CoChat
Sign Up
Cited by 0Labor market dynamics and wage inequalityLabour Market and MigrationEmployment, Labor, and Gender Studies

Brecha salarial de género en el Estado peruano: un análisis desde la ciencia de datos

Open access
León Leyva, Roberto·2024·Revista I+i
DOI 10.71701/revistaii.v.18.2024.91
Open this paper in CoChat — free accountOpen PDF

Abstract

The objective of this study is to analyze the existence of the gender wage gap in the Peruvian government and identify the profiles of these gaps at the regional level, based on open data provided by the National Authority of the Civil Service (Servir), applying data science methodologies and techniques. The research was applied and quantitative in nature, explanatory in level, and used a non-experimental longitudinal design (2017-2021). The methodology used was CRISP-DM (Cross Industry Standard). During the Exploratory Data Analysis it was found that at the national level the female participation is closed to parity getting an average of 47 % on the other hand, the gender pay gap in the Government had a downward trend from 13 % (2017) to 11 % (2021), an indicator better than the general average for Latin America which was 14 % (2019 ILO), additionally, the correlation showed that to more participation correspond less pay gap, however, when the information is disaggregated by regions, heterogeneity of the gaps is evident (interquartile range for Apurimac of 25 % and Piura of 2 %) and its evolution; in many regions the pandemic generates that the improvements in the gap are reversed or slowed down. To determine which regions made progress, a simple linear regression model was used based on the year; the negative slope would indicate progress, and the positive slope would indicate reversals in the period under study; the regions that made the best progress were Moquegua, Huánuco and Ancash, while those that showed deterioration were Huancavelica, La Libertad and San Martin. To identify the gap profiles, the year 2021 was used and variables obtained from the INEI were added, such as GDP per capita and coverage per public servant, the first factor associated with the progress of the region and the second associated with government services, therefore two clusters were determined with median values of pay gap high (12,7 %) and low (8,9 %). The Brown-Mood nonparametric hypothesis test implies that the pay gap exists regardless of the labor laws that banning the discrimination of rewarding in the labor market.

AI snapshot

Written from the paper’s full text.

TL;DR

Using Servir open data and CRISP-DM, the paper finds a gender pay gap in Peru’s public sector that looks small at the national aggregate but is heterogeneous across regions. Regional regressions and clustering identify improving/worsening regions and two profiles with median gaps of 12.7% and 8.9%, with a Brown-Mood test confirming different medians.

Key findings

  • At the national public-sector level, the gender pay gap fell from 13.0% in 2017 to 10.10% in 2020, then rose to 11.16% in 2021; female participation moved from 46.9% (2017) to 48.3% (2020) and 47.3% (2021), compared with a Latin America average gap of 14% in 2019 (OIT).
  • Regional medians varied widely: Pasco had the highest median gap (26%) and Ica the lowest (6.3%); Apurímac had the largest interquartile range (25%) and Piura the smallest (2%). Participation was highest in Pasco (median 56%) and lowest in Madre de Dios (39%).
  • Simple linear regressions across 25 regions, using a 0.10 significance level, found improvements in Huánuco, Moquegua, and Áncash; deteriorations in La Libertad, San Martín, and Huancavelica; and no significant evidence for 19 regions.
  • For 2021, clustering produced two profiles: cluster 1 had a median gap of 12.67%, participation 44.45%, GDP per capita 8.62 thousand soles, and 21.41 persons per public servant; cluster 2 had a median gap of 8.93%, participation 49.88%, GDP per capita 19.46 thousand soles, and 20.98 persons per public servant. Brown-Mood test: Z = 2.1669, p = 0.03025. Women also had higher completed university/advanced education than men (55% vs. 39%).

Methods

  • Applied quantitative, explanatory, non-experimental longitudinal design for 2017–2021, using CRISP-DM. Data came from Servir open data aggregated by region, gender, number of people, and average income, plus INEI population and regional GDP for 2021; missing 2018 public-servant counts were imputed as the semisum of 2017 and 2019.
  • Analysis included exploratory data analysis, simple linear regression per region with year as the independent variable and pay gap as dependent, clustering after standardization using hierarchical, k-means, and PAM algorithms with elbow/internal/stability validation, and Brown-Mood nonparametric median tests.
  • Regional profiles were characterized with descriptive statistics, using medians because of variable asymmetry.

Limitations

  • The paper states that the CRISP-DM deployment phase was not developed because this is a research study; it also states that more information and granularity are needed (job categories, education, experience, age; provinces, districts, sectors, public organisms) and that 2018 servant data were missing and imputed. It justifies a 0.10 significance level by noting sampling, instrument, and theoretical-model limitations in complex social problems.
  • Evident from the text: the analysis relies on aggregated regional/year data for 25 regions over five years, so it cannot control for individual-level characteristics; the conclusions mention a 5% significance level while the methods and table use 0.10, and some regions described as significant have p-values above 0.05.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.71701/revistaii.v.18.2024.91

Cites

0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 11 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Abstract

    La brecha salarial de género en el Estado peruano tenía una tendencia a la baja del 13 % en 2017 al 11 % en 2021.

    “en tanto que la brecha salarial de género en el Estado tenía una tendencia a la baja del 13 % (2017) a 11 % (2021)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.
  • Findingown data
    § Abstract

    Las regiones que presentaron mejores avances en la brecha salarial de género fueron Moquegua, Huánuco y Áncash.

    “Las regiones que presentaron mejores avances fueron Moquegua, Huánuco y Áncash”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.
  • Findingown data
    § Abstract

    A nivel nacional, la participación de las mujeres en el Estado peruano está cerca de la paridad, con un promedio del 47 %.

    “Durante el análisis exploratorio de datos, se encontró que a nivel nacional la participación de las mujeres está cerca de la paridad con un promedio del 47 %”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.
  • Findingown data
    § Abstract

    Las regiones que presentaron deterioro en la brecha salarial de género fueron Huancavelica, La Libertad y San Martín.

    “en tanto que las que presentaron deterioro fueron Huancavelica, La Libertad y San Martín”
    Quote in the paperSubstantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.
  • Findingcitation
    § Abstract

    El indicador de brecha salarial del Estado peruano en 2021 era mejor que el promedio general de Latinoamérica de 14 % en 2019 según la OIT.

    “indicador mejor al promedio general de Latinoamérica que era de 14 % (2019 OIT)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites OIT

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.
  • May combine two claims. 24
  • Classified as probably not a substantial claim. 2

  • May combine two claims.

    Findingown data
    § Abstract

    Al desagregar la información por regiones, se evidencia heterogeneidad en la participación y en la brecha salarial, con rango intercuartílico de 25 % para Apurímac y 2 % para Piura.

    “cuando se desagrega la información por regiones se evidencia heterogeneidad en participación y en la brecha salarial (rango intercuartílico para Apurímac de 25 % y Piura de 2 %)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.

  • May combine two claims.

    Findingown data
    § Abstract

    Existe brecha salarial de género en el Estado peruano, a pesar de contar con un marco legal que prohíbe la discriminación salarial.

    “La prueba de hipótesis no paramétrica de Brown-Mood permite concluir que existe la brecha salarial de género en el Estado peruano a pesar de contar con un marco legal que prohíbe la discriminación salarial.”
    • Also stated in: § INTRODUCCIÓN · plain
      “sin embargo, al analizar por regiones y observar la evolución durante el periodo estudiado, se encontró evidencia de que la brecha persiste.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    Scatter plot of salary gap vs participation for all regions and years shows scarce positive relationship, correlation 0.149.

    “se observa una escasa relación positiva, con una correlación de 0,149, como se muestra en la figura 10.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § CONCLUSIONES

    Cluster 2 comprises 11 regions, but concentrates those with greater number of servants (56%).

    “El clúster 2 comprende 11 regiones, pero concentra las que tienen mayor número de servidores (56 %).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § C. Fase III, preparación de datos

    Cluster 2 has salary gap of 8.9%, participation almost parity, and GDP per capita more than double cluster 1.

    “En el clúster 2, los indicadores son más favorables: la brecha salarial se reduce al 8,9 %, la participación es casi paritaria y el PBI per cápita es más del doble que en el clúster 1.”
    • Also stated in: § CONCLUSIONES · plain
      “Además, el clúster 2 presenta una participación casi paritaria (49,9 %) y un PBI per cápita más del doble que el clúster 1 (19 400 vs. 8600).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    Segmenting by terciles, no significant relationship in low and medium levels, correlations -0.089 and -0.015.

    “no se aprecia una relación significativa en los niveles bajo y medio, con correlaciones muy bajas de -0,089 y -0,015, respectivamente.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § D. Fase IV, modelado y Fase V, evaluación

    Highest median participation Pasco 56%, lowest Madre de Dios 39%.

    “La mayor participación mediana corresponde a Pasco, con un 56 %, mientras que la menor se registra en Madre de Dios, con un 39 %.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § C. Fase III, preparación de datos

    Simple linear regression identified six regions with significant coefficients: three with advances (Huánuco, Moquegua, Áncash) and three with regressions (La Libertad, San Martín, Huancavelica).

    “Esta tabla muestra modelos significativos para seis regiones, en las que se encontraron coeficientes significativos.Tres regiones lograron avances en la reducción de la brecha: Huánuco, Moquegua y Áncash, mientras que tres retrocedieron: La Libertad, San Martín y Huancavelica.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Additional information

    At state level, indicators show better levels than general averages of regions, with tendency to improve that was reversed by pandemic in 2020.

    “Del análisis exploratorio de datos se observó que, a nivel estatal, los indicadores muestran mejores niveles que los promedios generales de las regiones, con una tendencia a mejorar que fue revertida por la pandemia en 2020.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § C. Fase III, preparación de datos

    In two clusters, median salary gap values are 12.7 for cluster 1 and 8.9 for cluster 2, clearly distinct.

    “un diagrama de violín que muestra la mediana de ambas caracterizaciones evidencia que las distribuciones y los valores medianos (12,7 para el clúster 1 y 8,9 para el clúster 2) son claramente distintos”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    Salary gap reduced from 13.0% in 2017 to 10.10% in 2020, then increased to 11.16% in 2021.

    “La brecha salarial se redujo del 13,0 % en 2017 al 10,10 % en 2020, para luego aumentar al 11,16 % en 2021.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    Participation has reached levels close to parity, and salary gap has been decreasing slowly but progressively until before the pandemic; then both indicators deteriorated.

    “la participación ha alcanzado niveles cercanos a la paridad, y la brecha salarial ha ido disminuyendo de forma lenta, pero progresiva hasta antes de la pandemia.Posteriormente, ambos indicadores se deterioraron.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    In high level, slight relationship contradicting theory: greater participation, greater gap, correlation 0.319.

    “en el nivel alto se observa una ligera relación que contradice la teoría: a mayor participación, mayor es la brecha, con una correlación de 0,319.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § VOLUMEN 18, 2024

    Highest median salary gap in Pasco, 26%, lowest in Ica, 6.3%.

    “la mayor brecha salarial mediana en Pasco, con un 26 %, y la menor en Ica, con un 6,3 %.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § C. Fase III, preparación de datos

    Cluster 1 has greater salary gap, lower participation, and lower GDP per capita than cluster 2.

    “en el clúster 1, la brecha salarial es mayor, la participación es menor y el PBI per cápita de la región es más bajo.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § VOLUMEN 18, 2024

    Apurímac has highest dispersion (IQR 25%), Piura lowest (2%).

    “Apurímac presenta la mayor dispersión, con un 25 %, mientras que Piura tiene la menor, con un 2 %.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Fase II, entendiendo datos

    Participation improved from 46.9% in 2017 to 48.3% in 2020, retreating to 47.3% in 2021.

    “la participación mejoró del 46,9 % en 2017 al 48,3 % en 2020, retrocediendo al 47,3 % en 2021.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § D. Fase IV, modelado y Fase V, evaluación

    Pasco has highest IQR for participation (7%), Ica lowest (practically 0%).

    “Pasco presenta la mayor, con un 7 %, y la menor corresponde a Ica, con prácticamente un 0 %.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § INTRODUCCIÓN

    El nivel de calificación educativa de las mujeres es mayor que el de los hombres en todos los niveles del Estado peruano; en promedio, 55 % de las mujeres tiene educación universitaria o avanzada concluida, frente al 39 % de los hombres.

    “el nivel de calificación educativa de las mujeres es mayor que el de los hombres en todos los niveles del Estado; en promedio, un 55 % de las mujeres tiene educación universitaria o avanzada concluida, frente al 39 % de los hombres”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [3]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation
    § INTRODUCCIÓN

    El informe de Servir de 2024, con datos de 2022, indica que el sector público general representa el 16,7 % de la PEA asalariada, con una participación femenina del 47,3 %.

