Supersummary
Week 1
Article 1: Three worlds of welfare capitalism or more? A state-of-the-art report – Arts and
Gelissen, Review of Esping-Andersen
Main argument Esping-Andersen: The welfare state cannot be regarded as the sum total of
social policies, it is more than a numerical cumulation of discrete programmes. Welfare states
can be clustered together in three distinct subclasses.
Independent variable(s):
Degree of decommodification
Type of social stratification (and solidarities)
Historical and political forces:
the nature of class mobilization
class-political action structures
the historical legacy of regime institutionalization
Dependent variable(s): Welfare state typology
Results/Conclusions: We conclude that real welfare states are hardly ever pure types and are
usually hybrid cases; and that the issue of ideal-typical welfare states cannot be satisfactorily
answered given the lack of formal theorizing and the still inconclusive outcomes of
comparative research.
Based on the degree of decommodification and kind of social stratification one can identify 3
models or ideal-types of welfare states: liberal, conservative, social-democratic. These ideal-
types owe their origins to different historical forces and they follow qualitatively different
developmental trajectories. Contrary to the ideal world of welfare states, the real world is
likely to exhibit hybrid forms.
,Article 2: The Politics of Welfare State Retrenchment: A Literature Review - Peter Starke
Main argument: I argue that, because of the nature of these persisting issues (see below), the
pluralistic dialogue between different methods and approaches – as well as their combination
– remains the most promising way forward in the study of welfare state politics.
Problems: The debate on the relevance of political parties and ideas, by contrast, is still far
from settled. Further unresolved issues include the nature of the dependent variable in
retrenchment studies. Also, the exact motives for cutbacks are theoretically still little
understood, as are the political mechanisms through which they are realized.
Article 3: The Dependent Variable Problem within the Study of Welfare State Retrenchment:
Defining the Problem and Looking for Solutions - Christoffer Green-Pedersen
Main argument: Different research questions and theoretical perspectives should lead to
different conceptualizations of retrenchment and consequently also to different measurements
of it.
Results/Conclusions: When looking at the debate about welfare state retrenchment, two
different theoretical perspectives on welfare state retrenchment are prominent, namely
retrenchment as cutbacks in people’s welfare entitlements and retrenchment as institutional
change. Differently put, welfare state retrenchment can be conceptualized in two different
ways, namely as either cuts in entitlements or changes in institutional characteristics. These
two conceptualizations are not unrelated but the same changes may be evaluated very
differently from the two perspectives and, therefore, they should be kept apart. Which
conceptualization of retrenchment to use for a specific study is dependent on one’s theoretical
perspective. Thus, the main solution to the dependent variable problem is to be very clear-cut
about one’s theoretical perspective and research question.
Article 4: Panel data analysis in comparative politics: Linking method to theory - Thomas
Plümper, Vera E. Troeger & Philip Manow
Main argument: Re-analyzing a study of Garrett and Mitchell ‘Globalization, government
spending and taxation in the OECD’, we demonstrate that Garrett and Mitchell’s findings are
not robust. Instead, we show that partisan politics and socioeconomic factors such as aging
and unemployment as expected by theorists have a strong impact on the timeseries and cross-
sectional variance in government spending.
Problem: the choice of panel data analysis has often not been theoretically justified, but rather
bypassed with a general reference to the ‘standard’ that has emerged in the field over the last
years.
Independent variable(s):
Dependent variable(s):
Results/Conclusions: This article addresses four potential sources of problems in panel data
analyses with a lagged dependent variable and period and unit dummies (the de facto Beck-
Katz standard).These are: absorption of cross-sectional variance by unit dummies, absorption
, of time-series variance by the lagged dependent variable and period dummies, mis-
specification of the lag structure, and neglect of parameter slope heterogeneity. Based on this
discussion, we suggest substantial changes of the estimation approach and the estimated
model.
