Feedback Effects in Interregional Input-Output Models : What Have We Learned?

There has been a long-standing tradition, prompted by a 1963 paper by Ronald E. Miller, in the potential size orf the feedback effects on activities in a region arising intially from a stimulus in that region but due to its linkages through its trade with another region (or regions). Early results based on numerical simulations suggested that these feedback effects might be quite small in practice. This lead some researchers to question the efficacy and cost-effectiveness of constructing large and expensive interregional and multiregional input-output models for many applications. In light of reductions in the cost of computing power, researchers have gained a renewed interest in expanding input-output accounts not only multiregionally but also to endogenize more behavior (e.g., migration, income and capital flows, and flexible prices). But Miller's analytical findings remain relevant. This paper re-exmines the evidence on feedback effects and offers some assessment of what we have learned about them during the past 30 years