Policy analysis with beliefs about identifying restrictions

Información del proyecto

PABIR

Identificador del acuerdo de subvención: 682877

DOI

10.3030/682877

Proyecto cerrado

Fecha de la firma de la CE 22 Agosto 2016

Fecha de inicio 1 Octubre 2016

Fecha de finalización 30 Septiembre 2022

Financiado con arreglo a

• EXCELLENT SCIENCE - European Research Council (ERC)

Coste total

€ 1 090 686,00

Aportación de la UE

€ 1 090 686,00

Coordinado por

UNIVERSITY COLLEGE LONDON
United Kingdom

Periodic Reporting for period 4 - PABIR (Policy analysis with beliefs about identifying restrictions)

Período documentado: 2021-04-01 hasta 2022-09-30

Resumen del contexto y de los objetivos generales del proyecto

This project developed a toolkit of new econometric methods for policy analysis using Structural Vector Autoregressions (one of the leading approaches for conducting empirical macroeconomic policy analysis). The methods allow researchers to relax the strong
- and often dubious - assumptions that were previously made in the literature or to formally incorporate beliefs about the validity of such assumptions. The methods rely on Robust Bayesian analysis, which offers computational advantages and a formal way to incorporate beliefs, while removing the sensitivity to assumptions that plagued existing approaches.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

The project resulted in several articles published in Econometrica, Journal of Econometrics, Quantitative Economics and Journal of Business and Economic Statistics, which developed the methods and illustrated their empirical relevance.

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

One practical implication of the new methods is the suggestion to report set estimators, instead of point estimators. These sets correctly represent what can be learned from the data. The empirical findings in the published articles show that policy conclusions established in the literature can in fact be driven by arbitrary assumptions, thus offering a word of caution to empirical researchers as well as a strong motivation for using the new methods.

image.png