Mixing Methods – Blending Theories. The Dual Linkage in Policy Analysis

Mixed methods – the deliberate and pragmatic application of both qualitative and quantitative methods in order to answer a single research question in the most adequate way possible – finds ever more approval and application in the social sciences. Yet on the theoretical side and far from mixing, policy theories seem to diverge between the multi-causal and process-oriented frameworks like Punctuated Equilibrium, Advocacy Coalition, or Multiple Streams and mono-causal theoretical departures emphasizing the single or joint importance of certain economic, societal, political, or institutional factors. Aside distinct explanations, the two strands of theories imply a certain methodological proceeding as well. Process-oriented theories naturally favor qualitative approaches more suitable to trace the many details of political processes. In contrast, mono-causal effects are often explored via quantitative analyses facilitating the simultaneous testing of various single variables. In mixed method research, however, and although quantitative and qualitative approaches may call for different theories, the single parts of a study addressing the same research question should not rely on an entirely different theoretical background. To avoid a skewed comparison resulting from incongruous theories the literature strongly recommends a certain parallelism in the theoretical concepts. The paper follows this general argument and seeks to explore the explanatory potential for public policy analysis in both mixing methods and blending theories across traditional divides in order to arrive at empirical conclusions uncrippled by theoretical or methodological limits. Two questions are obtruding in this context: How can the needs of mixed methods be satisfied without distorting the applied theories? And what is gained from such a dual linkage for public policy analysis? These questions are addressed in three consecutive steps: First, the general argument for a theoretical nexus in mixed methods is developed, discussed, and reviewed with regard to the emerging literature on mixed method research. Second – and in an exemplary fashion – two policy theories, one process-, the other one variable-oriented, are merged on the basis of hypothesis building. For this, single political-institutional factors (e.g. parties, organized interests, or veto players) are blended with an updated version of the Multiple Streams Framework. Third, based on a mixed-method analysis of law and order legislation in the United Kingdom between 1994 and 2014 it is illustrated to what extent the quantitative analysis of single political-institutional variables in conjunction with a variable-oriented (and thus comparable but qualitative) utilization of the Multiple Streams Framework offers an increased explanatory potential. In sum, the paper features three distinct contributions: Generally, the study sketches a unique approach toward mixing methods and blending theories in order to arrive at more reliable causal explanations in public policy research. Specifically, the paper develops further the already pioneered fruitful link between single political-institutional factors and the Multiple Streams Framework by emphasizing its potential in mixed method research. Empirically, the approach is tested in a mixed-method analysis of British law and order legislation highlighting the explanatory potential gained from mixing methods and blending theories.