schwächer wurden. Insgesamt stützen die Befunde die These des "Wachstums zu Limits" und die These der "Neuen Politik" am ehesten.
 

Contents


 
1 Introduction
2 Politics and the Welfare State: Theoretical Accounts
3 Politics and the Welfare State: Evidence from Quantitative Research
4 Trends in Welfare Efforts in 21 OECD Democracies, 1980-1997
5 An Empirical Assessment of Welfare State Dynamics
5.1 Approach, Core Variables, and Data
5.2 Pooled Time-Series Cross-Seqction Analysis
5.2.1 Model Specification
5.2.2 Problems of Explaining Levels of Social Expenditure
5.2.3 The Baseline Model: Explaining Growth in Social Expenditure by Leftist and Christian democratic Parties and Institutional Rigidity
5.2.4 Stability Analysis
5.2.4.1 Turning the Perspective: The Impact of Conservative Parties
5.2.4.2 Temporal Stability
5.2.4.3 Dependency of the Findings on the Inclusion of Particular Countries
5.2.4.4 Partisan Complexion of Coalitions
5.2.5 TSCS Findings: Summary and Interpretation
5.3 Cross-Sectional Regression Analysis
5.3.1 Politics and the Growth in Social Expenditure, 1980-1997
5.3.2 Determinants of Net Social Expenditure
5.3.3 Political Determinants of Program-Related Expenditures
5.3.4 Cross-sectional Findings: Summary and Interpretation
6 Conclusion
References
Close window