Mexican Electoral Database: Precinct-Level Analysis (1994-2019)

This repository provides a complete workflow for processing, cleaning, and analyzing precinct-level electoral data for 31 Mexican states (excluding Mexico City) between 1994 and 2019. The data processing pipeline is implemented in R, with each states data processed independently before merging into a consolidated dataset.

Folder Structure

Data

This folder contains all the raw and auxiliary data required to process and generate the intermediate datasets as well as the final processed databases.

Subfolders:

1. extraordinary_elections

   • correct_extra_elec_final.csv: Lists extraordinary and normal elections that are removed during processing.
     
   • diff_year_extra_elec.csv & diff_year_extra_elec_flag.csv: Files necessary for processing extraordinary elections held in different years.
     

2. collapsed_database_manual_cases

   - state_collapsed_edited.csv: Manually sourced information about incumbents whose new electoral coalitions differ from, or are not a superset of, their coalitions in previous elections.

3. Raw_Electoral_Data.csv

   • Contains the raw electoral data used to initiate state-level processing via the process_raw_electoral_data.R script.

4. incumbent_data

   • Contains data on incumbents, used during the construction of state-level databases.

5. municipal_magar_data_splitcoal

   • Contains Magar's Municipal and State Election Data [1], used for correlation analysis and to construct municipal-level incumbent and runner-up data in FINAL_db_formation.R.

Processed Data

This folder organizes the processed data for each of Mexico's 31 states, produced by sequential execution of the script files.

Subfolders and Key Files:

1. State Subfolders:

   • Each state's folder contains outputs from the scripts, saved sequentially:
     • state_process_raw_data.csv: Produced by process_raw_electoral_data.R.
     • state_vote_manipulation.csv: Produced by vote_manipulation.R.
     • state_incumbent_manipulator.csv: Produced by incumbent_manipulation.R.
     • state_vote_calculator.csv: Produced by incumbent_vote_calculator.R.
     • state_final.csv: Produced by final.R.

2. Shared Files:

   • duplicate_cases.csv: Used to address minor inconsistencies during the final merge.
   • coalition_dic.csv: Contains all coalitions present in the dataset.

Final Data

This folder contains the consolidated dataset merging all 31 states: - all_states_final.zip: The final dataset after processing all states, compressed for distribution.

Correlation Data

This folder contains data and scripts used for the correlation analysis during technical validation.

Subfolders:

1. turnout_magar

   • Contains Magar's Municipal and State Election Data [1], processed by magar_mun.R for correlation.

2. generated_data

   • Includes the outputs magar_turnout.csv, used in graphs_correlation.R.

Scripts

This folder contains all R scripts used for data processing and analysis.

State-Specific Scripts:

1. process_raw_electoral_data.R

   • Imports precinct-level electoral data, standardizes variables, and aggregates polling booth-level data to the precinct level. Produces state_process_raw_data.csv. The municipal electoral precinct-level data for 31 states of Mexico (excluding the Federal District) was originally sourced by Larreguy (2012) [3], Marshall (2023) [4], and Enrquez et al. (2024) [5].

2. vote_manipulation.R

   • Cleans and selects relevant variables (e.g., municipality/state names, codes, votes). For some states, incorporates correct_extra_elec_final.csv, diff_year_extra_elec.csv, and diff_year_extra_elec_flag.csv. Produces state_vote_manipulation.csv.

3. incumbent_manipulation.R

   • Processes incumbent and runner-up data from multiple sources, including Magar's Municipal and State Election Data [1] and SNIM [2], and merges it into the precinct-level dataset. Produces state_incumbent_manipulator.csv.

4. incumbent_vote_calculator.R

   • Calculates precinct-level votes for incumbents and runner-ups, standardizing party names and handling coalitions. Produces state_vote_calculator.csv.

5. final.R

   • Computes votes for major parties (PRI, PAN, MORENA, PRD) and state incumbents. Integrates manually researched data from state_collapsed_edited.csv. Produces state_final.csv.

6. FINAL_db_formation.R

   • Merges all 31 state datasets into a single dataset. Computes vote shares for incumbents, runner-ups, and major parties, relative to valid votes and registered voters. Outputs all_states_final.zip.

correlation

Contains scripts for technical validation: 1. magar_.R
- Generates magar_.csv using Magar data [1]. 2. graphs_correlations.R
- Produces correlation graphs and tests using magar_.csv.

README.md

This file provides detailed instructions for replicating the data processing workflow.

Execution Order

For each of the 31 states: 1. process_raw_electoral_data.R 2. vote_manipulation.R 3. incumbent_manipulation.R 4. incumbent_vote_calculator.R 5. final.R

After all states: 6. FINAL_db_formation.R

Variable Descriptions

1. state

   • Name of the state where the election took place.

2. mun

   • Name of the municipality where the election took place.

3. state_code

   • A numerical code assigned by INEGI that uniquely identifies each state. For the first 9 states (including Mexico City), this is a single digit (19), while for the remaining 23 states, it is a two-digit code (1032).

4. mun_code

   • A numerical code assigned by INEGI that uniquely identifies each municipality. For 4-digit codes, the first digit corresponds to states 1 through 9; for 5-digit codes, the first two digits correspond to states 10 through 32, with the remaining digits representing the specific municipality.

5. precinct

   • A 1- to 5-digit code used to uniquely identify electoral precincts within each state.