    “el informe de 2024, con datos de 2022, indica que el sector público general representa el 16,7 % de la PEA asalariada, con una participación femenina del 47,3 %”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [3]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation
    § INTRODUCCIÓN

    Para 2022, el problema de brecha salarial en el Estado peruano parece haberse superado, persistiendo únicamente en el nivel de gobierno local.

    “Se observa que, para 2022, el problema parece haberse superado, persiste únicamente en el nivel de gobierno local”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [3]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data
    § Fase II, entendiendo datos

    At region-year level, there is great variability, suggesting possible existence of hidden gap when analyzing at state level.

    “se observa una gran variabilidad, lo que propone la posible existencia de una brecha oculta al analizar los datos a nivel estatal.”
    • Also stated in: § CONCLUSIONES · plain
      “A nivel regional, la realidad era diferente: la brecha salarial y la participación mostraron valores muy diversos y alejados del promedio general a nivel estatal.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § VOLUMEN 18, 2024

    Salary gap by region shows great dispersion, confirming existence of problem when disaggregating at state level.

    “La revisión de la brecha salarial por región muestra una gran dispersión, lo que confirma la existencia del problema al desagregar los datos a nivel estatal.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationown data
    § Fase II, entendiendo datos

    At state aggregate level, indicators are good, suggesting problem almost nonexistent.

    “La revisión a nivel agregado del Estado peruano muestra buenos indicadores, lo que podría sugerir que el problema es casi inexistente.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § INTRODUCCIÓN

    Con base en datos de la OIT, se observa un avance significativo en la brecha salarial de América Latina y el Caribe desde 2013 (18,3 %) hasta 2019 (14 %).

    “Con base en sus datos, se ha elaborado la tabla 1, en la que se observa un avance significativo desde 2013 (18,3 %) hasta 2019 (14 %) [12].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [12]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

25 of 25

  • The paper relies on this without support.

    El PBI per cápita está asociado al progreso de la región y la cobertura por cada servidor público está asociada a los servicios del Estado.

    “el primer factor asociado al progreso de la región y el segundo asociado a los servicios del Estado”

  • The paper relies on this without support.

    La información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir) es suficientemente precisa y completa para estimar la brecha salarial de género en el Estado peruano.

    “con base a la información de datos abiertos proporcionada por la Autoridad Nacional del Servicio Civil (Servir)”

  • The paper relies on this without support.

    The general gender salary gap in Latin America was 14% in 2019.

    “Comparado con la brecha general en América Latina, que era del 14 % en 2019”

  • The paper relies on this without support.

    The gender salary gap in the Peruvian State is a social problem that requires attention.

    “lo que confirma la existencia del problema al desagregar los datos a nivel estatal.”
4 of 4

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Six regions are called significant at the 5% level, but three exceed it

    Major

    The table reports p-values of 0.058 for Áncash, 0.063 for San Martín and 0.091 for Huancavelica. None is significant at the stated 5% level, so the conclusion that three regions advanced and three regressed at that level does not follow from the reported results.

1 of 1

Cites

The publisher hasn’t deposited this paper’s reference list.

Authors
León Leyva, R.
Venue
Revista I+i
Publisher
Tecsup No 1
Published
2024-12-27
Volume
18
Pages
94-112
ISSN
  • 1996-7551 (ISSN-L)
  • 2707-9368
DOI
10.71701/revistaii.v.18.2024.91
OpenAlex
W4417437342
Semantic Scholar
f196c8e4720a557e0cdfa85ddc32034e12381ddd
Cited by
0
References
0
FWCI
0.00
Type
journal article
Language
es
Open access
Licence
Locations
  • Revista I+i (publishedVersion, cc-by-nc)
  • Revista IBERC
Topics
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Labor market dynamics and wage inequality
  • Social Sciences › Gender Studies › Labour Market and Migration
  • Social Sciences › Safety Research › Employment, Labor, and Gender Studies
Keywords
gender pay gap, public sector wages, regional disparities, female labor force participation, COVID-19 pandemic impact, GDP per capita, labor discrimination

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-19.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-19.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review