-The Labor Market Effects of Facilitating Social Security Contributions under Part-Time Employment Contracts: Evidence from Colombia (joint with Brenda Samaniego de la Parra and Leonardo Morales). Labour Economics, Vol. 91 (December 2024)
We examine the impact of reducing rigidities caused by regulation on labor demand in a context with high informality. Using employer–employee matched administrative records and household survey data, we estimate the effects of a reform that eliminated a wedge in firms' regulatory costs of employing workers on different work schedules in Colombia, reducing the relative costs of formal part-time employment. We find that the reform increased the probability of entering the formal sector for previously informal part-time workers (the target population). Using pre-reform variation in firms’ demand for workers eligible for the policy, we find a 5.5 percent increase in formal employment at firms more exposed to the policy’s cost reduction. Mean wages at these firms rose by 1.6 percent after the reform relative to those at firms that tend to hire fewer workers with no formal sector experience. Firms with more exposure to the reform also experienced higher churn, consistent with the policy creating incentives for firms to rotate across workers at a faster pace.
-Heterogeneous Returns of Informality: Evidence From Brazil - (Latin American Economic Review, Vol 31, December 2022)
This paper estimates the marginal treatment effect of informality on wage rate for Brazil. The primary data source is the Pesquisa Nacional por Amostra de Domicilios (PNAD) for 2015, which is a household survey with information about workforce indicators, marital status and socio-economic characteristics. We use a combination of regional data on economic and geographic indicators at the state level and institutional data on labor inspections for identification. The results indicate that informality in Brazil responds to comparative advantage. Thus, workers self-select into the type of jobs that better reward their skills. Formal workers do not have, on average, higher wage rates than informal workers, but there is large and signicant heterogeneity in the returns of formality. Therefore, for workers with very low cost of being formal, formality offers signicantly higher wage premiums.
Desequilibrios en el mercado laboral de las ciudades colombianas: ¿Qué hace al Caribe different? (con Oriana Alvarez, Karina Acosta y Alexander Villegas). Conyutura Económica: Investigación económica y social, Núm. 54 (Diciembre 2024).
Este trabajo estudia los indicadores del mercado laboral en las principales ciudades colombianas. Asimismo, propone una alternativa empírica de la estimación de la curva Beveridge en contextos de alta informalidad, para tener una aproximación al potencial desequilibrio existente entre oferta y demanda laboral. El documento concluye que una menor creación de vacantes formales y las fallas en calidad y pertinencia de la educación de la fuerza laboral podrían explicar las diferencias en los indicadores laborales de las ciudades capitales de la región Caribe, cuyos resultados de equilibrio general exhiben peores indicadores laborales en comparación con las del resto del país.
Evolución de las transferencias monetarias en Colombia (con Karina Acosta, Bibiana Taboada-Arango y Jaime Bonet-Morón). Revista CEPAL No. 142 (Abril 2024).
Los programas de protección social se han convertido en una herramienta indispensable y muy efectiva para mejorar las condiciones de vida de la población en situación de pobreza y vulnerabilidad. Este documento revisa los cinco principales programas de transferencias monetarias en Colombia, y estudia su historia, focalización, cobertura, operatividad y los compromisos fiscales derivados de su implementación. Asimismo, hace una revisión de las evaluaciones de impacto de estos programas y sus resultados en múltiples indicadores sociales. Este trabajo es un primer intento de consolidar la información existente sobre el tema en Colombia, que permite entender de manera comprehensiva el alcance y retos de estos programas, así como identificar posibles áreas de investigación y mejoras.
