GOODHART'S LAW AND MACHINE LEARNING: A STRUCTURAL PERSPECTIVE

Christopher A. Hennessy and Charles A. E. Goodhart

International Economic Review, 2023, vol. 64, issue 3, 1075-1086

Abstract: We develop a simple structural model to illustrate how penalized regressions generate Goodhart bias when training data are clean but covariates are manipulated at known cost by future agents. With quadratic (extremely steep) manipulation costs, bias is proportional to Ridge (Lasso) penalization. If costs depend on absolute or percentage manipulation, the following algorithm yields manipulation‐proof prediction: Within training data, evaluate candidate coefficients at their respective incentive‐compatible manipulation configuration. We derive analytical coefficient adjustments: slopes (intercept) shift downward if costs depend on percentage (absolute) manipulation. Statisticians ignoring manipulation costs select socially suboptimal penalization. Model averaging reduces these manipulation costs.

Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://doi.org/10.1111/iere.12633

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wly:iecrev:v:64:y:2023:i:3:p:1075-1086

Ordering information: This journal article can be ordered from
http://www.blackwell ... bs.asp?ref=0020-6598

Access Statistics for this article

International Economic Review is currently edited by Michael O'Riordan and Dirk Krueger

More articles in International Economic Review from Department of Economics, University of Pennsylvania and Osaka University Institute of Social and Economic Research Association 160 McNeil Building, 3718 Locust Walk, Philadelphia, PA 19104-6297. Contact information at EDIRC.
Bibliographic data for series maintained by Wiley Content Delivery ().