Abstract: I estimate the causal effects of annuitization on old-age outcomes. Annuities insure retirees against outliving their savings by providing income for life, but there is no causal evidence on the consequences of retiring with an annuity rather than receiving comparable pension wealth as a lump sum. I exploit a rule in Colombia's private pension system that assigns a life annuity to retirees whose savings exceed an individual-specific threshold and a lump sum to those below it. I combine a fuzzy regression discontinuity design with administrative records on the universe of retirees from 2008 to 2023 linked to death records and the social-assistance registry, and estimate that annuitization adds 1.3 years of life in the observed period, and reduces reliance on means-tested social assistance among retirees observed more than a decade after retirement. The timing of these effects is consistent with gradual depletion of lump sum savings and shows that pension payout design can have lasting consequences for longevity and reliance on the social safety net.

Conference presentations: NTA (2025), APPAM (2025), RIDGE-LACEA (2026), CEPR Economics of Longevity (2026),  3rd IDB Research Conference (2026), CEGA Research Retreat (2026), LACEA LAMES (2026)

Abstract: We study how local economies in the United States respond to population aging. Using variation in the share of residents above retirement age across commuting zones, we document a set of empirical patterns: aging is associated with shifts in population, employment, wages, and the mobility of both younger and older cohorts. These correlations are puzzling in the context of a standard spatial equilibrium model in which retirees exit the labor force but continue to demand local goods and services. We show that incorporating housing markets and local amenities into the framework helps reconcile the evidence. Consistent with this extended model, we find that aging is strongly correlated with lower housing supply, stricter local land-use regulation, and more conservative political values. Together, these results highlight the role of housing and regulation in mediating the local economic consequences of demographic change.

Abstract: Most pension systems establish specific age thresholds that entail changes in the financial incentives to retire. These key retirement ages may themselves become reference points that influence the decision to retire beyond the effect of purely financial incentives. A growing literature has started to call attention to the relevance of these labeled key retirement ages. It is still unclear, however, how large the impact of these reference points is relative to underlying changes in financial incentives. Addressing this question is challenging since it is hard to find sources of variation that separately identify the effect of labelled retirement ages from the effect of financial incentives, as these usually vary together. This paper leverages the unique structure of the Colombian private pension system to address this challenge. I exploit the fact that workers in a specific pension pathway respond to labeled retirement ages even when these are not accompanied by financial incentives. I compare this response to those of workers in pension pathways where retirement ages and financial incentives vary together, enabling an estimation of the relative magnitudes of these effects. The findings provide new evidence on the extent to which labeled retirement ages independently shape retirement behavior. 

Abstract: Extending health coverage to excluded groups may change the health of everyone else who shares the same hospitals and budgets. In 2022, Colombia granted nearly two million Venezuelan migrants access to full health services. We explore the effects of this health care expansion on Colombian mortality rates, comparing municipalities more and less exposed to the program before and after the program's mass rollout. In a program that covered nearly four percent of the population, we find that a one standard deviation increase in municipal exposure to the program, representing about 30 newly covered migrants per 1,000 inhabitants, reduced mortality among Colombians by 1.8 percent, with declines of 2.2 percent in infant mortality and 2.7 percent among adults 65 and older. These effects are driven by congestion relief. Newly insured migrants entered financed routine care at scale, tilting the composition of their care sharply away from emergency rooms, while the emergency share of Colombians' own care fell and their routine care did not contract. Because emergency care was already publicly financed for undocumented migrants, full coverage increased net public spending only modestly. As a result, the net fiscal cost per Colombian life saved is USD \$58,000 under our preferred scenario, well below conventional benchmarks based on the value of a statistical life. These are benefits for an already insured host population, a channel that is generally omitted from standard fiscal estimates of expanding health coverage to excluded populations.

Abstract: We examine how forced displacement shapes child development beyond socioeconomic disadvantage. Using original data on roughly 1,600 Venezuelan refugees, Colombian internally displaced, and non-displaced children in Medellín, Colombia, we adapt rank-rank regressions from the intergenerational mobility literature to a setting of forced displacement, ranking each child's developmental outcomes against household wealth. We document substantial gaps that persist after holding wealth and parental background fixed, and we show that the cause of displacement shapes their nature. Venezuelan refugees, displaced by acute material deprivation, fall 19 and 13 rank points behind in cognitive and physical development; Colombian children displaced by armed conflict instead fall 12 rank points behind in depression and exhibit elevated trauma. Access to regularization and social services is associated with narrower gaps.

Abstract: This article investigates the effects of female immigration to the Dominican Republic (DR)—most of which is from Haiti and of low-education levels—on the labor supply of native women. Using individual-level data for 2003–2016 and exploiting geographic variation in early immigrant settlements together with time variation in female immigration inflows, we find that female immigration has led to disparate outcomes across women of different education levels and family structures. In line with the evidence from developed countries, female immigration to the DR is associated with an increase in the hours worked by highly educated native women with family dependents (relative to equally educated women without dependents). However, for low-educated native women, female immigration is associated with a decrease in hours worked and in earnings. Our results underscore the importance of studying the disparate effects of migration on vulnerable groups in developing countries

Abstract: We examine patterns of economic growth and social progress across Colombian “functional territories.” Unlike political or administrative divisions, which often arise for historical reasons unrelated to economic interactions, functional territories reflect the underlying patterns of spatial agglomeration and economic linkages within a region. Using a novel definition of functional territories, our analysis reveals substantial fragmentation: nearly 66% of municipalities—home to about 20% of the national population—have no significant economic ties to neighboring areas. A comparatively more integrated set of municipalities, though still only partially connected, forms a “rural–urban” space that accounts for roughly 31% of the population. The remainder of Colombians reside in “urban” or “metropolitan” clusters that are larger, more densely populated, and more interconnected. We analyze these territories along two dimensions: economic growth, or “dynamism,” and social progress, or “inclusion.” To guide this analysis, we propose a simple conceptual framework that organizes the diverse inputs shaping these outcomes. Larger and more urbanized agglomerations show clear advantages in these inputs, and long-run institutional factors provide the best explanation for differences across territories. Consistent with this, larger and more urbanized areas achieve stronger outcomes, particularly in economic activity. Moreover, dynamic territories tend also to be more inclusive, although recent gains in dynamism are not strongly correlated with improvements in inclusion.

Abstract: The delimitation of functional spatial units or functional territories is an important topic in regional science and economic geography, since the empirical verification of many causal relationships is affected by the size and shape of these areas. This paper proposes a two-step method for the delimitation of functional territories and presents an application for three developing countries: Mexico, Colombia and Chile. The first step of this method uses nighttime satellite images to identify the boundaries of urban continuums (conurbations). When these continuums extend over more than one municipality, we group and redefine them as a new single spatial unit. The second step calculates a dissimilarity index using bidirectional labor-commuting flows between the resulting areas of the first step and then applies a standard clustering procedure to delineate the definitive functional territories. Our results suggest that, using nighttime satellite images, our method can lead to a more accurate definition of functional territories, especially in developing or underdeveloped countries where the official data on labor-commuting flows are often outdated or unreliable.