Program evaluation as a decision problem

Research output: Contribution to journalArticle

Abstract

I argue for thinking of program evaluation as a decision problem. There are two steps. First, a counselor determines which program (treatment or control) each individual joins, based for example on maximizing the probability of employment or expected earnings. Second, the policymaker decides whether: to assign all individuals to treatment or to control, or to allow the counselor to choose. This framework has two advantages. Individualized assignment rules (known as profiling) can raise the average impact, improving cost effectiveness by exploiting treatment-impact heterogeneity. Second, it accounts system-atically for inequality and uncertainty, and the policymaker's attitude toward these, in the evaluation.

Original languageEnglish (US)
Pages (from-to)141-173
Number of pages33
JournalJournal of Econometrics
Volume125
Issue number1-2 SPEC. ISS.
DOIs
StatePublished - Mar 2005

Fingerprint

Program Evaluation
Decision problem
Cost effectiveness
Cost-effectiveness
Profiling
Join
Assign
Assignment
Choose
Uncertainty
Evaluation
Program evaluation
Politicians
Counselors

Keywords

  • Bayesian econometric
  • Profiling
  • Program evaluation

ASJC Scopus subject areas

  • Economics and Econometrics
  • Applied Mathematics
  • History and Philosophy of Science

Cite this

Program evaluation as a decision problem. / Dehejia, Rajeev.

In: Journal of Econometrics, Vol. 125, No. 1-2 SPEC. ISS., 03.2005, p. 141-173.

Research output: Contribution to journalArticle

Dehejia, Rajeev. / Program evaluation as a decision problem. In: Journal of Econometrics. 2005 ; Vol. 125, No. 1-2 SPEC. ISS. pp. 141-173.
@article{5b53ba3186e14a5f83816f372bf98761,
title = "Program evaluation as a decision problem",
abstract = "I argue for thinking of program evaluation as a decision problem. There are two steps. First, a counselor determines which program (treatment or control) each individual joins, based for example on maximizing the probability of employment or expected earnings. Second, the policymaker decides whether: to assign all individuals to treatment or to control, or to allow the counselor to choose. This framework has two advantages. Individualized assignment rules (known as profiling) can raise the average impact, improving cost effectiveness by exploiting treatment-impact heterogeneity. Second, it accounts system-atically for inequality and uncertainty, and the policymaker's attitude toward these, in the evaluation.",
keywords = "Bayesian econometric, Profiling, Program evaluation",
author = "Rajeev Dehejia",
year = "2005",
month = "3",
doi = "10.1016/j.jeconom.2004.04.006",
language = "English (US)",
volume = "125",
pages = "141--173",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier BV",
number = "1-2 SPEC. ISS.",

}

TY - JOUR

T1 - Program evaluation as a decision problem

AU - Dehejia, Rajeev

PY - 2005/3

Y1 - 2005/3

N2 - I argue for thinking of program evaluation as a decision problem. There are two steps. First, a counselor determines which program (treatment or control) each individual joins, based for example on maximizing the probability of employment or expected earnings. Second, the policymaker decides whether: to assign all individuals to treatment or to control, or to allow the counselor to choose. This framework has two advantages. Individualized assignment rules (known as profiling) can raise the average impact, improving cost effectiveness by exploiting treatment-impact heterogeneity. Second, it accounts system-atically for inequality and uncertainty, and the policymaker's attitude toward these, in the evaluation.

AB - I argue for thinking of program evaluation as a decision problem. There are two steps. First, a counselor determines which program (treatment or control) each individual joins, based for example on maximizing the probability of employment or expected earnings. Second, the policymaker decides whether: to assign all individuals to treatment or to control, or to allow the counselor to choose. This framework has two advantages. Individualized assignment rules (known as profiling) can raise the average impact, improving cost effectiveness by exploiting treatment-impact heterogeneity. Second, it accounts system-atically for inequality and uncertainty, and the policymaker's attitude toward these, in the evaluation.

KW - Bayesian econometric

KW - Profiling

KW - Program evaluation

UR - http://www.scopus.com/inward/record.url?scp=10044255625&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10044255625&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2004.04.006

DO - 10.1016/j.jeconom.2004.04.006

M3 - Article

AN - SCOPUS:10044255625

VL - 125

SP - 141

EP - 173

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 1-2 SPEC. ISS.

ER -