6. year

   • The year in which the election was held.

7. incumbent_party_candidate

   • The name of the candidate from the incumbent party or coalition who won the previous municipal election.

8. incumbent_party

   • The partisan composition of the coalition of the incumbent mayor who won the previous municipal election.

9. incumbent_party_component

   • The individual party or parties within the incumbent electoral coalition that contribute to the overall incumbent vote.

10. runnerup_party

    • The partisan composition of the coalition of the runner-up that came second in the previous municipal election.

11. state_incumbent_party

    • The political party or electoral coalition that held power at the state level during the municipal election.

12. mun_incumbent_party_vote

    • The number of votes received by the incumbent party or coalition in the previous municipal election at the municipal level.

13. mun_runnerup_party_vote

    • The number of votes received by the runner-up party or coalition in the previous municipal election at the municipal level.

14. mun_winning_margin

    • The winning margin of the incumbent party or electoral coalition, or the difference in its vote share and that of the runner-up party or electoral coalition, in previous municipal elections.

15. researched_incumbent_party

    • Manually researched incumbent party.

16. source_researched_incumbent_party

    • The provided source of the manually researched incumbent party.

17. incumbent_party_vote

    • The number of votes received in a given electoral precinct by the incumbent party or coalition in the municipal election.

18. share_incumbent_party_valid_vote

    • The percentage of votes cast in a given electoral precinct for the incumbent party or coalition as a proportion of valid votes.

19. share_incumbent_party_registered_voters

    • The percentage of votes cast in a given electoral precinct for the incumbent party or coalition as a proportion of registered voters.

20. runnerup_party_vote

    • The number of votes received in a given electoral precinct by the runner-up party or electoral coalition in the municipal election.

21. share_runnerup_party_valid_vote

    • The percentage of votes cast in a given electoral precinct for the runner-up party or coalition as a proportion of the valid votes.

22. share_runnerup_party_registered_voters

    • The percentage of votes cast in a given electoral precinct for the runner-up party or coalition as a proportion of registered voters.

23. state_incumbent_party_vote

    • The number of votes received in a given electoral precinct in the municipal election by the state incumbent party or the electoral coalition it ran under.

24. share_state_incumbent_party_valid_vote

    • The percentage of votes cast in a given electoral precinct in the municipal election for the state incumbent party or the electoral coalition it ran under as a proportion of valid votes.

25. share_state_incumbent_party_registered_voters

    • The percentage of votes cast in a given electoral precinct in the municipal election for the state incumbent party or the electoral coalition it ran under as a proportion of registered voters.

26. PRI_vote

    • The number of votes received in a given electoral precinct by the PRI or its electoral coalition.

27. share_PRI_valid_vote

    • The percentage of votes cast in a given electoral precinct for the PRI or its electoral coalition as a proportion of valid votes.

28. share_PRI_registered_voters

    • The percentage of votes cast in a given electoral precinct for the PRI or its electoral coalition as a proportion of registered voters.

29. PAN_vote

    • The number of votes received in a given electoral precinct by the PAN or its electoral coalition.

30. share_PAN_valid_vote

    • The percentage of votes cast in a given electoral precinct for the PAN or its electoral coalition as a proportion of valid votes.

31. share_PAN_registered_voters

    • The percentage of votes cast in a given electoral precinct for the PAN or its electoral coalition as a proportion of registered voters.

32. PRD_vote

    • The number of votes received in a given electoral precinct by the PRD or its electoral coalition.

33. share_PRD_valid_vote

    • The percentage of votes cast in a given electoral precinct for the PRD or its electoral coalition as a proportion of valid votes.

34. share_PRD_registered_voters

    • The percentage of votes cast in an electoral precinct for the PRD or its electoral coalition as a proportion of registered voters.

35. MORENA_vote

    • The number of votes received in a given electoral precinct by MORENA or its electoral coalition.

36. share_MORENA_valid_vote

    • The percentage of votes cast for MORENA or its electoral coalition as a proportion of valid votes.

37. share_MORENA_registered_voters

    • The percentage of votes cast in an electoral precinct for MORENA or its electoral coalition as a proportion of registered voters.

38. valid

    • The number of valid votes cast in a given electoral precinct, excluding invalid or spoiled ballots.

39. total

    • The sum of all votes cast in a given electoral precinct, including valid, invalid, and spoiled ballots.

40. registered_voters

    • The number of registered voters in a given electoral precinct.

41. turnout

    • The percentage of registered voters in a given electoral precinct that cast a vote in the election.

References

1. Magar, E. (2018). Recent Mexican Election Vote Returns. Retrieved from https://github.com/emagar/elecRetrns. Last revised on March 10, 2023.
2. SNIM (2024). Sistema Nacional de Informacin Municipal. Retrieved from http://www.snim.rami.gob.mx/.
3. Larreguy, H. A. (2012). Monitoring Political Brokers: Evidence from Clientelistic Networks in Mexico. Available at SSRN: https://ssrn.com/abstract=2225027.
4. Marshall, J. (2023). Political Information Cycles: When Do Voters Sanction Incumbent Parties for High Homicide Rates? Working Paper.
5. Enrquez, J. R., Larreguy, H., Marshall, J., & Simpser, A. (2024). Mass Political Information on Social Media: Facebook Ads, Electorate Saturation, and Electoral Accountability in Mexico. Journal of the European Economic Association.