Multidimensionality vs parsimony in institutional distance research

Abstract

Purpose – Should we prefer parsimony or multidimensionality when operationalizing institutional distance? In this paper, I explore the debate on whether current institutional distance operationalizations are too narrow, or if adding complexity in the operationalization needlessly complicates an otherwise lean construct. Design/methodology/approach – In order to explore the above question, I (1) test the statistical validity of the most common parsimonious operationalizations of institutional distance; (2) test whether a parsimonious or a multidimensional solution is more appropriate; (3) discuss whether the statistically valid operationalizations of institutional distance are psychometrically sufficient. Findings – The findings of this paper indicate that three of the most commonly used operationalizations for institutional distance do not possess statistical validity, nor converge on the effects of each underlying item. The fourth commonly used operationalization does possess statistical validity. Nonetheless, its unidimensional usage for institutional distance should be carefully weighed against the theoretical design of the study, where questions exists about whether it is psychometric sufficient for different theoretical arguments used in institutional distance research. Originality/value – While institutional distance is considered to be one of the core concepts of IB/M research, results on the topic have been somewhat ambiguous and sometimes contradictory. Simultaneously, several recent calls have highlighted the need for addressing methodological challenges in the field of IB/M. In this paper, I address one of the main methodological challenges for institutional distance research. I illuminate an underexposed debate and argue that multidimensional operationalizations of institutional distance are required, not just to address methodological challenges but to improve theory on institutional distance as well.