Dealing with the common econometric problems of count data with excess zeros, endogenous treatment effects, and attrition bias

In this article, the authors propose an approach to investigate regional economic structure, sectoral productivity, and relative efficiency by using a modified shift-share model and data envelopment analysis (DEA). To illustrate the proposed approachʹs usefulness, the authors applied it in two states to assess and compare the foundations of economic performance in manufacturing. The authors first investigate the impacts of productivity and output change on employment change in these states by employing the Haynes and Dinc extension of the shift-share model. The DEA method is then employed to diagnose the efficiency of potential lead sectors.