Matias Brum

Publications

Too little but not too late. Nowcasting poverty and cash transfers’ incidence during COVID-19’s crisis (With M. De Rosa) World Development (Volume 140, April 2021). 

Wage inequality on the rise: The role of workers’ characteristics (with G. Alves & M. Yapor). Journal of Income Distribution, vol. 22, n. 2, 2013 .

Income mobility and poverty traps: new evidence for Southern Cone countries (with R. Arim, A. Dean, M. Leites & G. Salas). Estudios Economicos, vol. 28, issue 1, 2013.

filler

Working Papers

Disentangling the incumbent’s advantage: Partisan and candidate separate effects for Uruguay (with R. Aguirre)

This paper separately estimates individual and partisan incumbency effects for regional elections in 1971-2020 for Uruguay. It contributes to a small but growing literature applying the novel method of Diff-in-Disc to solve endogeneity and bias problems. It exploits term limits and strong electoral enforcement. It yields a positive and significant individual incumbency effect of 56.3% (higher than the one reported in the only available precedent) and statistically non-significant partisan effects (in line with the literature.). This suggests that the candidate `owns’ his voters’ votes and sheds light on the intertwined relationships between parties, (potential) candidates and political parties. Results also show the importance and differential in size when using proper estimation tools vis-a-vis naive RDD methods. Results are robust to a battery of checks and extensions showing differential results before and after a constitutional reform that separated in time national and regional elections and limited the number of candidacies. We also explored two potential mechanisms behind the results. Overall, the paper shows that in Uruguay incumbency is a personal, individual matter and that when open elections take place, the new candidate of the incumbent party is not able to capitalize on his or her party’s past performance.

Do Dictatorships Affect People’s Long Term Beliefs and Preferences? An Empirical Assessment of the Latin American Case (SUBMITTED)

Does the political regime experienced during youth have long lasting effects on political beliefs and preferences? I exploit time and country variation in political regimes in Latin America using data from the 1995 to 2010 Latinobarometer and find that exposure to non-democratic regimes during youth reduces subsequent preference for democracy, satisfaction with democracy and confidence in institutions. These results suggest exposure to dictatorships during formative years permanently eroded democratic values. Exposure to non-democratic regimes also affects self-location in an ideology scale, reducing identification with the Right and increasing identification with the Left; which suggests dictatorships also shaped the political orientation of voters.

Italian Migration to the United States: The Role of Pioneers’ Locations (SUBMITTED)

Does the location and fortune of pioneering migrants have an effect on the size and concentration patterns of further migratory waves from the same origin communities? Filling a gap in the historical data, I focus on Italian mass migration to the US at the turn of the 20th century and combine scarcely used data with a surname matching technique to generate the first estimates of the yearly flow of migrants from each Italian municipality to each US county. I exploit variation across time, origins and destinations, and use instrumental variables based on settlement decisions of other non-italian immigrants to address the potentially endogenous location of Italian pioneers. I find that positive economic shocks in a county with high concentration of migrants from a given origin led to an increase in flows from that origin to that county and to the US as a whole. The former effect is smaller than the latter, which implies greater diversification of further migrants from the same origins across US counties.

Italian Migration to the United States: The Role of Migrant Networks (SUBMITTED)

This paper investigates the effect of network size on literacy levels of Italian migrants to the US at the turn of the twentieth century. I exploit unused ship manifest micro data and reconstruct migratory flows to the US by municipality of origin. This allows me to evaluate self-selection patterns at the sub-national level and by cohort, which shows positive selection for the lower tail of the literacy distribution. I use cross-sectional variation in the size of the migrant stock across municipalities of origin and use instrumental variables to asses measurement error. I find that increased migrant network size by municipality of origin is associated with a reduction in literacy for later migrants. Moreover, larger network size is also associated with lower immigrant age, and a higher proportion of women and agricultural workers in subsequent migratory flows. Results are consistent with migrant networks reducing migration costs and increasing emigration rates for lower quality (prospective) migrants.

esearch statement

Work in progress

Surfing the Latin American Pink Tide: Secular Trends or Sudden Change in Public Opinion? An Oaxaca-Blinder Approach (with P. Cazulo, G. Katz, H. Rueda, and S. Siniscalchi)

Employment effects of a social and labor inclusion program (with P. Blanchard, P. Carrasco, C. Parada, & I. Perazzo)

COVID-19 and the Informality Puzzle: The Case of Uruguay

Research statement