/barboni_comparison

Primary LanguageStata

Replication instructions for the Barboni, Field, and Pande (2021) comparison figure

Introduction

This document explains how to replicate the comparison figure used to compare the results on Barboni, Field, and Pande (2021) with similar papers for Sean Higgins's discussion at the NBER summer institute.

Data

The data sets used by this package were obtained from the puclicly replication codes of the papers used in the comparison figure. Microdata from the replication files from other studies is used to compare the effect size in the Barboni, et. al. (2021) study to those of other studies.

The data/ folder contains the following subfolders; the subsections below describe the source for the data that go into each subfolder.

  • angelucci15
  • augsburg15
  • banerjee15
  • crepon15

Specifically, the following replication data sets from the following studies are used:

In the mentioned subfolders inside the data/ folder, there are other dofiles that belong to replication packages. They are there to give a more complete version of the replication packages.

For other studies included in our comparison, we used the point estimates and standard errors from the paper; see scripts/barboni_comparison_dataprep.do for more detail.

The file data/barboni_comparison_metadata.xlsx includes metadata about the studies included in the comparison figure. It is included with the replication data.

Computational requirements

Most of the code was run on a computer with 8GB of RAM.

Software requirements

The following software programs and packages are required.

Stata 14.0

Earlier versions of Stata (13 or newer) should work as well but have not been tested. The Stata code was run using 4-core Stata-MP 14.0. All user-written .ado files required for replication are included in the scripts/programs/ folder, so they do not need to be separately installed to run the replication package. They include:

  • .ado files we've written
    • time
  • .ado files written by others
    • svmat2 (Nicholas J. Cox) and the .ado files needed for it to function.

R 3.6.3

The following list includes the R packages used and the version used; all of these packages and their dependencies are listed in the renv.lock file and can be installed using the renv package (installation instructions below).

  • tidyverse (1.3.0)
  • magrittr (1.5)
  • here (0.1)
  • metafor (3.0-2)
  • renv (0.12.0)

Folder structure

After downloading the replication code and data, unzip it into a folder on your computer. After unzipping, the project root directory containing the folders described below should be thought of as the main folder when editing the 00_run.do script to run on another machine. The folders inside of the zipped replication file should be placed in a folder that you will denote as global main in 00_run.do. These folders are:

  • data and its subfolders are for raw data.
  • figures is the folder to which the comparison figure produced by the scripts will be written.
  • logs is the folder in which log files will be written.
  • papers is the folder that contains all the papers used for comparison in a .pdf format.
  • proc is the folder in which processed data sets will be saved.
  • renv contains the information needed to install all needed R packages and dependencies.
  • scripts contains the replication codes and the files that produce the comparison figure.
  • programs subfolder in the scripts folder contains the .ado files required to run our Stata scripts.

Additional files

The project root directory contains the following files:

  • .here is included to enable R's here::here() function to work with relative file paths.
  • .Rprofile contains information to install all needed R packages and dependencies.
  • README.md is a markdown README file for the replication package.
  • README.txt is identical to README.md but included for those unsure how to open an .md file.
  • renv.lock contains information to install all needed R packages and dependencies.

Instructions

Set up

All the needed Stata packages (.ado files) are included in the programs/ folder and do not need to be installed. The list of needed R packages, including their versions and dependencies, are included in renv.lock, .Rprofile, and the renv/ folder, and can be installed by opening R from the project's root directory (or, equivalently, opening R and setting the working directory as the project's root directory), then running:

renv::restore()

Scripts

  • To create the comparison figure, first run scripts/barboni_comparison_dataprep.do.
  • Details on the scripts/barboni_comparison_dataprep.do file:
    • Change the directory in line 17 of the do file.
    • It contains locals. Thus, it is not recommended to be run in parts.
    • It also contains parts of the replication scripts from other papers, but has some modifications. The main one is that for some papers, it creates a new variable that measures total income. Do not modify those parts of the code.
  • Once the scripts/barboni_comparison_dataprep.do do file has been run, run the barboni_comparison_graph.R script.
  • The product of the R script is a graph saved both in a .eps file and a .jpg file.

References

Angelucci, Manuela, Dean Karlan, and Jonathan Zinman. 2015. "Microcredit Impacts: Evidence from a Randomized Microcredit Program Placement Experiment by Compartamos Banco." American Economic Journal: Applied Economics_, 7 (1): 151-82.

Augsburg, Britta, Ralph De Haas, Heike Harmgart, and Costas Meghir. 2015. "The Impacts of Microcredit: Evidence from Bosnia and Herzegovina." American Economic Journal: Applied Economics, 7 (1): 183-203.

Barboni, Giorgia, Erica Field, and Rohini Pande. 2021. Rural Banks Can Reduce Poverty: Experimental Evidence from 870 Indian Villages. Working paper.

Banerjee, Abhijit, Esther Duflo, Rachel Glennerster, and Cynthia Kinnan. 2015. "The Miracle of Microfinance? Evidence from a Randomized Evaluation." American Economic Journal: Applied Economics, 7 (1): 22-53.

Breza, Emily, and Cynthia Kinnan. 2021. "Measuring the Equilibrium Impacts of Credit: Evidence from the Indian Microfinance Crisis." The Quarterly Journal of Economics: Oxford University Press, vol. 136(3): pages 1447-1497.

Bruhn, M., and Love, I. 2014. "The Real Impact of Improved Access to Finance: Evidence from Mexico." Journal of Finance, 69: 1347-1376.

Callen, Michael, Suresh de Mel, Craig McIntosh, and Christopher Woodruff. 2019. "What Are the Headwaters of Formal Savings? Experimental Evidence from Sri Lanka." Review of Economic Studies: Oxford University Press, vol. 86(6): 2491-2529.

Crépon, Bruno, Florencia Devoto, Esther Duflo, and William Parienté. 2015. "Estimating the Impact of Microcredit on Those Who Take It Up: Evidence from a Randomized Experiment in Morocco." American Economic Journal: Applied Economics, 7 (1): 123-50.

Stein, Luke, Constantine Yannelis. 2020. "Financial Inclusion, Human Capital, and Wealth Accumulation: Evidence from the Freedman’s Savings Bank." The Review of Financial Studies, Volume 33, Issue 11: 5333–5377.