The "bunching" command uses bunching to partially and point identify the elasticity of income with respect to tax rates using different assumptions on unobserved heterogeneity, according to the paper by Bertanha, McCallum, and Seegert. The command "bunching" is a master function that combines three other commands included in the package: bunchbounds for partial identification, bunchtobit for point identification, and bunchfilter to filter out friction errors.
1. type "ssc describe bunching" in STATA for a description of the package
2. type "ssc install bunching" for automatic installation
3. type "net get bunching" to download a test data file to your current working directory
4. type "help bunching" for help and examples of how to use the command
1. download and unzip the file
"stata_bunching.zip"
2. copy the contents of the unzipped folder to your personal-ado directory. To find out the location of such directory, type "sysdir" in Stata and look for "personal":
MAC users: typically ~/Documents/Stata/ado or ~/Library/Application Support/Stata/ado depending on the version of Stata
Windows users: typically C:\Program Files\Stata16\ado or C:\ado\personal depending on the version of Stata
3. after installing the files, type "help bunching" for help
This package is in beta testing. This package is not yet a complete package (ado file program) and is currently being tested with real data. Feel free to test the code or build off of it for your own work. As a courtesy, please let me know if you find any errors.
1. download and unzip the file
"stata_crosscountry.zip"
2. This folder contains the following files.
a) 0_simulateData.do This file lets you simulate data to be used with the other files.
b) 1_fixedCost.do This file implements the fixed cost method described in the paper.
c) 2_iterativeOLS.do This file implements the two step method described in the paper.
d) 3_truncatedTobit.do This file implements the two step method described in the paper.
e) zerokink_loglike.ado This file is used in the truncatedTobit.do file.