Overcoming the Pitfalls of Decommodification and Generosity Indices: Towards a Fsqca-Based Index of Welfare Stateness

As the dependent variable issue continues to be a problem in comparative welfare state research, measurement issues alongside with conceptualization issues remain high on the agenda. This paper tackles these issues by introducing a new and potentially more 'valid' index based on Fuzzy Set Qualitative Comparative Analysis (fsQCA) for measuring welfare stateness. It shows that the leading state-of-the-art welfare state measurement instruments, Decommodification Index and Benefit Generosity Index, suffer from a set of validity problems that stem partly from their conceptualization and partly from their quantitative approach.

Regarding conceptualization, one common drawback lies in the ways in which they rank welfare states. Respective composite indicators of both indices, which are supposed to gauge welfare states in terms of their capacity to deliver the desired ‘outcome’, are problematic as they assess welfare states on a single scale. Given that contemporary welfare states are aimed at providing more than one ‘outcome’ and since those outcomes could appeal to different interests and indeed reflect potentially contrasting political motives, both indices run the risk of allowing for theoretically spurious aggregation.

As for the measurement, especially the Decommodification Index has serious problems with regard to discontinuities in the scoring, treatment of outliers and the aggregation of scores. While the Benefit Generosity Index avoids some of these problems by using more sophisticated mathematical techniques we show that it can remedy only some, but not all of these problems given that the quantitative approach employed in these datasets inevitably relies on uncalibrated data. Thus both indices may fall short of accurately reflecting differences between welfare states.

As an alternative to these instruments we develop an index of welfare stateness grounded in fsQCA. The index focuses on the Bismarckian, Beveridgean and redistributive elements of welfare states, thus providing a multi-dimensional picture of contemporary welfare states. We demonstrate that since QCA rests on calibrating data based on theoretically informed external criteria and allows for the aggregation of scores through logical operators, the Welfare Stateness Index that we postulate circumvents conceptualization and measurement problems associated with the state-of-the-art indices.

This paper, therefore, has two major contributions: First, it introduces a new and comprehensive welfare state index, namely, the Welfare Stateness Index, which represents an improvement over leading indices in terms of both conceptual and also measurement validity. Second, while introducing a new welfare state index, it also attests to the potential of QCA as a method that could be utilized not only in causal inferences and ideal-type analysis but also for measurement and ranking purposes.