The preceding discussion led to the following results. First, and partly contrary to what many
comparative political economists believe, unit fixed effects turn out to be problematic if
variables are time invariant or if the theory at test predicts level effects. Second, the inclusion
of a lagged dependent variable and/or period dummies tends not only to absorb large parts of
the trend in the dependent variable, but likely biases estimates. Third, simply assuming a
uniform lag structure may cause biased estimates and wrong inference. Fourth, if the time
dimension in panel analyses exceeds a rather limited number of time periods, it becomes
extremely important to think about and test for structural changes in slopes and error variance.
This is especially true for the impact of political institutions and actors. We think it has
become clear that the results derived from panel data analysis critically depend on a host of
crucial methodological decisions and theoretical assumptions which the use of a ‘standard’
like the de facto Beck-Katz method cannot avoid, but only hide.
First and foremost, we found that partisan effects matter. However, party preferences’
influence on government spending is not stable over time. Second, our results say that
unemployment and the aging of the society tend to put an upward pressure on government
budgets, while growth reduces the government share of the economy. Third, international
economic openness does not seem to have a similarly important influence on government
spending. We would like to emphasize that these findings stand in contrast to Garrett’s and
Mitchell’s results gained on the basis of the same data, but with a different methodological
approach.
Week 2
Article 1: The New Politics of the Welfare State – Paul Pierson 3
Main arguments:
The welfare state is resilient to retrenchment because of the welfare state’s enduring
popular support and its institutional inertia. -> path dependence.
The new context leads politicians to refrain from radical and highly visible cuts, to
pursue, instead, ‘the politics of blame avoidance’ in order to pursue retrenchment.
Under the conditions of the ‘new politics of the welfare state’ the explanatory
relevance of parties and left power resources had faded.
Retrenchment is not simply the mirror image of welfare state expansion.
Problem: One can’t use the same framework to explain retrenchment of the welfare state after
the 1980’s, because there are more narrow political actors and ‘new politics of the welfare
state’. Theories about the ‘old politics’ of expansion, including socio-economic functionalism
and class-based power resources theory, fail to account for developments after the end of the
‘Golden Age’.
Independent variable(s): Rather than emphasizing cuts in spending per se, the focus is on
reforms that indicate structural shifts in the welfare state.
Week 1
Article 1: Three worlds of welfare capitalism or more? A state-of-the-art report – Arts and
Gelissen, Review of Esping-Andersen
Main argument Esping-Andersen: The welfare state cannot be regarded as the sum total of
social policies, it is more than a numerical cumulation of discrete programmes. Welfare states
can be clustered together in three distinct subclasses.
Independent variable(s):
Degree of decommodification
Type of social stratification (and solidarities)
Historical and political forces:
the nature of class mobilization
class-political action structures
the historical legacy of regime institutionalization
Dependent variable(s): Welfare state typology
Results/Conclusions: We conclude that real welfare states are hardly ever pure types and are
usually hybrid cases; and that the issue of ideal-typical welfare states cannot be satisfactorily
answered given the lack of formal theorizing and the still inconclusive outcomes of
comparative research.
Based on the degree of decommodification and kind of social stratification one can identify 3
models or ideal-types of welfare states: liberal, conservative, social-democratic. These ideal-
types owe their origins to different historical forces and they follow qualitatively different
developmental trajectories. Contrary to the ideal world of welfare states, the real world is
likely to exhibit hybrid forms.
,Article 2: The Politics of Welfare State Retrenchment: A Literature Review - Peter Starke
Main argument: I argue that, because of the nature of these persisting issues (see below), the
pluralistic dialogue between different methods and approaches – as well as their combination
– remains the most promising way forward in the study of welfare state politics.
Problems: The debate on the relevance of political parties and ideas, by contrast, is still far
from settled. Further unresolved issues include the nature of the dependent variable in
retrenchment studies. Also, the exact motives for cutbacks are theoretically still little
understood, as are the political mechanisms through which they are realized.
Article 3: The Dependent Variable Problem within the Study of Welfare State Retrenchment:
Defining the Problem and Looking for Solutions - Christoffer Green-Pedersen
Main argument: Different research questions and theoretical perspectives should lead to
different conceptualizations of retrenchment and consequently also to different measurements
of it.
