Acceso usuarios
Subjective measures used in models of political choice are typically open to criticism with respect to their endogeneity. Economic perceptions have been subject to particular criticism in this respect. We address these concerns through the application of estimation procedures introduced from econometrics and the introduction of a new measure of economic perceptions that acts as an instrumental variable. Together, these eliminate simultaneity bias and bias due to unobserved heterogeneity, reduce omitted-variable bias, and reduce noise/measurement error in economic perceptions. An analysis of three-wave panel surveys produces estimates of the effect of economic perceptions on party evaluation that are not biased upward by the presence of endogeneity and that help address the discrepancy between competing subjective models of economic effects on government approval.
