Typology Construction for Comparative Country Case Study Analysis of Patterns of Growth in Sub-Saharan Africa

Abstract

This study has been motivated by the limitations of cross-country regressions and unstructured comparative case studies in providing policy-relevant findings on the determinants of patterns of growth. It presents a methodology to improve upon existing comparative case study research by situating cases withing a typological framework and subsequently using cluster analysis to improve the matching of cases with respect to a number of ‘weakly exogenous’ variables. Such an approach performs a taxonomic function, distinguishing different types of cases and an explanatory function by facilitating the comparison of similar cases in terms of variables in the typology (‘like with like’ comparisons) or of cases with one or more known differences with respect to these variables. The approach is illustrated using data on poverty and growth in SSA and uncovers a number of good comparator cases situated within a typological framework for subsequent comparative analysis.