-Desigualdades en el Mercado Laboral Urbano-Rura en Colombia, 2010-2019 (Joint with Edson Acosta) (Revista CS, No. Especial, Julio 2022)
La persistencia de la desigualdad al interior de Colombia tanto en ingresos como en oportunidades ha sido ampliamente documentada. Sin embargo, poco se sabe sobre las desigualdades en el contexto urbano-rural. Este trabajo estudia las desigualdades urbano-rurales desde el contexto del mercado laboral mediante el uso de una descomposición de tipo Blinder-Oaxaca, en la cual se encuentra que cerca del 60% de la brecha en ingresos laborales se explica por factores observados de los trabajadores, en donde la educación es el elemento principal y el restante 40% se debe a características no observadas como preferencias de los empleadores, habilidad de los trabajadores o calidad de la educación. En este trabajo se identificaron tres características del mercado laboral rural que pueden contribuir a explicar las diferencias actuales en ingresos entre trabajadores urbanos y rurales: (i) la participación laboral y la tasa de ocupación femenina en zonas rurales es baja comparada no solo con la de los hombres sino también con respecto a la de las mujeres ubicadas en las cabeceras; (ii) las tasas de informalidad laboral para trabajadores rurales son significativamente más altas que las urbanas; y, (iii) el trabajo infantil aún es una práctica común en las zonas rurales que, aunque ha disminuido en el tiempo, todavía hay a una proporción significativa de niños, niñas y adolescentes que trabajan de forma ilegal y, además, no asisten al colegio por estar trabajando.
-The Impact of Hard Discount Stores on Local Labor Markets: Evidence from Colombia (joint with Lukas Delgado-Prieto and Andrés Calderón). Submitted.
Hard discount stores have reshaped the retail sector by selling low-cost products. While this business model has gained market shares in many countries, how it affects the labor market is unclear. To fill this gap, we study the impact of discount stores on local labor markets in Colombia, where these stores had a rapid, staggered geographic expansion. Our results show that discount stores boost local formal employment, especially in retail, manufacturing, and agriculture, suggesting significant spillover effects from retail to other sectors. Consistent with this finding, we also document increases in local tax revenues from manufacturing and commerce activities.
-Live or Let Die: Formality, Firm Creation and Firm Growth in Colombia (joint with Christian Posso, Leidy Gómez, and Gustavo García).
This paper aims to study startup survival determinants by evaluating the impact of a regulatory reform implemented in Colombia that intended to promote the creation of formal start-ups by cutting down registration costs and taxes. Using a rich combination of administrative records that allow us to follow in time the universe of formal firms in the country, we analyze static and dynamic effects on new formal firms' creation and the quality of the new firms created by testing whether they had a higher probability of survival due to the reform, and whether the firms created faced fewer financial constraints through access to formal credit. We find that the reform increased the number of formal small firms registered by 30%, but we do not find evidence that the newly formalized firms that benefited from the reform had a higher probability of survival, which means that reducing formalization costs could not be enough to improve the quality of the new firms created. We, therefore, consider Sterk, Sedlacek, and Pugsley (2021) statistical model to rationalize the determinants of firm size, finding that small-sized firms experience higher ex-ante heterogeneity than larger firms, which can potentially translate into higher exit rates as firms with higher ex-ante heterogeneity are less suited for growth and more vulnerable in the presence of financial frictions. Draft available upon request!
-The Heterogeneous Labor Market Effects of the Venezuelan Exodus on Female Workers: Evidence from Colombia (joint with Ana María Tribín and Tatiana Mojica).
We study the labor market effects of the Venezuelan migration shock on female labor market outcomes in Colombia using a Bartik-type approach. Our findings suggest that in the labor market, female migrants can act as substitutes for native-born women or complements depending on the education level of native women. Migrant workers are substitutes for native-born low-educated women because they compete for similar jobs. Therefore, labor force participation and earnings decrease for low-educated women. At the same time, unpaid domestic work and care increase for low-educated native women, possibly preventing the job search for this group. However, there is an increase in participation of 1.5 p.p. for highly educated women with the presence of minors at home and a 1 p.p. higher likelihood of becoming entrepreneurs due to the migratory shock, which supports the complementary-skill hypothesis. We find no effects on the unemployment rate or working hours for highly educative native female workers nor evidence that the migratory shock induced households to outsource more home production as a means for women to spend more time at paid work.
