← Search

American Economic Review Vol. 105 No. 5 2015

Prediction Policy Problems

Jon Kleinberg1; Jens Ludwig2; Sendhil Mullainathan3; Ziad Obermeyer4

1 Cornell University, Ithaca, NY 14853 (e-mail: ) · 2 University of Chicago, 1155 East 60th Street, Chicago, IL 60637 and NBER (e-mail: ) · 3 Harvard University, 1805 Cambridge Street, Cambridge, MA 02138 and NBER (e-mail: ) · 4 Harvard Medical School, Boston, MA 02115 and Brigham and Women's Hospital (e-mail: )

open access

Abstract

Most empirical policy work focuses on causal inference. We argue an important class of policy problems does not require causal inference but instead requires predictive inference. Solving these “prediction policy problems” requires more than simple regression techniques, since these are tuned to generating unbiased estimates of coefficients rather than minimizing prediction error. We argue that new developments in the field of “machine learning” are particularly useful for addressing these prediction problems. We use an example from health policy to illustrate the large potential social welfare gains from improved prediction.

DOI
10.1257/aer.p20151023
Volume
105
Issue
5
Pages
491-495
Language
en
Sources
bibtex:phds-export.bib crossref openalex

Cite