Science Advances

Supplementary Materials

This PDF file includes:

  • section S1. A comprehensive literature review of the effects of electricity access
  • section S2. Summary statistics
  • section S3. Study setting
  • section S4. Site selection and external validity
  • section S5. Preanalysis plan
  • section S6. Additional estimates based on the preanalysis plan
  • section S7. Balance statistics
  • section S8. Additional descriptive data
  • section S9. Different LATE
  • section S10. Testing for geographic spillovers
  • section S11. Additional regression output for energy access effects
  • section S12. Multiple comparisons
  • section S13. Placebo tests from randomization inference
  • section S14. Additional regression output for socioeconomic effects
  • section S15. Socioeconomic effects: Full results
  • section S16. MGP
  • section S17. Robustness: Energy access effects without flooded villages
  • section S18. Robustness: Energy access effects without contaminated villages
  • section S19. Robustness: Energy access effects without suspicious case
  • section S20. Robustness: Energy access effects without treatment from wait list
  • section S21. Robustness: Energy access effects without households with 24 hours of electricity
  • table S1. Summary statistics for outcome and other key variables across all three survey waves.
  • table S2. Summary statistics for outcome and other key variables, separate for treatment, control, and remote control group and by wave.
  • table S3. Mean values of different variables at the village level across different samples, all on a scale of 0 to 1 of population shares.
  • table S4. Effect of MGP solar microgrids on gender schooling equality.
  • table S5. Effect of MGP solar microgrids on gender job equality.
  • table S6. Balance statistics at habitation level (pretreatment).
  • table S7. Balance statistics at household level (pretreatment).
  • table S8. Primary lighting sources by survey wave.
  • table S9. Primary lighting sources by survey wave and treatment status.
  • table S10. Number of households by subgroup and wave.
  • table S11. Reasons for discontinuation of MGP services.
  • table S12. Effect of MGP solar microgrids with a different LATE (see text) on household spending on kerosene in the private market and on kerosene through the PDS.
  • table S13. Effect of MGP solar microgrids with a different LATE (see text) on household electrification and hours of electricity per day.
  • table S14. Effect of MGP solar microgrids with a different LATE (see text) on household electrification and hours of electricity per day.
  • table S15. Effect of MGP solar microgrids on household kerosene spending including remote control habitations.
  • table S16. Effect of MGP solar microgrids on household electrification and hours of electricity including remote control habitations.
  • table S17. Effect of MGP solar microgrids on household spending on kerosene by wave, separate for private, public, and total kerosene expenditures.
  • table S18. Effect of MGP solar microgrids on household lighting satisfaction and hours of lighting.
  • table S19. Effect of MGP solar microgrids on the use of kerosene as the main source of lighting (=1 if the household uses kerosene for lighting).
  • table S20. Benjamini and Hochberg (27) corrections of P values for the energy access family of outcomes.
  • table S21. Effect of MGP solar microgrids on socioeconomic outcomes by wave.
  • table S22. Effect of MGP solar microgrids on socioeconomic outcomes by wave.
  • table S23. Socioeconomic effects of MGP solar microgrids on household savings in rupees per month.
  • table S24. Socioeconomic effects of MGP solar microgrids on household expenditures in rupees per month.
  • table S25. Socioeconomic effects of MGP solar microgrids on household business ownership, measured as a dichotomous indicator that takes a value of 1 if the household head owns a business.
  • table S26. Socioeconomic effects of MGP solar microgrids on the amount of work hours per day (female module).
  • table S27. Socioeconomic effects of MGP solar microgrids on household use of lighting for studying, measured as a dichotomous indicator that takes a value of 1 if the respondent or the children use lighting to study at night.
  • table S28. Socioeconomic effects of MGP solar microgrids on household expenditures to charge mobile phones, measured in rupees per week.
  • table S29. Socioeconomic effects of MGP solar microgrids on household expenditures to charge mobile phones, measured in rupees per week, controlling for electrification status of the household.
  • table S30. Effect of MGP solar microgrids on prevalence of domestic violence against women in the habitation (female module).
  • table S31. Effect of MGP solar microgrids on prevalence of eve teasing of women in the habitation (female module).
  • table S32. Effect of MGP solar microgrids on perceived safety in habitation because of better lighting (female module).
  • table S33. Effect of MGP solar microgrids on belief there is enough light to go outside in habitation (female module).
  • table S34. Effect of MGP solar microgrids on women’s time spent cooking per day (female module).
  • table S35. Effect of MGP solar microgrids on household kerosene spending, without flooded villages.
  • table S36. Effect of MGP solar microgrids on household electrification and hours of electricity, without flooded villages.
  • table S37. Effect of MGP solar microgrids on household kerosene spending, without contaminated villages.
  • table S38. Effect of MGP solar microgrids on household electrification and hours of electricity, without contaminated villages.
  • table S39. Effect of MGP solar microgrids on household kerosene spending, without suspicious household.
  • table S40. Effect of MGP solar microgrids on household electrification and hours of electricity, without suspicious household.
  • table S41. Effect of MGP solar microgrids on household kerosene spending, without treatment habitations from wait list.
  • table S42. Effect of MGP solar microgrids on household electrification and hours of electricity, without treatment habitations from the wait list.
  • table S43. Effect of MGP solar microgrids on household kerosene spending, without households with 24 hours of electricity per day.
  • table S44. Effect of MGP solar microgrids on household electrification and hours of electricity, without households with 24 hours of electricity per day.
  • fig. S1. Locations of study habitations in the Barabanki district.
  • fig. S2. Spending on kerosene on the private (black) market in the pretreatment period (baseline survey).
  • fig. S3. Spending on kerosene through the PDS in the pretreatment period (baseline survey).
  • fig. S4. Hours of electricity per day in the pretreatment period (baseline survey).
  • fig. S5. Placebo estimates for electricity access, private kerosene expenses, and total kerosene expenses.
  • References (29–38)

Download PDF

Files in this Data Supplement: