Identification of entire counterfactual distribution of potential outcomes: Evidence from NSW Demonstration job training program Open Access

Abstract

When evaluating the effect of any social program using non-experimental data, if a more comprehensive econometrics model is used, then a more precise estimation of the program's (treatment) effect can be made. The heterogeneity of treatment effect and the unobserved characteristics of individuals in economic evaluations of job training programs are addressed in this study. To evaluate the effect of attending a job training program (National Supported Work Demonstration, NSW) on real earnings, this study non-parametrically estimates the entire counterfactual distribution of potential outcomes. The results show that for most quantiles of the earnings distribution, the effect of the NSW program is positive, and individuals in the lower levels of earnings benefit more from the program. Additionally, the estimated average treatment effect of attending this program drops significantly when in addition to the observable characteristics of individuals; their unobservable characteristics are controlled in the estimation.

Table of Contents

Table of Contents

1 Introduction.............................................................................................................................................................................1

2 Model Structure and Identification ...............................................................................................................................................4

2.1 Identification in the Simple Case (No Covariates and Equal Returns to Unobservable).................................................................. 6

2.2 Identification with Allowing for Covariates........................................................................................................................... 8

2.3 Identification with Allowing for Different Returns to Unobservable............................................................................................ 9

3 Data .................................................................................................................................................................................... 10 4 Estimation and Results ............................................................................................................................................................. 12 5 Conclusion ............................................................................................................................................................................. 20 References................................................................................................................................................................................ 22

List of Tables

1 Descriptive Statistics (Individual Observables).............................................................................................................................11

2 ATT (Average Treatment Effects) Estimates of the NSW
Job Training Attendance on Real Earnings, 1978 .............................................................................................................................13

List of Figures

1 (DID) Counterfactual and realized distribution of real earnings, specification1.................................................................................... 14

2 CDF of counterfactual and realized real earnings, specification1...................................................................................................... 15

3 Quantile treatment effects, specification1................................................................................................................................... 15

4 (DID) Counterfactual and realized distribution of real earnings, specification2..................................................................................... 16

5 CDF of counterfactual and realized real earnings, specification2 ..................................................................... ................................ 16

6 Quantile treatment effects, specification2 ....................................................................................................................................17

7 (DID) Counterfactual and realized distribution of real earnings, specification3 ..................................................................................... 17

8 CDF of counterfactual and realized real earnings, specification3 ........................................................................................................18

9 Quantile treatment effects, specification3 ....................................................................................................................................18

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Economics
Degree	MA
Submission	Master's Thesis
Language	English
Research Field	Economics, Labor
Keyword	Mincer equation average and quintile treatment effects counterfactual distribution non-parametric estimation
Committee Chair / Thesis Advisor	Maasoumi, Esfandiar, Emory University
Committee Members	Markowitz, Sara, Emory University Mialon, Sue, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Identification of entire counterfactual distribution of potential outcomes: Evidence from NSW Demonstration job training program ()	2018-08-28 14:39:48 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions