Startups, Labor Market Frictions, and Business Cycles (Job Market Paper) [PDF]

This paper studies how the labor market conditions affect the formation and growth potential of new businesses over the business cycle. I develop a dynamic occupational choice model in which the labor market conditions shape the business formation through two channels in downturns. First, a higher unemployment rate increases the entry as a stopgap activity. Second, the lower job finding probability, by increasing the cost of a business failure, discourages the entry from employment, especially for highly educated individuals who are precisely those more likely to start high-growth startups. Using U.S. individual-level data, I provide empirical support for these mechanisms. Then, I use the calibrated model to study the macroeconomic implications of these individual-level dynamics during and after the Great Recession. I show that consistently with U.S. firm-level data, (i) entry due to stopgap motives increases while entry into entrepreneurship declines, and (ii) the change in the composition of entrepreneurial founders toward fewer highly educated individuals makes new cohorts of businesses have fewer high-growth startups. Both features hinder job creation recovery, keeping the labor market depressed longer and thus the entrepreneurship entry persistently low.

Presented at: European Winter Meeting of the Econometric Society, 2020 (Nottingham, virtual). Federal Reserve Board Pre-Job Market Conference, 2020 (Washington, DC, virtual). Central Bank of Chile Seminar, 2020 (Sanitago, virtual). Productivity Workshop I: Understanding Productivity, 2019 (Santiago). LACEA-LAMES, 2019 (Puebla).

Earnings Effects of Working at Startups, working paper available upon request

with Nathalie Gonzalez and Alvaro Silva

This paper studies the short- and medium-term effects of working for a startup on earnings. Using Chilean administrative data, we find that those who took jobs at startups earned, on average, 21% less over the next five years than those who transitioned to established firms. However, when considering selection, the difference is reduced to 14%, implying that a significant part of the observed difference comes from sorting. These results are consistent with previous literature, such as Sorenson et al. (2021). We further decompose this effect and find that 11 percentage points of the overall 5-year effect come from lower average wages. The remaining 3% arises from more frequent or more prolonged unemployment spells. Additionally, we find a smaller contemporaneous earnings differential than the average 5-year effect, suggesting the existence of scarring effects. Finally, we also find heterogeneous effects over the business cycle. Transitioning to a startup has a smaller negative impact during recessions.

Composition of Founding Teams over the Business Cycle and Firm Dynamics, work in progress

Does the type of individuals deciding to work for startups vary over the cycle in terms of previous occupation, education, and experience? Does the ability of startups to form founding teams changes in recessions? To address these questions, I develop a dynamic occupational model with firm and worker dynamics, frictional labor markets, and on-the-job search. In this framework, startups offer less job security than older firms, especially during recessions. At the same time, the lower job finding probability reduces job-to-job transitions, implying that individuals might get stuck in an unsuccessful startup, or even worse, fall of the ladder if the business exit. In this environment, individuals might become reluctant to accept jobs at startups during downturns, especially highly educated and experienced workers in higher rungs. This diminishes startups' ability to build organizational capital at the beginning of the operation, affecting their survival probability and future performance. Using administrative data for the U.S., I empirically study how the composition of the founding team in terms of previous occupation, education, and experience varies over the cycle and how these characteristics relate to future business performance. Then, I use the calibrated structural model to assess how the change in the founding team composition over the cycle affects the firm dynamics.

Trade Liberalization and Inequality: Role of Institutions and Financial Markets, work in progress

This paper studies the role of institutions and domestic financial markets in the effect of trade liberalization on income distribution. Not all agents in the economy are able to take advantage of the new opportunities arising from international integration. Their possibilities of undertaking entrepreneurial activities are shaped by access to financial markets, quality of institutions, and level of corruption. A cross country analysis is performed using a panel data sample for 16 developed and 20 emerging economies over the period 1996 - 2005. A standardized Gini coefficient is constructed for each country as measure of income inequality using data from the World Income Inequality Database, UNU-Wider. The results show that in countries with a higher rule of law and more developed financial markets, trade liberalization is associated with a reduction of income inequality. A counterfactual exercise shows that if the emerging economies had had the quality of the legal institutions of the developed countries between 1996 and 2005, the increase in trade opening during this period would have reduced in 1.97 percentage points their Gini coefficient.

Business Formation in the U.S.: Individual-level Analysis using SIPP data, work in progress

Other Published Work

Understanding Domestic Savings: An Empirical Approach [Journal Link] [IDB Working Paper]

with Rodrigo Cerda, Rodrigo Fuentes, and Jose Ignacio LLodra, Applied Economics, 52 (9), 905-928, 2020. Formerly, Inter-American Development Bank, Working Papers Series Nº IDB-WP-626

This paper constructs time series data on savings per type of agent for Chile during the period 1960-2012, and documents two unusual and important features in the evolution of the savings rate. First, the economy increased the average savings rate by 11 percentage points in the period 1985–2013 compared to 1960–1984, mainly due to a large change in private savings rate (10 percentage points), and an additional 1 percentage point from the public sector. The second feature is related to the change in the composition of private savings. After several years of nearly no corporate savings, this component became an important part of total savings reaching an average of almost 10% of Gross National Disposable Income (GNDI) during the period 1986–2012. Our results show that the 1984 tax reform, the boost in the marginal productivity of capital and the deepening of the financial market were the main drivers that explain the dramatic increase in corporate savings. We also found that the reduction in personal income tax after the tax reform and the higher income per capita growth helped to explain the increase in household savings, while the structural balance rule helped to explain the increase in public savings.

The Impact of a Carbon Tax on the Chilean Electricity Generation Sector [Journal Link]

with C. Benavides, L. Gonzales, M. Diaz, R. Fuentes, R. Palma, C. Ravizza, Energies, 8(4), 2674-2700, 2015.

This paper aims to analyse the economy-wide implications of a carbon tax applied on the Chilean electricity generation sector. In order to analyse the macroeconomic impacts, both an energy sectorial model and a Dynamic Stochastic General Equilibrium model have been used. During the year 2014 a carbon tax of 5 US$/tCO2e was approved in Chile. This tax and its increases (10, 20, 30, 40 and 50 US$/tCO2e) are evaluated in this article. The results show that the effectiveness of this policy depends on some variables which are not controlled by policy makers, for example, non-conventional renewable energy investment cost projections, natural gas prices, and the feasibility of exploiting hydroelectric resources. For a carbon tax of 20 US$/tCO2e, the average annual emission reduction would be between 1.1 and 9.1 million tCO2e. However, the price of the electricity would increase between 8.3 and 9.6 US$/MWh. This price shock would decrease the annual GDP growth rate by a maximum amount of 0.13%. This article compares this energy policy with others such as the introduction of non-conventional renewable energy sources and a sectorial cap. The results show that the same global greenhouse gas (GHG) emission reduction can be obtained with these policies, but the impact on the electricity price and GDP are lower than that of the carbon tax.