Embedded Files
Job Market Paper
Job Market Paper
Power Distribution and Efficiency: Theory and Evidence from Unionized Firms
Power Distribution and Efficiency: Theory and Evidence from Unionized Firms
Abstract: Power is irrelevant for efficiency in a complete contract world, but when some actions are noncontractible it can affect both the size and distribution of surplus. This paper studies groups with little formal power that engage in coercion—noncontractible actions that shift the bargaining disagreement point in their favor while destroying surplus—and shows that transferring bargaining power toward them can crowd out coercion and enhance efficiency. I investigate this mechanism in union–management disputes, providing early causal evidence on power's efficiency effects in incomplete-contract settings. Using administrative data, including population-level tax records from a 2017 reform of the Chilean labor code, I run an event-study design exploiting institutions that fix the timing of collective bargaining to obtain quasi-random assignment to the new versus the old regime across firms. The reform leaves profits statistically unchanged while increasing productivity and workers' remuneration without reducing employment, and reduces measures of conflict and unionization. Efficiency gains are concentrated where unions were initially weak, while wage and unionization responses are larger where unions start stronger—a nonlinear response to bargaining power. A bargaining model with coercion rationalizes these findings, characterizes surplus-maximizing power distributions, and shows that power can become a burden: management's welfare can fall as its own bargaining power rises when reduced coercion expands surplus. A calibration implies the union's surplus share rises from 0.39 to 0.48 without harming profits. More broadly, the results suggest that equality and efficiency need not trade off when contracts are incomplete.
Working Papers
Working Papers
Partisan Universities and Political Donations
Partisan Universities and Political Donations
Under Review. (with Franco Calle)
Abstract: Does the ideological orientation of universities shape their students' partisan engagement? We link the universe of Chilean college applications to the universe of political donations (2004-2021) and exploit score-based discontinuities in the centralized admission system to estimate the causal effect of enrolling in universities with different ideological leanings. Enrolling in a left-leaning university raises the probability of donating to left-wing campaigns by 0.41 percentage points, 40 percent of the baseline donation rate, with effects three times as large at public left-leaning universities and amplified by electoral exposure during college. The effects are asymmetric: private right-leaning universities do not mobilize right-wing donors but instead reduce donations to the left, while by field of study only business and economics majors increase right-wing donations, by 0.78 percentage points. Effects are strongest for the earliest cohorts, indicating durable rather than transitory exposure. The pattern points to a public, collectivist institutional culture on the left and to market-oriented curricula and professional networks on the right.
Divide and Conquer: Bargaining Scope and the Distribution of Rents
Divide and Conquer: Bargaining Scope and the Distribution of Rents
(with José M. Pascual)
Abstract: This paper studies how legal firm boundaries that constrain the scope of collective bargaining affect wages, rent sharing, and productivity. We analyze a 2014 Chilean labor-code reform that prevented corporations from fragmenting their workforce into smaller legal entities, thereby expanding the potential size of bargaining units. Using population-wide matched employer–employee administrative data and a dynamic difference-in-differences design with same-industry controls, we find that average annual earnings rose by 4.7%, bottom-decile wages by 10%, and within-firm wage dispersion (standard deviation) by $520–$710 (USD), corresponding to 6–9% of its pre-reform level. Employment and firm profits did not change, while value added per worker increased. Effects are strongest in initially smaller businesses and in multi-establishment groups with co-located units, consistent with coordination-cost frictions. We develop an extensive-form model in which management chooses whether to fragment or consolidate its business, and payoffs are determined by subsequent bargaining with workers. Under consolidation, workers can form an encompassing union that worsens management’s outside option during a strike, while the firm internalizes cross-unit complementarities in production. The model predicts higher worker effort, wages, and wage dispersion under consolidation, matching our empirical results and rationalizing their heterogeneity. Overall, the analysis shows that expanding bargaining-unit scope can be efficiency-consistent: coordination gains offset rent reallocation, allowing efficiency and equity to coexist within firms.
