Policy learning in policy domains with value conflicts: the Austrian cases of abortion and Assisted Reproductive Technologies (1).

1 Introduction

This paper addresses the question of how we can account for learning in policy domains, in which according to Sabatier's Advocacy Coalition Framework (ACF, Sabatier 1988, 1998, Sabatier & Jenkins-Smith 1993, 1999) learning is unlikely.

For a better understanding of the policy process the ACF starts from the assumption that "policy subsystems" or policy domains are structured by several - usually two - competing advocacy coalitions (AC). An AC is "composed of actors from various governmental and private organisations which both (a) share a set of normative and causal beliefs and (b) engage in a non-trivial degree of coordinated activity over time" (Sabatier 1998: 103). Policy beliefs again can be divided into deep core values, policy core beliefs and secondary aspects. Deep core values relate to deeply held and rather stable "fundamental normative and ontological axioms" (ibid. 112), which often are even not reflected and span across all policy fields. Policy core beliefs refer to fundamental policy positions, whereas secondary aspects are about instrumental decisions. Both categories of beliefs are particular to a policy subsystem.

Sabatier attributes policy change to three different sources: First, changes in the "real world", i.e. changes in socio-economic conditions, public opinion, systemic governing coalition, or policy outputs from other policy domains; second, turnover in personnel; and third, policy-oriented learning. He defines policy oriented learning as "relatively enduring alterations of thought or behaviour intentions, which result from experience and/or new information and which are concerned with the attainment or revision of policy objectives" (Sabatier 1998: 104). Sabatier considers policy learning in policy core beliefs - in contrast to secondary aspects - as being rather unlikely and requiring external perturbation (Sabatier & Jenkins-Smith 1999: 123).

Sabatier presents several hypotheses, which identify conditions for and the likelihood of policy change and policy learning (Sabatier & Jenkins-Smith 1999: 124). Among other factors, he proposes that policy learning is more likely: in areas related to the natural environment and not the social world (learning hypothesis 3); if there are accepted quantitative data, explanations and indicators of success (learning hypothesis 2).

Available empirical research using the ACF has been predominantly concerned with policy subsystems, which are strongly informed by technical knowledge. Much less has this approach been used to understand policy domains that are characterised by normative or identity issues (Sabatier 1998: 100, 122).

In this paper we present two cases from Austrian life politics, which according to Sabatier's hypothesis are rather unlikely to involve policy learning, because they both belong to the social world and lack a problem perception in terms of quantitative data and explanations. Nevertheless, we can detect learning in both cases. The first case is concerned with the Austrian debate on regulating abortion in the early 1970s, which led to a rather permissive law. The second case deals with the discussion about assisted reproductive technologies (ART) policy in Austria during the 1980s and early 1990s, which culminated in the enactment of the Reproductive Medicine Act (Fortpflanzungsmedizingesetz, FMedG) in 1992, a relatively restrictive regulation.