Comparative Welfare State Analysis with Survey-Based Benefit: Recipiency Data: The ‘Dependent Variable Problem’ Revisited

Published date01 September 2013
Date01 September 2013
224 Intersentia
W  O*
is paper aims to contribute to the debate on the ‘dependent variable problem’ in
comparative welfare state analysis, by focusing on ‘ bene t recipiency’ as a hitherto
largely neglected type of indicator for und erstanding cross-national and longitudinal
variation between welfare states.  e pros and cons of the commonly used indicators of
social rights and social ex penditure are summarised in order to provide a background
for a conceptual and methodological a ssessment of the added value provided by the
inclusion of bene t recipiency data in analysis. In addition, we explore the empir ical
potentialities of bene t recipiency data. We present and discuss the resu lts of analyses
of EU-SILC data for the EU 27 countries, using them to ca lculate and compare rates
of acc ess to a se ries of be ne ts among working age populations, as well as the relative
amounts of bene ts rece ived. Ou r main con clusion is that b ene t recipiency data are
a promising source for comparative welfare state analysis. However, like other data
sources, they are not without problems, and more work need s to be done to assess
their value, relative to more commonly used indicators.
Keywords: bene t recipiency; comparative; indicators; social sec urity; welfare state
In cross-national and longitudinal comparative welfare state ana lysis there is an
important debate about the ty pes of indicator that are best employed for measuring
di erences between and developments within welf are states. Until now, analysis of the
* Wim van Oorschot i s Professor of Social Pol icy at the Center for Sociolo gical Researc h (CESO),
University of Leuven, Address: Parkstraat 45, 3000 Leuven, Belgium; tel: 0032.(0);
Comparative Welfare S tate Analysis wit h Survey-based Bene t Recipiency Data
European Jour nal of Social Secu rity, Volume 15 (2013), No. 3 225
so-cal led ‘dependent va riable problem’ has focused most speci cally on the relative
quality and adequacy of two types of indicator: the  rst relating to social rights and
the second to social expenditure. Social rights-based measures compare formal
legislation in countries that establish and regulate citizens’ access a nd entitlements
to social provisions (e.g. Esping-Andersen 1990; Scrug gs and Allan 2006), while
expenditure-based measures compare how much states spend in any year on social
provision (in total and by ty pe of bene ts and serv ices) (e.g. Castles 2002). In a recent
volume, which discuss es various aspects of the ‘de pendent variable problem’ in detail,
these two ty pes of indicators are given prime attention (Clasen and Siegel 2007).
However, there is recognition in the literature of a third type of indicator relating
to the number of citizens who receive (speci c types of) bene ts that might be
helpful and might complement the two conventional i ndicators. In Peter Flora’s early
comparative study of European welf are states over the period 1950–1980, it was argued
that information on ‘welfare clienteles’ referring to their (relative) numbers and the
(relative) amount of bene t they receive is essentia l for generating a coherent account
of the character and mag nitude of welfare states (Flora 1986: xxx i).  e study ’s analyses
of administrat ion-based ‘case load’ numbers of rec ipients of a range of social secu rity
bene ts demonstrated that, as a general trend, welfare state protect ion expanded
steadily over the period in all countries, while some particular di erences in levels
of welfare protection between groups of countries became visible (the largest welfare
clienteles were identi ed in the Scandinavian countries, followed by West-European
countries, and the smallest clienteles in the Southern European countr ies). It was
clear that caseload data cou ld be used to detect the overall trends in, and di erences
between, welfare states that characterised thisGolden Age period of Western-
European welfare state development. However, since then, very few comparative
welfare state studies have ta ken up this approach by using welfare clientele or bene t
recipiency dat a as indicators; t he most likely reason for this is that such data is hard to
access for large numbers of countries and years. Notable exceptions are Arents et al.
(2002), Immervoll et al.(2004) and OECD (2003a), three stud ies which form a series
in which administrative caseload data are used and analysed to track and compare
trends in social security protection. However, the series touches only brie y upon the
relative value of such data in comparison with social rights and social expenditure
data. A recent comparative study of caseload d ata locates its empirical analysis d irectly
in the context of such an eva luation, focusing particula rly on the relative value of such
data compared to social expenditure data (De Deken and Clasen 2011, 2013).  e
study concludes that caseloads, despite generating some particular methodological
problems, are not necessari ly inferior as an indicator of welfare state change. Bec ause
expenditure-based data are not  awless either, and more importantly because both
types of indic ator may tap into di erent aspects and manifestations of welfa re reforms,
it is concluded that both are disti nctive indicators of welfare state change, telling t heir
own stories, and that t herefore each may be appropriate in its own right for addressing
di erent types of research questions. Final ly, De Deken and Clasen (2011) suggest

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT