OIT - Returns to education in heterogenous labour markets The case of Peru

X Returns to education in heterogenous labour markets The case of Peru Authors / Juan Chacaltana, Martin Moreno

July / 2025 ILO Working Paper 147© International Labour Organization 2025 Attribution 4.0 International (CC BY 4.0) This work is licensed under the Creative Commons Attribution 4.0 International. See: https:// creativecommons.org/licenses/by/4.0/. The user is allowed to reuse, share (copy and redistribute), adapt (remix, transform and build upon the original work) as detailed in the licence. The user must clearly credit the ILO as the source of the material and indicate if changes were made to the original content. Use of the emblem, name and logo of the ILO is not permitted in connection with translations, adaptations or other derivative works. Attribution – The user must indicate if changes were made and must cite the work as follows: Chacaltana, J., Moreno, M. Returns to education in heterogenous labour markets : The case of Peru . ILO Working Paper 147. Geneva: International Labour Office, 2025.© ILO. Translations – In case of a translation of this work, the following disclaimer must be added along with the attribution: This is a translation of a copyrighted work of the International Labour Organization (ILO). This translation has not been prepared, reviewed or endorsed by the ILO and should not be considered an official ILO translation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the translation. Adaptations – In case of an adaptation of this work, the following disclaimer must be added along with the attribution: This is an adaptation of a copyrighted work of the International Labour Organization (ILO). This adaptation has not been prepared, reviewed or endorsed by the ILO and should

not be considered an official ILO adaptation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the adaptation. Third-party materials – This Creative Commons licence does not apply to non-ILO copyright materials included in this publication. If the material is attributed to a third party, the user of such material is solely responsible for clearing the rights with the rights holder and for any claims of infringement. Any dispute arising under this licence that cannot be settled amicably shall be referred to arbitration in accordance with the Arbitration Rules of the United Nations Commission on International Trade Law (UNCITRAL). The parties shall be bound by any arbitration award rendered as a result of such arbitration as the final adjudication of such a dispute. For details on rights and licensing, contact: rights@ilo.org. For details on ILO publications and digital products, visit: www.ilo.org/publns.

ISBN 9789220423394 (print), ISBN 9789220423387 (web PDF), ISBN 9789220423400 (epub), ISBN 9789220423417 (html). ISSN 2708-3438 (print), ISSN 2708-3446 (digital) https://doi.org/10.54394/MQIL6393

The designations employed in ILO publications, which are in conformity with United Nations practice, and the presentation of material therein do not imply the expression of any opinion whatsoever on the part of the ILO concerning the legal status of any country, area or territoryor of its authorities, or concerning the delimitation of its frontiers or boundaries. See: www.ilo. org/disclaimer. The opinions and views expressed in this publication are those of the author(s) and do not necessarily reflect the opinions, views or policies of the ILO.

Reference to names of firms and commercial products and processes does not imply their endorsement by the ILO, and any failure to mention a particular firm, commercial product or process is not a sign of disapproval. Information on ILO publications and digital products can be found at: www.ilo.org/researchand-publications ILO Working Papers summarize the results of ILO research in progress, and seek to stimulate discussion of a range of issues related to the world of work. Comments on this ILO Working Paper are welcome and can be sent to chacaltana@ilo.org.

Authorization for publication: Sangheon Lee, Director, Employment Policy Department ILO Working Papers can be found at: www.ilo.org/research-and-publications/working-papers Suggested citation: Chacaltana, J., Moreno, M. 2025. Returns to education in heterogenous labour markets : The case of Peru , ILO Working Paper 147 (Geneva, ILO). https://doi.org/10.54394/MQIL639301 ILO Working Paper 147

Abstract This paper examines the private returns to education in the context of heterogeneous labour markets, using Peru as a case study. While education is widely considered a key driver of individual earnings and economic growth, its actual returns may be constrained by structural features of labour demand. Drawing on nationally representative household survey data from 2016, 2019, and 2022, we estimate Mincerian earnings equations with Heckman selection correction to assess how returns to education vary by employment status (employee vs. own account) and sector (formal vs. informal). Our findings confirm significant positive returns to education overall, particularly at the tertiary level. We also find that while own-account and informal workers earn significantly less on average; returns to education vary by education level and sector. At the primary and secondary levels of education, returns are higher among informal and own-account workers, while at the post-secondary non-tertiary and tertiary levels, returns are higher in formal employment. Among own-account workers, returns at the tertiary level are comparable to those of salaried workers. These findings suggest that improving educational outcomes alone

mary and secondary levels of education, returns are higher among informal and own-account workers, while at the post-secondary non-tertiary and tertiary levels, returns are higher in formal employment. Among own-account workers, returns at the tertiary level are comparable to those of salaried workers. These findings suggest that improving educational outcomes alone is not sufficient; targeted policies are also needed to expand formal, high-productivity employment, strengthen school-to-work transitions, and enhance the earning potential of workers in informal and own-account roles. About the authors Juan Chacaltana is a Senior Employment Policies Specialist at the International Labour Organization (ILO) in Geneva. He holds a PhD in Economics from the Pontificia Universidad Católica del Perú and an MSc in Economics from Texas A&M University. Previously, he led the ILO's programme on formalisation for Latin America and the Caribbean, worked as an ILO regional economist for Latin America and the Andean region, and coordinated the UN MDG youth employment programme in Peru. His research spans informality, youth employment, human capital, and labour-market policy. Martin Moreno is an independent researcher and consultant. He holds an MA in Sociology from The Pennsylvania State University and doctoral studies in Demography and Sociology from the same university. Mr Moreno has worked extensively in the fields of education and development, including analysis on student learning in Latin America and Africa, and co-authoring and contributing to several reports and publications.02 ILO Working Paper 147 Abstract 01 About the authors 01 X Introduction 05 X 1 Related literature 07 X 2 Data 10 X 3 Overall returns to education 14 X 4 Returns to education with heterogeneous labour demand 18 Employees / own account work 19 Formal / informal 23 X Conclusion 27 Annex 29 References 44 Acknowledgements 47 Table of contents03 ILO Working Paper 147 List of Figures Figure 1. Peru. Share of higher education and formal employment in total employment (%) 06 Figure 2. Peru. Educational attainment of the workforce (%) 12

Annex 29 References 44 Acknowledgements 47 Table of contents03 ILO Working Paper 147 List of Figures Figure 1. Peru. Share of higher education and formal employment in total employment (%) 06 Figure 2. Peru. Educational attainment of the workforce (%) 12 Figure 3. Peru. Average real hourly income, deflated to Dec 2022 13 Figure 4. Estimated coefficients by estimation method 16 Figure 5. Estimated earnings premiums by educational levels and employment status (2016–

1. 22

Figure 6. Estimated earnings premiums by educational levels and formality status (2016–2022) 26 Figure A1. Peru. Income density distributions 29 Figure A2. Earnings profiles in Peru 3004 ILO Working Paper 147 List of Tables Table 1. Descriptive Statistics 11 Table 2: Overall returns to education levels 15 Table 3. Returns to education of employees and own account workers 20 Table 4. Returns to education with informality interaction terms 24 Table A1: Overall returns to education levels, full specification 31 Table A2. Returns to education of employees and own account workers, full specification 34 Table A3. Returns to education of employees and informality status in unit of production, full specification 37 Table A4. Earnings premium by education level and employment status 40 Table A5. Earnings premium by education level and formality status 4205 ILO Working Paper 147 X Introduction Education is widely recognised as an investment that yields substantial rewards, both for individuals and for the broader economy. Empirical evidence supports the fact that people with higher education tend to have higher earnings than less educated ones. An increasing number of empirical studies on private returns to education has allowed identifying important features of these returns. Psacharopoulos and Patrinos (2018) for example summarise that these returns are positive at around 10% per additional year of education; returns to primary education are the highest; and these returns are higher in lower income countries due to the relative scarcity

these returns. Psacharopoulos and Patrinos (2018) for example summarise that these returns are positive at around 10% per additional year of education; returns to primary education are the highest; and these returns are higher in lower income countries due to the relative scarcity of educated workforce. In terms of trends, some studies show a decreasing trend of returns to education over time, although there is some discussion around it. Many governments around the world embark in massive educational programs and policies to boost education based on these findings. However, there is an open discussion on whether these private returns lead to benefits beyond the individual level. While some studies find evidence of positive social returns1 other authors find different results2. In particular, Pritchet (2001) using cross-national data show no association between increases in rising educational attainment of the labour force and the rate of growth of output per worker. This led him to pose his famous question: where has all education gone? He provided three possible responses to this question. ● First, the institutional/ governance environment could have been sufficiently perverse that the accumulation of educational capital does not enhance (or even lowers) economic growth. ● Second, marginal returns to education could have fallen rapidly as the supply of educated labour expanded while demand remained stagnant. ● Third, educational quality could have been so low that years of schooling created less (or even no) human capital. In this paper, we focus on the limits imposed by labour demand and introduce the idea that this limit is even tighter in the context of heterogeneous labour demand. We assess labour demand heterogeneity in terms of status in employment (wage employment/own account work) and informality (employment in the formal or informal sectors). Our hypothesis is that heterogeneous labour demand– especially when the labour market produces low productivity jobs - impose an additional limit to overall productivity growth. For this, we use data from Peru a developing country with high informality and own account work rates and, at the same time, with rapidly increasing post-secondary education. According to data from the National Institute of Statistics of Peru (INEI) workers with tertiary education inadditional limit to overall productivity growth. For this, we use data from Peru a developing country with high informality and own account work rates and, at the same time, with rapidly increasing post-secondary education. According to data from the National Institute of Statistics of Peru (INEI) workers with tertiary education increased from 23% in 2004 to 32% in 2022 (Fig 1). However, the share of workers with formal employment increased from 20% to 24% in the same period. An increasing share of workers with tertiary education and a slower growth of formal employment means that an increasing number of workers with tertiary education end up working in the informal economy, which must influence the overall returns to education. 1 Canton (2007) using data for OECD countries finds that the long-term social returns to education are between 11%-15% and the short-term social returns are between 7.5%-10%, similar to private returns. Cui, Y., & Martins, P . S. (2021) performed a meta-analysis of 32 studies on the social returns to education across 15 countries since 1993 and find that the social returns to education are consistently higher than the private returns, with an average rate of return of 10.3%. The study also found that the social returns to education are higher in developing countries than in developed countries. 2 Psacharopoulos and Patrinos (2018) find that social returns are universally lower than private returns because of the public subsidisation of education and because studies include all costs but not all benefits.06 ILO Working Paper 147 X Figure 1. Peru. Share of higher education and formal employment in total employment (%) Source. Data processed by INEI based on INEI-ENAHO for the 2004-2022 period.07 ILO Working Paper 147 X 1 Related literature

The international discussion on returns to education has seen a significant recent advance with the dissemination of databases containing empirical estimates for an increasing number of countries (Card, 1991; Psacharopoulos 1994; Psacharopoulos and Patrinos 2014; Montenegro

X 1 Related literature

The international discussion on returns to education has seen a significant recent advance with the dissemination of databases containing empirical estimates for an increasing number of countries (Card, 1991; Psacharopoulos 1994; Psacharopoulos and Patrinos 2014; Montenegro and Patrinos 2013; Psacharopoulos and Patrinos 2018). These studies provide global evidence on the levels of returns to education, their evolution over time, and highlight important levels of individual heterogeneity, especially by gender, age, educational attainment, experience, and socio-economic characteristics. There is less evidence on heterogeneity from the demand side, that is, differences across the sectors of the labour market in which workers are employed. Some studies have addressed the differentials between self-employment and wage employment and findings are not always straightforward. Lee (1987) incorporates occupational choice into a model of wage determination in wage and self-employed activities, finding that the two groups have different earnings structure. While the results for employees confirm the robustness of the human capital model, the same model explains only a small fraction of the variation of self-employed earnings.3 Yamada (1996), using Peruvian household survey data from 1985–86 and 1990 finds that self-employed workers in the urban informal sector enjoy significant earnings premium over wage earners, as self-employed individuals tend to possess higher entrepreneurial ability. Conversely, wage earners attributes like discipline or teamwork, are more valued in structured employment but less critical in self-employment. Idrus and Cameron (2000), examine returns to education between self-employed and wage-earners in rural Malay and finds that returns to education do not significantly differ between self-employed and wage-employed workers. However, returns to education do vary by education level, with the highest returns occurring at the secondary education level (14.2%), compared to primary (12.7%) and university (13.5%). In Uganda Kavuma, Morrissey, and Upward (2015) find that returns to education are roughly equal across sectors,

to education do vary by education level, with the highest returns occurring at the secondary education level (14.2%), compared to primary (12.7%) and university (13.5%). In Uganda Kavuma, Morrissey, and Upward (2015) find that returns to education are roughly equal across sectors, but the earnings profile differs: convex for wage employees and concave for the self-employed. This discussion is also present in developed countries.Iversen et al (2010) using data from the Danish Labour market conclude that the traditional log linear specification seems to work well for wage employment, but it seems to be highly nonlinear in the case of the self-employed.4 In the case of the self-employed they find low returns for lower educational levels but significantly higher returns for 18 years of more years of education. Williams (2003), using longitudinal German data, finds that returns to education are significantly lower for the self-employed than for wage earners, with education having no significant effect on self-employment earnings. He also shows that self-employment experience is less rewarded when individuals return to wage work, suggesting limited transferability of sector-specific human capital. Iglesias, Carmona, and Ferradás (2016) using EU data, distinguish between wage earners, own-account workers, and employer entrepreneurs and find that tertiary education offers the highest returns for employer entrepreneurs relative to wage employees. In the case of own-account workers they benefit most from secondary education, with returns surpassing those of wage earners at this level, whereas tertiary education does not confer additional advantages for this group. In general, 3 Lee-Ying Soon (1987). Self-employment vs wage employment: Estimation of earnings functions in LDCs. Economics of Education review. 4 Iversen, J; Maclchow-Moller, N; and Sorensen, A (2010) Returns to schooling for the self-employed. Economic letters.08 ILO Working Paper 147 they suggest that self-employment, regardless of educational attainment, is associated with a general earnings disadvantage compared to wage employment5.

they suggest that self-employment, regardless of educational attainment, is associated with a general earnings disadvantage compared to wage employment5. Other studies perform breakdowns by formal/informal sectors and seem to indicate that the same education levels yield less returns in the informal sector. In China, a study by Albert Park and Qu (2013) found that the returns to education in the informal sector in China were lower than in the formal sector. The study found that each additional year of schooling increased earnings by 4.2% in the informal sector, compared to 11.1% in the formal sector. They argue that differentials in returns could create arbitrage opportunities for some worker by moving from one sector to another if mobility is unrestricted. Therefore, persisting differentials could indicate restrictions in mobility. Another study by Herrera, Lopez and Montellon (2013) in Colombia, based on the fact that overeducation seems to be larger in the informal sector, estimate return to education for those in undereducation, those in actual education and those in overeducation. They find that in the informal sector not only the returns to correct years of education are lower, but the penalty that informal workers face due to educational mismatches are higher than in the formal sector. Mendiata Ossio (2022) using data from Bolivia, finds nonlinearity in the earnings equation and hypothesises that this pattern could be due to informal and self-employed workers having lower returns to education than those in formal employment. Specifically, they find returns to education of 6.6% for the self-employed, 10.1% for informal employees and 41.6% for formal employees. More recently, Montenegro and Patrinos (2022) estimate returns to education for two sectors of the economy, public and private, using data for 28 European and Central Asian countries. They find that the effect of education on earnings is stronger in the private sector and argue that this provides evidence that wage determination in this sector is determined by economic variables, such as education, and that strong screening is not as widespread as in the public sector. In the case of Peru, there are several studies with estimates of overall returns to education and

