Skip to content

Replication repository for The Long-Run Development Impacts of Agricultural Productivity Gains: Evidence from Irrigation Canals in India

Notifications You must be signed in to change notification settings

devdatalab/paper-india-irrigation-canals

Repository files navigation

india-irrigation-canals-paper

This is the replication repository for "The Long-Run Development Impacts of Agricultural Productivity Gains: Evidence from Irrigation Canals in India" (Asher, Campion, Gollin, Novosad)

This repository contains the code required to replicate all tables and figures from the paper. Replication data can be found in this Google Drive folder: link.

Dataset list

  • canals_analysis_data.dta: main analysis dataset at the village and town level
  • canals_spillovers.dta: additional variables used for the spillovers analysis at the village and town level
  • canal_construction_data.dta: canal data on construction, at the canal segment level (matching to the GIS data)
  • pc81_canals_working.dta: National Sample Survey data on migration at the individual level, with district-level variation in canal coverage
  • pc51_balance_data.dta: subset of main data matched to 1951 villages with canals constructed after 1951
  • town_panel_equal_area.dta: town panel data

Replication Instructions

  1. Open canals_config.do
  2. Change the following filepaths saved as globals:
Filepath Description
ccode the base filepath for the repository
cdata the folder where the data files (from Google Drive link above) are saved
out the folder where you want all outputs saved
tmp a scratch folder where intermediate outputs can be saved
  1. Run canals_config.do in Stata. This will set all configurations, no need to change any other files.
  2. Open make_canals_analysis.do and run in Stata.

Computational requirements

Most code is run in Stata, with the relevant packages included in the ado folder in this repository.

The exception are the coefficient plots which are created using python. Only a very few basic packages are required: pandas, matplotlib, and numpy. Ensure that whatever python environment you are using is activated (i.e. Stata knows where to access the environment) before running the make_canals_analysis.do file in Stata, as it will call the code requiring python.

Or, if you do not want to run the python-dependent code, comment out the lines in the Figures section of the make_canals_analysis.do file before running it.

Figures and Tables

Main Text:

Exhibit Filename Created by
Table 1 sample_description.tex a/gen_sample_description.do
Table 2 rd_elev_balance.tex a/gen_balance_table.do
Table 3 rd_elev_results_irr_bal.tex a/gen_main_results.do
rd_elev_results_ag_bal.tex
rd_elev_results_ec_bal.tex
rd_elev_results_ed_bal.tex
Table 4 rd_land_ownership_results.tex a/gen_landownership.do
Table 5 spillovers_table_irr.tex a/spillovers_analysis.do
spillovers_table_ag.tex
spillovers_table_ec.tex
spillovers_table_ls.tex
Table 6 town_did_growth.tex a/analyze_town_diff_in_diff.do
town_did_appear.tex a/analyze_town_diff_in_diff.do
Figure 1 canal_completion_time.png a/plot_canal_completion_time.py
Figure 2 rel_elev_results_bal.pdf a/gen_rd_result_grid.do
Figure 3 all_elev_outcomes_coefplot_bal.png a/make_all_coefplots.py
Figure 4 land_ownership_coefplot.png a/land_ownership_coefplot.py
Figure 5 main_ln_pop_fill_20.pdf a/analyze_town_diff_in_diff.do
sfe_ln_pop_fill_20.pdf
main_appeared5000_20.pdf
sfe_appeared5000_20.pdf

Appendix:

Exhibit Filename Created by
Table A2 pc51_balance.tex a/gen_pc51_balance_table.do
Table A3 rd_additional_outcomes_bal.tex a/gen_main_results.do
Table A4 rd_comm_balance.tex a/gen_balance_table.do
Table A5 appendix_irr.tex a/gen_appendix_tables.do
Table A6 appendix_ag.tex a/gen_appendix_tables.do
Table A7 appendix_ec.tex a/gen_appendix_tables.do
Table A8 appendix_ed.tex a/gen_appendix_tables.do
Table A9 rd_comm_results_irr_bal.tex a/gen_main_results.do
rd_comm_results_ag_bal.tex
rd_comm_results_ec_bal.tex
rd_comm_results_ed_bal
Table A10 sinuosity_table.tex a/gen_appendix_tables.do
Table A11 sensitivity_table.tex a/make_sensitivity_table.do
Table A12 spillovers_table_irr_ebal_app.tex a/spillovers_analysis.do
Table A13 spillovers_table_ag_ebal_app.tex a/spillovers_analysis.do
Table A14 spillovers_table_ec_ebal_app.tex a/spillovers_analysis.do
Table A15 spillovers_table_ls_ebal_app.tex a/spillovers_analysis.do
Table A16 town_did_robust.tex a/analyze_town_diff_in_diff.do
Table A17 nss_migration_table.tex a/migration_nss.do
Figure A3 main_ln_pop_fill_10.pdf a/analyze_town_diff_in_diff.do
sfe_ln_pop_fill_10.pdf
main_appeared5000_10.pdf
sfe_appeared5000_10.pdf
main_ln_pop_fill_30.pdf
sfe_ln_pop_fill_30.pdf
main_appeared5000_30.pdf
sfe_appeared5000_30.pdf

About

Replication repository for The Long-Run Development Impacts of Agricultural Productivity Gains: Evidence from Irrigation Canals in India

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •