Talks and presentations

Democratization, Elections, and Public Goods: The Evidence from Deforestation

April 20, 2018

Invited Talk, Working Group on African Political Economy (WGAPE) Annual Meeting, Los Angeles, CA

This paper offers a political explanation for increases in deforestation: competitive elections. The protection of forested areas provides long-term, public goods while their destruction provides short-term, private goods for local voters. Politicians facing a competitive election offer voters access to forested areas for commercial use of timber and small-scale farming in exchange for electoral support. I test this theory of political deforestation using satellite-verified global forest cover data and the results of over 500 national-level elections between 1975 and 2005. The findings suggest that the transition to democracy is associated with higher rates of deforestation, that election years may have slightly higher rates of deforestation than non-election years, and that close elections have 25\% higher deforestation rates than elections in which one side won by a wide margin. This suggests that democratization is associated with underprovision of environmental public goods and that contested elections are partially responsible for this underprovision.

Political Deforestation Cycles

December 12, 2017

Poster, American Geophysical Union Annual Meeting, New Orleans, LA

When do politicians’ re-election strategies cause serious environmental damage? This poster argues that for a politician interested in retaining office protecting forested areas is efficient in most years, but giving targeted access to pivotal voters is efficient in election years. I test the theory that competitive elections are associated with higher rates of deforestation using remote sensed satellite data of forest cover and data on national elections.

Detecting Land-Use Change and On-Farm Investments at the Plot Scale

December 10, 2017

Poster, American Geophysical Union Annual Meeting, New Orleans, LA

We use satellite images to measure the impact of a land-titling project on agricultural land use in Benin. Because control units are not delineated we generate synthetic controls for each treated plot based on pre-treatment match in the dependent variable (NDVI).

Geospatial Synthetic Controls for agricultural impact evaluation

October 20, 2017

Talk, CEGA Research Retreat, Berkeley, CA, Berkeley, California

We use satellite images to measure the impact of a land-titling project on agricultural land use in Benin. Because control units are not delineated we generate synthetic controls for each treated plot based on pre-treatment match in the dependent variable (NDVI).

Mass Digitization of Chinese Court Decisions

August 30, 2017

Conference proceedings talk, APSA Annual Conference, San Francisco, CA

We explore the potential uses for a massive new dataset of Chinese court cases, focusing on our efforts to estimate the length of time between the start of the case (inferred) to the decision. We use this to measure what types of cases take more or less time to be decided.