provides evidence that wage determination in this sector is determined by economic variables, such as education, and that strong screening is not as widespread as in the public sector. In the case of Peru, there are several studies with estimates of overall returns to education and some references to demand side heterogeneity. Rodríguez (1993) found that primary education presented the highest rates of return using the National Living Standards Measurement Household Survey (ENNIV 1991). Yamada (2007) provides a comprehensive outlook on returns to education and finds that that from 1985 to 2004 the average returns to education in Peru fluctuated around 10% with procyclical fluctuations and with presence of "convexity" in the returns to education, meaning that individuals with higher educational attainment, particularly at the university level, see a disproportionately larger increase in income. This study also found that the average return per additional year of education for salaried workers was higher than for independent (self-employed) workers. These findings are confirmed some years later in Arpi and Arpi (2016) who report a pattern of lower returns for independent workers. Regarding the formal and informal divide, Saavedra and Chong (1999) highlight that after controlling for observable characteristics like education and experience, earnings differentials between formal and informal wage earners remain in favour of formal work, suggesting that factors, such as the nature of the work (repetitive tasks requiring few intellectual abilities), could limit the returns to education for informal wage earners. Interestingly, they also find that returns to potential experience were higher for formal self-employed workers than for formal wage earners and informal self-employed workers. 5 Complementing these findings, van der Sluis, van Praag, and Vijverberg (2004) conduct a meta-analysis of 94 studies examining how formal schooling affects entry into self-employment and entrepreneurial performance. They find that education does not significantly influence the likelihood of becoming self-employed but is positively and significantly associated with performance once self-employment is chosen—particularly in terms of earnings. The authors underscore methodological differences that complicates cross-study

formal schooling affects entry into self-employment and entrepreneurial performance. They find that education does not significantly influence the likelihood of becoming self-employed but is positively and significantly associated with performance once self-employment is chosen—particularly in terms of earnings. The authors underscore methodological differences that complicates cross-study comparisons.09 ILO Working Paper 147 Other studies highlight other sources of heterogeneity. Schaffner (1998) estimates returns to education by employer size using data from the 1985 Peru Living Standards Measurement Study (LSMS) and find that returns to education are significantly higher in larger firms than in smaller firms6. Calónico and Ñopo (2007) and Yamada (2007) suggest that attending private schools generally results in numerically higher returns compared to public schools. Yamada and Castro (2002) found that returns were higher in urban areas and formal employment sectors, with women experiencing marginally higher returns than men, particularly at the tertiary level. Arpi and Arpi (2015) also highlight heterogeneity by residence area sex and employment status. In sum, the literature highlights considerable heterogeneity in the returns to education, shaped not only by individual characteristics such as gender, experience, and education level, but also by demand-side factors like employment status, sector (formal vs. informal), firm size, and geographic location. 6 Schaffner (1998) "Premiums to Employment in Larger Establishments: Evidence from Peru"10 ILO Working Paper 147 X 2 Data

We use data from the National Household Surveys (ENAHO) produced by the National Institute of Statistics (INEI) of Peru. This survey includes a module specialised on employment applied to all household members ages 14 or older with national coverage and representative of urban and rural areas, and regions. In both cases, we use original microdata files harmonized by the ILO Department of Statistics for 2016, 2019 and 2022.7 The criteria used for inclusion in the analytical sample restricted cases of individuals of ages between 15 and 65 years old who at the time of the survey were part of the labour force.

