Abstract:  We study the design of income-based financial aid in Chile’ s college-admissions system. Chile’s Gratuidad program, with its threshold design (full tuition subsidy for students below the 60th income percentile and none above), creates an affordability gap around the threshold and a significant enrollment distortion. We build a novel college-admissions model with tuition-charging colleges and students heterogeneous in affordability. Developing an axiomatic framework,we characterize a class of feasible aid policies which addresses this distortion. A redistributiveness relation over aid policies establishes a conceptual link between this class and the current policy and yields a unique policy that eliminates the distortion while preserving Gratuidad’s redistributive objectives. The results are implementation-friendly: they are robust to misspecification of behavioral responses, the slope is pinned downby the highest tuition in the market, and the policymaker’ s decision reduces to choosing a single parameter to adjust the cost of policy.

Abstract: An agency needs to hire a group of individuals as diplomats for a predetermined set of missions, each lasting for a certain duration. Diplomats are reassigned to new missions periodically throughout their career. Missions may require different qualifications and vary in desirability among the diplomats. This paper investigates the design of recruitment procedures that results in a suitable pool of diplomats for the hiring entity to be able to assign an adequately qualified diplomat to each mission while maintaining balance in mission assignments in terms of their desirability. By introducing a new Hall’s type theorem for hypergraphs, we characterize such recruitment procedures. In addition, we show a particular relation structure within the characterized set which constitutes an isomorphism with the merit comparison of the diplomat sets selected under different recruitment procedures.

Abstract: Motivated by France’s public-school teacher (re)assignment system, we study how to (re)assign workers to positions in a dynamic market with entry of new workers and exit of retirees. When worker preferences are correlated, unguided reassignment produces persistent regional segregation in worker quality. We formalize distributional targets at each school and introduce a feasibility requirement that every reassignment weakly improves each school’s type distribution relative to its target while preserving individual rationality (IR). We design choice rules that implement these targets via ceilings/floors and analyze a generalized deferred-acceptance mechanism. We show the intrinsic tension between feasibility and standard stability, and propose a weaker fairness notion (allocational stability) that is compatible with IR and the distributional constraints. The resulting mechanism yields step-by-step improvements toward the policymaker’s distributional objectives without harming currently employed workers and offers a practical path to reducing long-run segregation. 

Abstract:  Amid ongoing debates on affirmative action, we provide a clean distinction between opportunity-enhancing policies and affirmative-action policies in markets where individuals compete for seats (college admissions and labor). In a large-market environment, individuals choose effort endogenously as a function of policy. By exploiting the policy-induced changes in equilibrium effort, we locate the boundary between equality of opportunity and affirmative action, and express this boundary in terms of redistributive objectives. 

Abstract: This paper studies the design of equity-oriented, income-based financial aid in centralized college admissions, using Chile as the motivating case. I build a large-market model in which students are assigned through student-optimal deferred acceptance and tuition enters preferences via an income-dependent value of money. Aid is modeled as a schedule that scales with posted tuition, so changes in the schedule shift students’ true rankings over colleges before the match is run. I introduce the notion of hidden envy: a student experiences hidden envy for a college when she would strictly prefer that college to her assignment and her score would have cleared its cutoff, but net tuition prevents her from ranking or reaching it. This concept isolates barriers created by prices rather than by priorities or strategy. 

Using multidimensional Leibniz rule defined for polytopes, I characterize stable outcomes through market cutoffs and derive comparative statics for any marginal change to the aid schedule. The results separate a direct effect—beneficiaries’ preferences move toward higher-tuition options—from a matching effect—seat cutoffs adjust in general equilibrium, which can help some groups while shifting congestion to others. I also lay out a transparent budget mapping that highlights nonlinear spending across incomes. Together, these tools provide implementable diagnostics and policy trade-offs, including designs that reduce overall hidden envy or protect the worst-off group, while remaining compatible with centralized assignment.

Abstract: We study a parametric family of school choice mechanisms that bridges the Deferred Acceptance (DA) mechanism and the Boston mechanism (BM). Our Generalized Secure Boston mechanism introduces a vector parameter that grants each school a controlled ability to “honor love” (top-choice claims) while still respecting priorities. At one end of the parameter space the mechanism coincides with DA; at the other, with BM. This unification lets policymakers tune the trade-off between priority fairness/stability and welfare (as captured by ranks). We show monotone comparative statics in the parameter: higher paraameter weakly reduces manipulability and priority violations, while lower parameter improves rank efficiency. Using simulated environments with varying correlation in preferences and priorities, we evaluate a loss that combines welfare and fairness (e.g., average rank plus a penalty for blocking pairs) and find that the optimal parameter depends systematically on the joint distribution of preferences and priorities—providing a practical, data-driven calibration of "love", with a guaranteed minimum of "being loved".