Age in Peru 2017 census

[MarkRMont]

In the R and .csv files for Peru 2017, I am not able to locate variables (or value labels) for the household member's age in the personas files.

Sex is described in persona_labels_c5p2, and for infants whose age is coded only in months, the months are labelled in persona_labels_c5p41m. For other household members, I would have expected to see age (either in completed years, or in birth-date form) in variables named something akin to c5p3 or c5p4.

I hope I am not missing something obvious! Otherwise these files are beautifully organized.

Mark R. Montgomery

[pachadotdev]

Hi Mark,

Thanks for the positive comment. I'm glad this is useful.

I went to the census page https://censos2017.inei.gob.pe/redatam/, but it's offline

I did this

censo <- readRDS("PER2017.rds")

colnames(censo$persona)

 [1] "persona_ref_id" "hogar_ref_id"   "parea"          "pencare"       
 [5] "c5p2"           "c5p41m"         "c5p5"           "c5p5bcod"      
 [9] "c5p6"           "c5p6bcod"       "c5p7"           "c5p7bcod"      
[13] "c5p81"          "c5p83"          "c5p84"          "c5p85"         
[17] "c5p86"          "c5p91"          "c5p92"          "c5p93"         
[21] "c5p95"          "c5p96"          "c5p97"          "c5p11"         
[25] "c5p12"          "c5p13niv"       "c5p13gra"       "c5p13anios"    
[29] "c5p13anioe"     "c5p14"          "c5p15"          "c5p15bcod"     
[33] "c5p16"          "c5p17"          "c5p18"          "c5p19coda"     
[37] "c5p20coda"      "c5p19cod"       "c5p20cod"       "c5p21"         
[41] "c5p23"          "c5p23bcod"      "c5p24"          "c5p25"         
[45] "c5p26"          "c5p27"          "c5p271"         "c5p28"         
[49] "c5p281"         "c5p291"         "c5p292"         "factorpond"    
[53] "edquinq"        "p03esc"         "p03gq15m"       "pet"           
[57] "p09disc"        "p08afilia"      "c5p5di"         "c5p6di"        
[61] "c5p7di"         "c5p15di"        "c5p5pv"         "c5p6pv"        
[65] "c5p7pv"         "c5p15pv"       

censo$persona_labels_c5p41m

# A tibble: 14 × 2
   c5p41m c5p41m_description 
    <int> <fct>              
 1      0 Edad menos de 1 mes
 2      1 Edad de 1 mes      
 3      2 Edad de 2 meses    
 4      3 Edad de 3 meses    
 5      4 Edad de 4 meses    
 6      5 Edad de 5 meses    
 7      6 Edad de 6 meses    
 8      7 Edad de 7 meses    
 9      8 Edad de 8 meses    
10      9 Edad de 9 meses    
11     10 Edad de 10 meses   
12     11 Edad de 11 meses   
13     13 MISSING            
14     12 NOTAPPLICABLE

quantile(censo$persona$c5p41)
  0%  25%  50%  75% 100% 
  NA   NA   NA   NA   NA 
Warning message:
Unknown or uninitialised column: `c5p41`. 

this means the microdata downloaded from https://redatam.org/cdr/descargas/censos/poblacion/CP2017PER.zip does not have the age column :\

you would need to contact the Peruvian authorities here: https://www.gob.pe/20399-solicitar-acceso-a-la-informacion-publica?child=25841

I know the variables because I have this Excel file where I looked for "edad"

censo-2017-peru-variables.xlsx

[pachadotdev]

@MarkRMont also relevant, if you use our datasets or software, please cite the open access article

Vargas Sepúlveda, Mauricio and Barkai, Lital. 2025. "The REDATAM format and its challenges for data access and information creation in public policy." Data & Policy 7 (January): e18. https://dx.doi.org/10.1017/dap.2025.4.

[MarkRMont]

Thanks for responding so quickly! This does seem to be an odd situation. I'll confess that I cannot quite see how my own individual request to the Peruvian authorities would induce them to reissue the 2017 census microdata, which would then enter the Redatam data-processing pipeline. For now I will stay with the IPUMS-I microdata sample for Peru 2017, which has the age variable as well as all other variables that I need---although obviously the full set of records would be far preferable to the IPUMS sample. I have been citing your 2025 article at every opportunity---it expresses very well the motivations many of us have for using Open Redatam. In my own work, I chain together the census identifiers in precisely the way you describe in the article, although I prefer doing it from the top (level-1) down to personas. The article will have a truly beneficial cumulative effect, I predict. Best, Mark

…

On Mon, Apr 14, 2025 at 7:22 PM pachadotdev ***@***.***> wrote: @MarkRMont <https://github.com/MarkRMont> also relevant, if you use our datasets or software, please cite the open access article Vargas Sepúlveda, Mauricio and Barkai, Lital. 2025. "The REDATAM format and its challenges for data access and information creation in public policy." *Data & Policy* 7 (January): e18. https://dx.doi.org/10.1017/dap.2025.4. — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADKWCBCX6JLC7KQR3LSJF3D2ZQ7K7AVCNFSM6AAAAAB3DWWK6SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQMBTGI4TAMZTGI> . You are receiving this because you were mentioned.Message ID: ***@***.***> *pachadotdev* left a comment (pachadotdev/redatam-microdata#1) <#1 (comment)> @MarkRMont <https://github.com/MarkRMont> also relevant, if you use our datasets or software, please cite the open access article Vargas Sepúlveda, Mauricio and Barkai, Lital. 2025. "The REDATAM format and its challenges for data access and information creation in public policy." *Data & Policy* 7 (January): e18. https://dx.doi.org/10.1017/dap.2025.4. — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADKWCBCX6JLC7KQR3LSJF3D2ZQ7K7AVCNFSM6AAAAAB3DWWK6SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQMBTGI4TAMZTGI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

-- Mark R. Montgomery Professor of Economics Stony Brook University Stony Brook, NY 11794 ***@***.***

[pachadotdev]

Thanks for responding so quickly!

This does seem to be an odd situation. I'll confess that I cannot quite see
how my own individual request to the Peruvian authorities would induce them
to reissue the 2017 census microdata, which would then enter the Redatam
data-processing pipeline.

For now I will stay with the IPUMS-I microdata sample for Peru 2017, which
has the age variable as well as all other variables that I need---although
obviously the full set of records would be far preferable to the IPUMS
sample.

I have been citing your 2025 article at every opportunity---it expresses
very well the motivations many of us have for using Open Redatam. In my own
work, I chain together the census identifiers in precisely the way you
describe in the article, although I prefer doing it from the top (level-1)
down to personas. The article will have a truly beneficial cumulative
effect, I predict.

Best, Mark
…

welcome, probably they can send you a CSV file with only that column

if Spanish is an issue, I can contact them, plus I taught them at P. U. Catolica del Peru :)