ILO Department of Statistics for 2016, 2019 and 2022.7 The criteria used for inclusion in the analytical sample restricted cases of individuals of ages between 15 and 65 years old who at the time of the survey were part of the labour force. The dependent variable is the log of the hourly income in the main (principal) job including employees and own account workers. The variable is derived from the monthly income divided by the total number of days per months and the number of hours worked per week. Educational attainment is defined as the maximum number of years of schooling attained. As per the educational levels, we group them into 5 categories: primary or less; lower secondary, upper secondary, post-secondary non tertiary, and bachelor or above.8 Informal sector is defined at the unit of production level, following the ILO definition comprising “economic units that are producers of goods and services mainly intended for the market to generate income and profit and that are not formally recognised by government authorities as distinct market producers and thus not covered by formal arrangements”.9 Two subsets of additional variables are included in the analysis. A subset of individual characteristics including gender (male, female), potential work experience (measured as years as the age minus the number of years of schooling minus 6 years or the age of entry into the school system); marital status, and geographic location (rural, urban). The second subset of variables are job-related characteristics including: firm economic sector, broad sectorial classification based on ISIC revision 4: agriculture, industry and services. Table 1 shows descriptive statistics for these variables and this sample. Employees account for over 55% of the Peruvian workforce, reaching 56.48% in 2022. Own-account workers constituted a significant share as well, making up 39.66% in 2022, while employers remained a relatively small proportion, declining to 3.86% in 2022. 7 Note that the period under study includes structural but also cyclical components —notably COVID-19 and political instability—which could affect the trend in returns.

small proportion, declining to 3.86% in 2022. 7 Note that the period under study includes structural but also cyclical components —notably COVID-19 and political instability—which could affect the trend in returns. 8 Primary or less includes non-schooling, pre-primary, primary incomplete and complete. Lower secondary includes from 7 to 9 years of education. Upper secondary includes 10 and 11 years of education. Post secondary non tertiary includes technical and vocational education and incomplete university education. Bachelor or higher includes bachelor or equivalent and post graduate studies. 9 See ILO (2023). Resolution concerning statistics on the informal economy. 21st International Conference of Labour Statisticians. Geneva, 11–20 October 202311 ILO Working Paper 147 X Table 1. Descriptive Statistics 2016 2019 2022 Status in employment (ICSE 93) - Main job 1 - Employees 56.30 55.59 56.48 2 - Employers 4.85 4.39 3.86 3 - Own-account workers 38.85 40.03 39.66 Education Levels Primary or less 21.98 20.13 17.82 Lower secondary 10.92 10.17 9.42 Upper Secondary 42.85 43.97 46.03 Post secondary Non-Tertiary / Short-cycle Tertiary 11.66 12.27 11.90 Bachelor or higher 12.59 13.45 14.82 Economic activity (Sector) - main job Agriculture 19.27 18.64 18.33 Industry 19.71 18.89 20.13 Services 61.02 62.47 61.54 Geographical coverage Urban 81.18 82.31 82.94

Rural 18.82 17.69 17.06 Region (Geographic Domain) Northern Coast 14.99 15.41 13.81 Central Coast 7.16 7.08 7.92 Southern Coast 2.18 2.14 2.99 Northern Highlands 5.25 5.26 4.63 Central Highlands 10.74 10.73 10.44 Southern Highlands 13.12 12.98 14.11 Jungle/Amazonia 12.37 12.37 12.11 Lima Metropolitan Area 34.19 34.02 34.00 Sex Male 60.60 59.62 59.09 Female 39.40 40.38 40.91 Marital Status Single 40.26 42.32 44.70 Married 59.74 57.68 55.30 Age (mean) 38.54 39.17 39.19 Share of children below 15 0.23 0.19 0.21 Share of members ages 65+ 0.05 0.06 0.06 Informality Informal Sector (unit of production) 46.49 48.83 49.78 Informal employment 66.56 67.48 68.70 Source. Own’s elaboration based on INEI-ENAHO 2016, 2019 and 2022. Employment in the informal sector increased in the period, and accounted for 49.78% of total employment in 2022, while informal employment, also increased and accounted for 68.7% in

2022. This means that informal employment outside the informal sector (in the formal sector and in households explai