Como medir o invisível? Guerras, pizzarias do Pentágono e o uso de variáveis proxy em econometria
Guilherme Vianna, Victor Rangel
TL;DR
The paper tackles identification when key economic drivers are latent by formalizing how omitted latent factors induce bias and how proxy variables can mitigate it. It develops a practical protocol based on four properties—relevance, conditional sufficiency, exogeneity, and stability—to evaluate proxies, and demonstrates the approach with a micromobility proxy around Arlington as a latent signal for geopolitical risk (GPR), modeled via cointegration and a bivariate VEC. The analysis shows that a perfect proxy can completely remove the bias, while imperfect proxies leave a residual bias that is identifiable in direction and significance, with the proxy coefficient attenuated by the proxy quality. The work provides a theory-grounded, actionable method for practitioners to measure and interpret latent factors using observable proxies, with implications for policy-relevant empirical work where direct measurement is infeasible.
Abstract
Many economically relevant variables (risk, confidence, uncertainty) are latent and therefore not directly observable, which creates identification challenges in applied regressions. This text formalizes how omitting latent factors generates omitted-variable bias and discusses when including a proxy variable can mitigate it. We distinguish the case of a perfect proxy, which can eliminate the bias, from the more realistic case of an imperfect proxy, where residual bias remains and the estimated effect is attenuated. We propose a practical evaluation protocol based on four properties: relevance, conditional sufficiency, exogeneity, and stability. As an illustration, we use micromobility data from Arlington together with the U.S. Geopolitical Risk Index, estimating cointegration and a bivariate VEC model to interpret local activity as a high-frequency signal of the latent component of geopolitical tension.