-Blame it on the Rain: The Effects of Weather Shocks on Rural Formal Employment in Colombia (joint with Camilo Bohorquez-Penuela)
Episodes of excessive or low rainfall have not only become more frequent, but also more severe. These events can aect agricultural production and local labor markets. By combining cm security records, that allow us to measure formal employment in rural areas and average earnings, with administrative data from weather stations, we estimate the eects of municipality-level precipitation shocks on formal rural employment in Colombia and the heterogeneous eects of having access to irrigation technologies for mitigating these shocks. Fixed eect estimates show that monthly episodes of excessive rainfall -measured as those that are above the 90th percentile of historical mean precipitation in the last 30 years for each municipality- have a negative impact on formal employment in rural areas for both the agricultural and non-agricultural sector, On the other hand, episodes of lack of rainfall (i.e., below the 10th percentile) affect the formal rural labor market in the opposite direction. These results vary depending on the crop type. Additionally, we explore if such weather shocks translate into productivity shocks by looking at price changes in some agricultural goods in the absence of production data, finding that episodes of lack of rainfall act as a negative productivity shock.
-Heterogenous effects of informality: An Application to Labor Regulation Policy in Russia (with K. Peter)
This paper estimates the heterogeneous returns of informality in Russia using a Marginal Treatment Effects model (MTE) and regional variation in the degree of enforcement of the current labor regulation. The results indicate that self-selection accounts for a relatively large fraction of the difference in wages between formal and informal sector with the wage gap falling from 6.4% to 2.5%, ceteris paribus, and the earnings gap become negative. The findings support a “comparative advantage” hypothesis, as workers are self-selecting into the sector that better rewards their skills. There is also evidence of significant heterogeneity in the size of the formal-informal wage gap depending on individual unobserved cost of being formal. The study also finds significant sector differences in non-pecuniary labor market outcomes, as formal workers have a higher likelihood of receiving benefits such as supplemental medical insurance and paid vacation, and they report to be more satisfied with their jobs, but at the same time formal workers are more concerned about job loss, not finding a job if they get laid off, and they are less satisfied with their pay.
Work in Progress
-Race against the Machines? AI-Augmented Job Search Support for Older Workers in Bogotá (joint with María del Pilar López, Manuel Rojas, Christopher Neilson, Cristina Butnaru, Hernando Zuleta, and Christian Posso).
Older job seekers in Bogotá face significant barriers to re-entering the labor market—from limited digital fluency to self-doubt using online platforms. Adults 50+ in Colombia face low formal employment, partly due to stereotypes about productivity and tech skills. Over 40% of older men and 60% of older women active in the labor market are informally employed or unemployed (Posthuma & Campion, 2009; Medina et al., 2013). These challenges motivate our central research question: Can AI-powered tools, alone or paired with human coaching, improve employment outcomes for older adults?
We explore this through three interrelated sub-questions:
1. Can AI-based tools, such as chatbots, support the re-skilling of older job seekers and enhance their labor market outcomes, including job applications, interview rates, and hiring success?
2. Can older adults effectively use AI tools for re-skilling independently with video tutorials, or is human coaching essential for success?
3. How baseline psychological traits shape engagement with AI-based re-skilling tools, and how these traits may in turn evolve as a result of exposure to such tools?
We ground our RCT design in theories of technology adoption and self-efficacy, which highlight how confidence drives behavior. Prior studies show that even with access to digital tools, fear or frustration often blocks meaningful engagement. Yet modest interventions, like structured support or encouragement, can bridge the gap between intent and action in the job search process (Abel et al., 2019)
-The Revolution is Still Not Welcome Inside Home: Time-use and Gender Roles in Colombia (joint with Juliana Jaramillo, Ana María Tribín, and Juanita Villaveces).
During the XX Century, important social transformations happened in Colombia. Women went from having high fertility, low schooling, low labor market participation to having low fertility, more schooling than men, and a rising labor force participation (Iregui, Melo, Ramírez, and Tribín, 2020). Doing a cross-cohorts descriptive analysis, we find that in 2021, the average time spent doing unpaid work by women is 4 hours and 40 minutes higher than the time spent by men doing the same activities. The gap in unpaid work remains if we only look at the younger cohorts (Millennials and Gen Z), but it almost disappears for older cohorts with higher education degrees. Thus, in order to rationalize such results, we propose that it is the accumulation of feminist discourses and experiences, a form of human capital, that makes women better at bargaining within the household for more redistribution of unpaid work.
-Paternity leave, social norms and labor market outcomes in Colombia (joint with Juliana Jaramillo, Ana María Tribín, Natalia Ramírez and Julián Gallego).