Data on crude monthly municipal-level homicide rates is available in mexico-muni-month-homicide-rates-2000-2022.csv
; state-level aggregations are available in mexico-state-month-homicide-rates-2000-2022.csv
. Note that both files use |
as the separator.
If you use R
you can use readr
package to load the file and specify the separator with readr::read_delim("PATH_TO_FILE", delim = "|")
The code to reproduce the homicide rate calculations is in the code
subdirectory and divided into three groups of tasks: census-data
, deaths-data
, and homicide-rates
. The census-data
and deaths-data
tasks assume that the requisite data is in a top-level directory called data
.
To replicate the results, first run the import
sub-task within the census-data
task using the Makefile
in the census-data/import
directory. This task reads in data from the 2000, 2010, and 2020 censuses, extracts municipal-level population counts, and verifies that they sum to the total population. This task expects three census data files from INEGI in a sub-directory called census
within the top-level data
directory. The files are:
- cgpv2000_iter_00.csv, retreived from https://www.inegi.org.mx/programas/ccpv/2000/#Datos_abiertos (download data for "Estados Unidos Mexicanos")
- iter_00_cpv2010.csv, retreived from https://www.inegi.org.mx/programas/ccpv/2010/#Datos_abiertos (download data for "Estados Unidos Mexicanos")
- conjunto_de_datos_iter_00CSV20.csv, retreived from https://www.inegi.org.mx/programas/ccpv/2020/#Datos_abiertos (download data for "Estados Unidos Mexicanos")
Next run the interpolate
sub-task within the census-task
using the Makefile
in the census-data/interpolate
directory. This task uses the population counts from the import
sub-task to linearly interpolate mid-year (1 July) population counts for each municipality from 2000-2022.
After running both sub-tasks in the census-data
task, run the sub-tasks in the deaths-data
directory. Again, this task begins with an import
sub-task, which you can run using the Makefile
. This task reads in death certificate files published in .dbf
format by INEGI and writes their contents to .csv
files. This task expects death certificate files from 2000-2022 in a sub-directory called death-certificates
within the top-level data
directory. These files can be downloaded from https://www.inegi.org.mx/programas/mortalidad/#Microdatos.
Next, run the homicide-counts
sub-task using the Makefile
. This task uses the death certificate files imported in the deaths-data/import
task to generate counts of homicide deaths in each municipality in each month from January 2000-December 2022. The cause of death classification file, found in the hand
subdirectory follows the cause of death classification scheme used by Elo, Beltrán-Sánchez and Macinko (2014). Note that deaths that occurred outside of Mexico and deaths that were missing cause of death, country of occurrence, or municipality of occurrence were excluded from these calculations. We also opted to use data from the location where the death occurred rather than the location where the individual was from because this information was more complete for homicides. One day we might impute this information and recalculate the counts accordingly.
Finally, run the top-level homicide-rates
task to calculate the monthly municipal-level crude homicide rates for January 2000-December 2022.
If you use this data please use the BibTeX entry below or see the OSF repository for other citation formats:
@misc{Gargiulo_Aburto_Floridi_2023,
title={Monthly municipal-level homicide rates in Mexico (January 2000–December 2022)},
url={osf.io/u8dc3},
DOI={10.17605/OSF.IO/U8DC3},
publisher={OSF},
author={Gargiulo, Maria and Aburto, José Manuel and Floridi, Ginevra},
year={2024},
month={March}
}