Results/Conclusions: When looking at the debate about welfare state retrenchment, two
different theoretical perspectives on welfare state retrenchment are prominent, namely
retrenchment as cutbacks in people’s welfare entitlements and retrenchment as institutional
change. Differently put, welfare state retrenchment can be conceptualized in two different
ways, namely as either cuts in entitlements or changes in institutional characteristics. These
two conceptualizations are not unrelated but the same changes may be evaluated very
differently from the two perspectives and, therefore, they should be kept apart. Which
conceptualization of retrenchment to use for a specific study is dependent on one’s theoretical
perspective. Thus, the main solution to the dependent variable problem is to be very clear-cut
about one’s theoretical perspective and research question.
Article 4: Panel data analysis in comparative politics: Linking method to theory - Thomas
Plümper, Vera E. Troeger & Philip Manow
Main argument: Re-analyzing a study of Garrett and Mitchell ‘Globalization, government
spending and taxation in the OECD’, we demonstrate that Garrett and Mitchell’s findings are
not robust. Instead, we show that partisan politics and socioeconomic factors such as aging
and unemployment as expected by theorists have a strong impact on the timeseries and cross-
sectional variance in government spending.
Problem: the choice of panel data analysis has often not been theoretically justified, but rather
bypassed with a general reference to the ‘standard’ that has emerged in the field over the last
years.
Independent variable(s):
Dependent variable(s):
Results/Conclusions: This article addresses four potential sources of problems in panel data
analyses with a lagged dependent variable and period and unit dummies (the de facto Beck-
Katz standard).These are: absorption of cross-sectional variance by unit dummies, absorption
, of time-series variance by the lagged dependent variable and period dummies, mis-
specification of the lag structure, and neglect of parameter slope heterogeneity. Based on this
discussion, we suggest substantial changes of the estimation approach and the estimated
model.
The preceding discussion led to the following results. First, and partly contrary to what many
comparative political economists believe, unit fixed effects turn out to be problematic if
variables are time invariant or if the theory at test predicts level effects. Second, the inclusion
of a lagged dependent variable and/or period dummies tends not only to absorb large parts of
the trend in the dependent variable, but likely biases estimates. Third, simply assuming a
uniform lag structure may cause biased estimates and wrong inference. Fourth, if the time
dimension in panel analyses exceeds a rather limited number of time periods, it becomes
extremely important to think about and test for structural changes in slopes and error variance.
This is especially true for the impact of political institutions and actors. We think it has
become clear that the results derived from panel data analysis critically depend on a host of
crucial methodological decisions and theoretical assumptions which the use of a ‘standard’
like the de facto Beck-Katz method cannot avoid, but only hide.
First and foremost, we found that partisan effects matter. However, party preferences’
influence on government spending is not stable over time. Second, our results say that
unemployment and the aging of the society tend to put an upward pressure on government
budgets, while growth reduces the government share of the economy. Third, international
economic openness does not seem to have a similarly important influence on government
spending. We would like to emphasize that these findings stand in contrast to Garrett’s and
Mitchell’s results gained on the basis of the same data, but with a different methodological
approach.
Week 2
Article 1: The New Politics of the Welfare State – Paul Pierson 3
Main arguments:
The welfare state is resilient to retrenchment because of the welfare state’s enduring
popular support and its institutional inertia. -> path dependence.
The new context leads politicians to refrain from radical and highly visible cuts, to
pursue, instead, ‘the politics of blame avoidance’ in order to pursue retrenchment.
Under the conditions of the ‘new politics of the welfare state’ the explanatory
relevance of parties and left power resources had faded.
Retrenchment is not simply the mirror image of welfare state expansion.
Problem: One can’t use the same framework to explain retrenchment of the welfare state after
the 1980’s, because there are more narrow political actors and ‘new politics of the welfare
state’. Theories about the ‘old politics’ of expansion, including socio-economic functionalism
and class-based power resources theory, fail to account for developments after the end of the
‘Golden Age’.
Independent variable(s): Rather than emphasizing cuts in spending per se, the focus is on
reforms that indicate structural shifts in the welfare state.
