Receiving my PhD in political science from Emory University, I have training and experience in causal inference and experimental design, survey and measurement methodologies, statistical analysis and predictive modeling, as well as data management and visualization in the programming language R.
Projects
Democratic Space Barometer
Funded by IRI and USAID, this Varieties of Democracy (V-Dem) Institute project estimates the probability that a country will experience at least one opening event (shifts towards more democratic governance) or at least one closing event (shifts towards more autocratic governance) over a two-year window (2019-2020) for six dimensions of the Democratic Space: the formal and informal institutional environment of a polity in a given year.
While acknowledging that the Democratic Space captures a number of formal and informal institutions, the Democratic Space Barometer focuses on six institutional spaces that can have an out-sized influence on the nature and character of a country's regime: associational, economic, electoral, governing, individual, and informational spaces.
We use a set of 12 random forest forecasting models with a vast country-year dataset with global coverage from 1970 to 2019 to estimate the probability that there will be an opening or closing event in a country within the next two years (2020-2021).
By focusing on both opening and closing events we hope to help policymakers identify countries at a potential crossroad in their regime trajectory. The goal is to provide policymakers more concrete abilities to identify potential opportunities and entry-points for program implementation to help counter autocratization efforts and bolster democratization efforts. To these ends, we will be launching an interactive dashboard with visualization tools in March 2020, when we have our formal public release.
While acknowledging that the Democratic Space captures a number of formal and informal institutions, the Democratic Space Barometer focuses on six institutional spaces that can have an out-sized influence on the nature and character of a country's regime: associational, economic, electoral, governing, individual, and informational spaces.
We use a set of 12 random forest forecasting models with a vast country-year dataset with global coverage from 1970 to 2019 to estimate the probability that there will be an opening or closing event in a country within the next two years (2020-2021).
By focusing on both opening and closing events we hope to help policymakers identify countries at a potential crossroad in their regime trajectory. The goal is to provide policymakers more concrete abilities to identify potential opportunities and entry-points for program implementation to help counter autocratization efforts and bolster democratization efforts. To these ends, we will be launching an interactive dashboard with visualization tools in March 2020, when we have our formal public release.
Experiment & Survey experience: The Dissertation
As a US Army Psychological Operations soldier in Afghanistan, one of my tasks was to assess how US military and development operations might influence the support preferences of the local population. US military doctrine suggests that by improving the economic conditions and security situation of an area, the population will be more willing to provide vital information to counter-insurgent forces.
Arguing that an extensive road network would improve the economic conditions of an area as well as the government's ability to patrol through an area and secure the population, the US military saw road development as an important, dual use counterinsurgency tool. My dissertation tests the assumption that an increase in armed presence patrols (think cops patrolling neighborhoods in squad cars) will lead to an increase in popular support for the government. I don't think this was quite correct.
Instead, I argue that when ethnicity is salient to an insurgency, an increased presence of government forces can lead to a decrease in support for the government. However, when ethnicity is less salient, more government forces can increase support. A government-funded rural road development plan and on-going ethnic and ideological insurgencies in India provided an opportunity to test this relationship.
My dissertation focuses on rural villages within two conflict areas in India: the districts of Assam where the National Democratic Front of Bodoland, an insurgency where ethnicity is salient, is active, and the districts in southern Bihar where the Communist Party of India -- Maoist insurgency, where ethnicity is less salient, is active.
I leveraged the allocation procedures of government of India's rural roads development scheme, Pradhan Mantri Gram Sadak Yojana (PMGSY), in a survey-based research design that allowed me to assess (1) whether the expansion of a rural road network increases the probability that counter-insurgents patrolled through an area and (2) how an increased presence of counter-insurgents in an area affects the willingness of the population to express support for the government.
In brief, I drew a random sample of villages with populations just below (230 to 249) and just above (250 to 270) the 250-person population threshold a village had to surpass to receive a PMGSY road. Similar to a regression-discontinuity design, this allowed me to survey villages that did and did not receive a road but had a similar probability of being awarded one.
To measure the presence of and popular support for the government in these villages, I developed a household survey and a village leader semi-structured interview template. With the generous help of a small Indian market research firm (and NSF funding), we traveled to these areas (Assam from April to June 2015 and Bihar from June to August 2016) and recruited and trained local enumerators, who then helped us refine, test, and field our survey. Using a number of survey experiments, the survey measured the frequency of patrols in villages and government support.
Data from these village-level surveys provide support for my theory. Road development increases the probability that counter-insurgents will patrol through an area, and an increase in police presence in ethnic conflicts tends to decrease support for the government. Other research suggests that this latter relationship is mitigated when the ethnic make up of the security forces is similar to the area in which they operate.
Arguing that an extensive road network would improve the economic conditions of an area as well as the government's ability to patrol through an area and secure the population, the US military saw road development as an important, dual use counterinsurgency tool. My dissertation tests the assumption that an increase in armed presence patrols (think cops patrolling neighborhoods in squad cars) will lead to an increase in popular support for the government. I don't think this was quite correct.
Instead, I argue that when ethnicity is salient to an insurgency, an increased presence of government forces can lead to a decrease in support for the government. However, when ethnicity is less salient, more government forces can increase support. A government-funded rural road development plan and on-going ethnic and ideological insurgencies in India provided an opportunity to test this relationship.
My dissertation focuses on rural villages within two conflict areas in India: the districts of Assam where the National Democratic Front of Bodoland, an insurgency where ethnicity is salient, is active, and the districts in southern Bihar where the Communist Party of India -- Maoist insurgency, where ethnicity is less salient, is active.
I leveraged the allocation procedures of government of India's rural roads development scheme, Pradhan Mantri Gram Sadak Yojana (PMGSY), in a survey-based research design that allowed me to assess (1) whether the expansion of a rural road network increases the probability that counter-insurgents patrolled through an area and (2) how an increased presence of counter-insurgents in an area affects the willingness of the population to express support for the government.
In brief, I drew a random sample of villages with populations just below (230 to 249) and just above (250 to 270) the 250-person population threshold a village had to surpass to receive a PMGSY road. Similar to a regression-discontinuity design, this allowed me to survey villages that did and did not receive a road but had a similar probability of being awarded one.
To measure the presence of and popular support for the government in these villages, I developed a household survey and a village leader semi-structured interview template. With the generous help of a small Indian market research firm (and NSF funding), we traveled to these areas (Assam from April to June 2015 and Bihar from June to August 2016) and recruited and trained local enumerators, who then helped us refine, test, and field our survey. Using a number of survey experiments, the survey measured the frequency of patrols in villages and government support.
Data from these village-level surveys provide support for my theory. Road development increases the probability that counter-insurgents will patrol through an area, and an increase in police presence in ethnic conflicts tends to decrease support for the government. Other research suggests that this latter relationship is mitigated when the ethnic make up of the security forces is similar to the area in which they operate.
Last updated: 07/21/2020
