(Joint work with Juan Arroyo and Roberto Gómez, Banco de México.)
We use a novel data set on fi
rm vacancies and job seekers from a Mexican government job placement service to analyze whether changes in matching frictions can explain the large and persistent increase in Mexican unemployment after the 2008 global
financial crisis. Although we
find evidence of a signi
cant reduction in the efficiency of the matching function after 2008, the estimated effect can only explain about 70bp of the 233bp observed increase in the unemployment rate. Even with the decrease in efficiency, our measured market tightness implies that the recovery since 2010 should have been enough for the unemployment rate to recover to pre-crisis levels. We conclude that matching frictions cannot explain the abnormal persistence of unemployment.
(Joint work with Karen Kaiser and Alejandrina Salcedo, Banco de México.)
In 2007 banks in Mexico were mandated to offer basic deposit accounts for workers in the formal sector which are free of bank fees for some basic services and have standardized access across all banks. If deposit accounts provide access to other financial services, financial development in a country may deeply benefit from policy measures such as this one. Using a triple-differences strategy, we test whether the introduction of payroll basic accounts had a positive impact on the credit card holdings of the affected households. We exploit variation over time (before and after the reform) and compare those eligible for the basic accounts (formal workers) with those ineligible (informal workers) in municipalities with and without bank branches. We find evidence of a positive effect of the reform on credit card holdings.
(Banco de México Working Paper: 2011-09. For now, only available in Spanish.)
In 2010, charging ATM bank fees for users of the same bank was prohibited in Mexico. In this paper we build on the Hotelling model of spatial competition developed by Massoud and Bernhardt (2002) to analyze the consequences of the prohibition. We find that banning fees for account holders also reduces the fees charged on non-account holders but increases fixed fees. This latter increase is on average smaller than the decrease in the other two fees, which leads to an increase in consumer welfare. We also find that the prohibition decreases total surplus but that this decrease is absorbed by the banks' profits. The model does not consider the decision of banks to open or close down ATMs, which we leave for future research.
(First draft: December 2010. Revised February 2011 to add appendix with regressions for Senate. )
I find evidence that Political Action Committees act strategically when giving campaign contributions in the U.S. House of Representatives. They do so by targeting legislators with a higher probability of being pivotal for a vote. I find that increasing the pivotality of a legislator by one standard deviation increases her contributions by 39,000 dollars (10.8% of the average contributions). I also show that pivotality is not related to distance to the median voter when there is uncertainty on how legislators vote, although the distribution of pivot probabilities has as a similar spatial property: it's single-peaked. Finally, my estimations show that Republicans disproportionately gained pivotality after the Republican takeover of 1994 and continued to increase it while they were in control of the House.
(Banco de México Working Paper: 2010-16.)
This paper proposes a model of voting competitions (political campaigns and strategic lobbying) where voters are in°uenced by the opinion of their neighbors on a social network. In the unique pure strategy nash equilibrium, resources are targeted toward individuals with an influential position in the network. This finding contrasts with previous theories of strategic spending which predict that parties (or lobbies) should spend more on individuals who have a higher probability of being pivotal for the vote. The paper then tests the model using data on campaign contributions by interests groups in the US. House of Representatives. The estimations show that both network in°uence and pivotality are significant predictors of campaign contributions
“Price competition on a network.”
(Banco de México Working Paper: 2011-04.)
We present a model of imperfect price competition where not all firms can sell to all consumers. A network structure models the local interaction of firms and consumers. We find that aggregate surplus is maximized with a fully connected network, which corresponds to perfect competition, and decreases monotonically as the network becomes less connected until firms become local monopolists. When we study which networks are likely to form in equilibrium, we find that stable networks are not fully connected but are connected enough to rule out local monopolists. Our results extend to oligopolistic competition when consumers can either buy from a single firm or from all firms.
“The Leverage of Weak Ties: How Linking Groups Affects Inequality”
Joint work with Ben Golub. (Revised April 2010)
Network centrality measures based on eigenvectors are related to investment decisions, transmission of information, and local public goods provision. We study how the centrality of each member of a society changes when initially disconnected groups begin interacting with each other. We find that arbitrarily weak intergroup connections can have arbitrarily large effects on the distribution of centrality. For instance, if a high-centrality member of one group begins interacting with a low-centrality member of another, the latter's group will have a proportionally larger share of the total centrality. Our results suggest that eigenvector centralities should be treated with care in empirical work: the centrality across sparsely linked groups can be very sensitve to measurement error.
“Lorenz undominated points in a facility-location problem.”
(Revised March 2009)
An important dilemma in welfare economics is the trade-off between equality and surplus maximization, but this trade-off is not always present. Lorenz undominated points identify policies were there is no trade-off: either it’s no longer possible to improve aggregate surplus without increasing inequality, or to decrease inequality without decreasing aggregate total surplus. Unfortunately, these are hard to compute. We present an algorithm for finding Lorenz undominated points in an interesting social choice problem: where to place a desirable facility on an interval. In developing the algorithm we are able to highlight some interesting properties of Lorenz undominated points: they come from take averages from the worst-off individuals in society, they shrink under concave transformations and they form a convex set.