Work in Progress
Work in Progress
(Un)Ethical Allocations in Democratic Competition: A Dynamic Probabilistic Voting Model
(Un)Ethical Allocations in Democratic Competition: A Dynamic Probabilistic Voting Model
Abstract: Ethical preferences significantly influence individuals' judgments on political actions and policies, and deviations from perceived ethical standards lead to political dissatisfaction. This paper explores the conditions under which democratic competition fails to implement ethical allocations, using a dynamic probabilistic voting model to understand how candidates learn about citizens’ ethical preferences. This study characterizes a set of ethically consistent redistribution policies (allocations), defined as equilibrium outcomes implementable by a social welfare function —ranging from Rawlsian to utilitarian— derived from normative axioms consistent with a particular ethical framework. Through a communication game, continuous learning by politicians leads to a steady-state convergence, identifying society’s true preference profile and implementing a utilitarian optimum (as in classic probabilistic voting models). However, a sequence of uninformative election outcomes, termed as 'uninformative traps', can obstruct this learning process, leading to potentially unethical allocations. The paper investigates the role of communication frictions, like voluntary voting and authoritarian regimes, in impeding the learning process, while highlighting how high-quality institutions can facilitate accurate learning and ethically consistent allocations. The findings underline the importance of informed electoral processes and institutional quality in promoting ethically consistent policies, contributing to satisfaction with democratic outcomes.
Non-Consequentialist Social Choice
Non-Consequentialist Social Choice
[Draft coming soon] [Link to preliminary slides]
Abstract: This paper develops a formal framework for collective decision-making in societies where individuals may hold non-consequentialist moral views. Standard social choice theory and welfare economics typically evaluate outcomes solely in terms of the welfare they produce. In contrast, many individuals care about both the moral quality of outcomes and the procedures by which they are produced. I extend the Arrow–Sen framework to incorporate such ethical pluralism by modeling a two-stage process: a meta-level selection of institutional rules and an operational-level implementation of social choices. I introduce two axioms, Outcome Non-Consequentialism (ONC) and Procedural Non-Consequentialism (PNC), to capture moral commitments beyond consequences. I prove that any social choice function (SCF) or meta-social choice function (MSCF) satisfying either ONC or PNC must violate weak Pareto efficiency and neutrality. I also introduce a Consistency Axiom requiring alignment between meta- and operational-level decisions, and show that ethical disagreement across agents leads to the impossibility of coherent institutional design. The MSCF framework can accommodate any mechanism or aggregation procedure that implements social states, including game-form rights (Nozick), evolutionary bargaining (Binmore), and other abstract institutional designs (Rawls). The results highlight the normative and institutional limits of deontological approaches and the structural tension between procedural legitimacy and social efficiency.
Publications (Pre - Ph.D.)
Publications (Pre - Ph.D.)
Earned Income Tax Credit: a public policy against inequality
with Andrés Hernando
Estudios Publicos, 146: 49-85, May 2017.
Earned Income Tax Credit: a public policy against inequality
with Andrés Hernando
Estudios Publicos, 146: 49-85, May 2017.
Abstract: This article presents the earned income tax credit (EITC), an instrument that consists of topping up lower-paid workers’ earnings with cash transfers. As a social policy, it has the advantage of being an effective mechanism for tackling inequality that does not greatly interfere with the trade-off between leisure and work but encourages labor market participation and formalization, cuts poverty, and reduces the stigma attached to being a social program beneficiary. This instrument is offered for two purposes: (i) as an actual policy option and (ii) as a benchmark for contrasting the potential cost-effectiveness of any other public policy intended to reduce inequality. The paper applies a genetic algorithm for numerical optimization to evaluate the parameters that optimally minimize income inequality and shows that, for example, spending 5,000 million dollars a year would reduce the Gini by between 4.7 and 6.1 points, which is more than the entire reduction recorded in Chile between 1990 and 2015.
An Inequality Trap: the consequences of assortative matching on income distribution
Estudios Publicos, 144: 61-107, October 2016.
An Inequality Trap: the consequences of assortative matching on income distribution
Estudios Publicos, 144: 61-107, October 2016.
[Link to paper — In Spanish] [Media Coverage 1] [Media Coverage 2]
Abstract: This paper looks at the effects of assortative mating on education for variables such as inequality, income, and the level of education in society. For Chile, it simulates a scenario where the parents of this generation would have mated randomly, without regard to their mate’s level of education. It demonstrates that in that hypothetical scenario, the inequality of household income, measured using the Gini coefficient, would fall from 0.48 to 0.43, a drop equivalent to the reduction in inequality between 1990 and 2013 (for this same variable). On average, income would simultaneously fall 15 percent, the uneducated would decrease by 5 percent, and the college graduates would also drop 5 percent.
My research has been generously supported by the following institutions:
Page updated
Google Sites
Report abuse