DOC PREVIEW
Using Randomization in Development

This preview shows page 1-2-3-4-5-6-41-42-43-44-45-46-84-85-86-87-88-89 out of 89 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 89 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Using Randomization in Development Economics Research: A Toolkit Esther Duflo, Rachel Glennerster, and Michael Kremer CID Working Paper No. 138 December 2006 © Copyright 2006 Esther Duflo, Rachel Glennerster, Michael Kremer, and the President and Fellows of Harvard College at Harvard UniversityCenter for International DevelopmentWorking PapersUsing Randomization in Development Economics Research: AToolkit∗Esther Duflo†, Rachel Glennerster‡and Michael Kremer§December 12, 2006AbstractThis paper is a practical guide (a toolkit) for researchers, students and practitioners wish-ing to introduce randomization as part of a research design in the field. It first covers therationale for the use of randomization, as a solution to selection bias and a partial solutionto publication biases. Second, it discusses various ways in which randomization can be prac-tically introduced in a field settings. Third, it discusses designs issues such as sample sizerequirements, stratification, level of randomization and data collection methods. Fourth, itdiscusses how to analyze data from randomized evaluations when there are departures fromthe basic framework. It reviews in particular how to handle imperfect compliance and ex-ternalities. Finally, it discusses some of the issues involved in drawing general conclusionsfrom randomized evaluations, including the necessary use of theory as a guide when design-ing evaluations and interpreting results. JEL Classification: I0; J0; O0; C93. Keywords:Randomized evaluations; Experiments; Development; Program evaluation.∗We thank the editor T.Paul Schultz, as well Abhijit Banerjee, Guido Imbens and Jeffrey Kling for extensivediscussions, David Clingingsmith, Greg Fischer, Trang Nguyen and Heidi Williams for outstanding researchassistance, and Paul Glewwe and Emmanuel Saez, whose previous collaboration with us inspired parts of thischapter.†Department of Economics, Massachusetts Institute of Technology and Abdul Latif Jameel Poverty ActionLab‡Ab dul Latif Jameel Poverty Action Lab§Department of Economics, Harvard University and Abdul Latif Jameel Poverty Action Lab1Contents1 Introduction 32 Why Randomize? 42.1 The Problem of Causal Inference . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Randomization Solves the Selection Bias . . . . . . . . . . . . . . . . . . . . . . . 72.3 Other Methods to Control for Selection Bias . . . . . . . . . . . . . . . . . . . . . 102.3.1 Controlling for Selection Bias by Controlling for Observables . . . . . . . 102.3.2 Regression Discontinuity Design Estimates . . . . . . . . . . . . . . . . . 112.3.3 Difference-in-Differences and Fixed Effects . . . . . . . . . . . . . . . . . . 122.4 Comparing Experimental and Non-Experimental Estimates . . . . . . . . . . . . 132.5 Publication Bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.5.1 Publication bias in non-experimental studies . . . . . . . . . . . . . . . . 152.5.2 Randomization and publication bias . . . . . . . . . . . . . . . . . . . . . 173 Incorporating Randomized Evaluation in a Research Design 193.1 Partners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203.2 Pilot projects: From program evaluations to field experiments . . . . . . . . . . . 223.3 Alternative Methods of Randomization . . . . . . . . . . . . . . . . . . . . . . . . 243.3.1 Oversubscription Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.3.2 Randomized Order of Phase-In . . . . . . . . . . . . . . . . . . . . . . . . 253.3.3 Within-Group Randomization . . . . . . . . . . . . . . . . . . . . . . . . . 263.3.4 Encouragement Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274 Sample size, design, and the power of exp eriments 284.1 Basic Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284.2 Grouped Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314.3 Imperfect Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334.4 Control Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344.5 Stratification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.6 Power calculations in practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3815 Practical Design and Implementation Issues 405.1 Level of Randomization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405.2 Cross-Cutting Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425.3 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455.3.1 Conducting Baseline Surveys . . . . . . . . . . . . . . . . . . . . . . . . . 455.3.2 Using Administrative Data . . . . . . . . . . . . . . . . . . . . . . . . . . 466 Analysis with Departures from Perfect Randomization 476.1 The Probability of Selection Depends on the Strata . . . . . . . . . . . …


Using Randomization in Development

Download Using Randomization in Development
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Using Randomization in Development and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Using Randomization in Development 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?