Extract: The specification and estimation of simultaneous-equations models (henceforth SEMs) has a long tradition in economics. Although SEMs were originally established in the field of macroeconomics, various applications can also be found in applied regional science and economic geography. Among many others, one prominent example is the Carlino and Mills (1987) study on the simultaneous evolution of regional population and employment densities, which gave rise to the famous ‘chicken-or-egg’ quest for causality within the framework of regional adjustment models. And indeed, specifying and estimating an SEM has much to do with getting causal relationships right. This is why applied economists and econometricians generally valorize the SEM approach for its capacity to formulate an explicit structural model with more than just one endogenous variable and the statistical power to control for correlated residuals among the individual equations of the system. While the first argument is of crucial importance for the consistency of the estimated model parameters, the second point is mainly concerned with the notion of estimation efficiency.
Mitze, T. & Stephan, A. (2015). “Simultaneous-equations analysis in regional science and economic geography“. In C. Karlsson, M. Andersson, & T. Norman (Eds.), Handbook of Research Methods and Applications in Economic Geography (pp. 158-171). Cheltenham, UK: Edward Elgar Publishing.