Research

This paper examines the effects of a labor reform in Chile that increased the cost of using temporary agency workers (TAW). Using a dynamic industry equilibrium model and a unique plant-level dataset, we find TAWs are imperfect substitutes for regular workers, have lower productivity and exert less wage bargaining power. The reform increased both fixed and variable costs associated with TAWs, constraining firms' ability to cope with volatility while increasing TAWs' bargaining power. Our model predicts a rise in output prices, with labor misallocation accounting for half of this surge, contingent upon the elasticity of substitution between worker types.

Advertising plays a critical role in shaping societal perceptions of self and others, especially regarding body image. In this paper, we employ computer vision to analyze body image representation in advertisement images, looking for potential differences vis-à-vis the general population. We apply these methods to a sample of Chilean newspapers (2016-2021). Our findings reveal a statistically significant difference between the demographic characteristics of faces portrayed in ads and those of the country's broader population. Both men and women in ads had lower Body Mass Index (BMI) than the population but with a larger bias for women. The estimated effect is, as if, female images faced a 5-to-10 pounds "weight penalty" to be portrayed in ads. We also find that women are under-represented in ads. While the previous literature has already documented some biases in ads, these studies often relied on humans for coding the characteristics of portrayed faces. Our study highlights the scalable and low-cost potential for monitoring ad biases and assessing their impact.

In this paper we exploit the introduction in February 2018 of a new paid parental leave program to care for a seriously ill child in Chile (SANNA) to identify the role of both economic incentives and gender norms on families' decisions regarding market versus home production specialization. To measure the impact of economic incentives, we utilize the design of the SANNA program, which covers the beneficiary's wages up to a specific threshold, beyond which the benefit remains fixed. The efficient allocation of this benefit depends on the income levels of family members and whether their income exceeds the threshold. To investigate the role of gender norms, we compare the effect of economic incentives among older, more traditional families and younger families. Our results indicate that both gender norms and economic incentives affect parental leave allocation. We estimate that older families pay a cost of USD 1,200 for adhering to traditional gender norms compared to younger families.

Labor regulation matters for the economy. But measuring firms' compliance with this regulation poses significant challenges as inspections, enforcement, or reporting might be correlated to compliance risk factors of each firm. We propose estimating the compliance rate by a randomized controlled trial (RCT) of standard labor inspections. We run the experiment on small and medium firms (SME) in Chile. We estimate a non-compliance rate of 13%. Our results are robust to various alternative estimation methods. Additional results look at the effects of enforcing labor laws, showing no significant differences on firms' growth; neither on the intensive (employment and sales) nor the extensive margin (exit). Nonetheless, the experiment had limited power for these additional results. Our methods and results matter for the monitoring and enforcement of labor policies.

This paper tests the granular hypothesis introduced by \citet{Gabaix:vn} in the US, Germany, Canada, and the UK. We find that firm-level idiosyncratic shocks significantly impact aggregate fluctuations in only two of the four countries analyzed. Compared to the US and Germany, the UK and Canada show greater granularity, but also a negligible firm-level contribution to aggregate volatility. Additional results look at the role of the transportation sector as a driver of granular effects in the US and Germany. The explanatory power of firms' idiosyncratic shocks significantly decreases in the US and Germany after removing the largest firms from the sample, many of which belong to the transportation sector, whereas removing firms of comparable importance in Canada and the UK has little to no impact.

Technological advancements and automation have reshaped work. In this paper, we examine gender differences in response to these changes. We identify a narrowing and then reversing gender gap in the probability of automation over the last four decades, which cannot be attributed to a single mechanism. This shift is explained by newer cohorts of women entering lower-risk jobs, who have more years of schooling and are better prepared to access lower-risk jobs than their predecessors. They also face fewer gender norms, which lessens their home production responsibilities and shifts their labor preferences toward jobs with longer hours and less predictable schedules. Often seen as less favorable for maintaining a work-life balance, these occupations typically face a lower probability of automation. We do not find evidence that the closing gender gap in automation probability is explained by increasing returns to social skills or by women's comparative advantages in occupations requiring more social skills.

In this paper I analyze the equilibrium effects of convex capital adjustment costs on quantity dynamics and asset prices in a real business cycle model when the representative agent has Epstein-Zin preferences. Capital adjustment costs make it costly for agents to smooth fluctuations in consumption through the production sector, inducing them to take more consumption risk. I show this model accounts for the main statistical features of macroeconomic aggregate quantities. At the same time, adjustment costs increase the equity risk premium, with the mean stock return and its standard deviation in the order of magnitude consistent with the data. The model also produces a stable risk-free rate, and comes close to matching its average return.

This paper aims to quantify the costs associated with job displacement following a mass layoff, specifically examining the effects on wages and employment. To measure these costs, we consider the duration of unemployment after separation and relevant characteristics of employers and employees within the context of an emerging economy, including firm size, informality, and gender. Our findings reveal that individuals only recover 70% of their earnings after one year of a mass layoff which heavily depends on the low recovery rate of employment (66%), as wages tend to increase (5%).  Individuals entering their first job after 2 months of facing a mass layoff experience wage variations ranging from -3.4% to -0.7% depending on the duration of the unemployment spell before the new job, with these losses generally dissipating after one year of employment. Wage losses are larger for individuals with prior employment in large firms or sectors characterized by high formality levels, but employment recovery rates are also larger. Women generally have a lower rate of wage employment recovery compared to men. However, when women secure new jobs, they tend to experience reduced wage losses compared to their male counterparts.

The decline in the price of new digital technologies, and the increased investment in industrial robots in high-income countries, has reduced production advantages of low-labor-cost countries. Jobs in sectors more susceptible to automation, which were previously offshored to developing and emerging economies, have been replaced by robots. This process has given rise to a reshoring process in which countries with high endowments of robotic capital has started to rebuild multiple productive segments that they had been previously offshored, increasing their share in the generation of value in Global Value Chains. To examine the effects of robotization on trade patterns, wages and welfare we extend the multisector general equilibrium model of Caliendo and Parro (2015) by introducing heterogenous labor at the occupation level, and trade in robots, with robot-labor complementarities.