My research interests are primarily in the fields of development economics and applied econometrics, with secondary interests in applied microeconomics and economic education. My current research can be divided into three areas: (I) the measurement of and policy responses to underperforming service delivery, (II) poverty measurement, and (III) economic education as it pertains to students’ knowledge attainment and retention of economics.
I. The Measurement of and Policy Responses to Underperforming Service Delivery
There has been a surge of interest in the measurement of service delivery with the observation that, despite an increase in health and education funding, many countries health and education outcomes have stagnated and these countries will likely not be able to meet these MDGs. In “The Service Delivery Underperformance Index: A Multidimensional Approach to Measuring the Inadequacies in Service Delivery” a new approach to the measurement of service delivery is introduced. The Service Delivery Underperformance Index (SDUI) adapts the Alkire and Foster (2011) methodology used for poverty measurement to measure the underperformance, or multiple inadequacies, in service delivery. The index satisfies numerous properties. It focuses on underperformance, satisfies dimensional monotonicity, subgroup decomposability, and decomposability by dimensions and indicators. Dimensions and indicators that could be used to populate the index are discussed, referencing past work that has shown the negative impact of the poor delivery of services on health and education of individuals. It is demonstrated how the Alkire and Foster (2011) methodology is applied to calculate the index using facilities as the unit of analysis. Significance and robustness of the resulting rankings are discussed.
In “Healthcare Delivery in Rwanda, Uganda, and Tanzania: Applying the Service Delivery Underperformance Index to the Healthcare Sector” the SDUI is used to analyze healthcare delivery in Rwanda, Uganda and Tanzania using data from Demographic and Health Surveys Service Provision Assessment. A cross-country ranking of healthcare delivery is done for these countries, where it is found that Rwanda has the best performing healthcare delivery despite being ranked below Tanzania and Uganda in terms of some health outcomes. A more extensive analysis of healthcare delivery in Rwanda shows that there are significant disparities in the performance of different types of healthcare facilities. An analysis is done of healthcare delivery in facilities that did and did not participate in policies intended to improve healthcare delivery. It is found that facilities that participated in community involvement performed significantly better than facilities that did not. This observation calls for future work to be done using the SDUI as an impact evaluation tool to analyze how policies impact underperforming healthcare delivery.
In “The Delivery of Education Services in Papua New Guinea: Applying the Service Delivery Underperformance Index to the Education Sector of Papua New Guinea” the delivery of education services in Papua New Guinea is explored using the SDUI. Results from this analysis show that there are significant differences in the delivery of education services across different managerial types. It is also found that there are significant differences across managing authority and types of facilities depending on the province, which is expected considering that the education sector in PNG is highly decentralized. An evaluation is done of facilities in PNG that did or did not participate in multigrade classroom use and that did or did not have community involvement. These are two policies that the government was interested in expanding at the time of the survey. It is found that facilities that participated in either policy performed better than facilities that did not. This observation calls for future work to be done using the SDUI as an impact evaluation tool to analyze how policies impact underperforming education delivery. The impact of the poor delivery of education services, as measured by the SDUI, is explored using school level enrolment and national grade 8 test scores. It is found that schools classified as underperforming have significantly lower enrolment rates and lower test scores than schools not classified as underperforming. When analyzing results for male and female students, it is found that the impact is significantly larger for female students than male students.
My long-term research agenda is to expand the number of countries evaluated and do a comparison of these countries service delivery across time to further explore the relationship between service delivery and health and education outcomes. I also plan to use this index as an impact evaluation tool in both healthcare and education to establish the impact that policies such as decentralization have on service delivery. Many policies, such as decentralization, have had mixed results using the usual impact evaluation tools, likely because they evaluate the impact on indicators separately (without aggregation) and ignore the relationship between these indicators. Evaluation using the SDUI will give more definitive results on how the policy impacts overall service provision as well as specific indicators that the policy improves. There has also been interest in adapting the approach to other areas of service delivery such as public safety, water services, transportation, and natural resource management. I am currently working on adapting the approach for education to developed countries, with an application in the US.
II. Poverty Measurement
In a consulting position with the Poverty Reduction and Economic Management Group of the World Bank, I calculated poverty rates for the country of Lesotho from national survey data. I then presented these results to the Bureau of Statistics in Lesotho and taught them the techniques used to calculate poverty rates. This position allowed me to apply my knowledge of poverty measurement to a country specific case and expand my knowledge of the adaptation of poverty measures for individual countries. An important reason for involving the World Bank in poverty calculations within Lesotho was that these poverty calculations would be used in their proxy means testing for projects that target the poor in Lesotho. This inspired a paper to explore the benefits of using multidimensional targeting instead of proxy means testing.
In another consulting position with the Poverty, Gender, and Equity Group of the World Bank and the government of Brazil, I worked with members of the Chief Economists office in the Latin America and the Caribbean Region to quantify how much initial conditions affect poverty reduction given a level of growth across the states of Brazil, and therefore inhibit states from meeting the first Millennium Development Goal (reducing poverty by half). To do this we used a new framework to calculate counterfactual poverty reduction when all state are given the same initial conditions but still maintain their original growth rates. Results showed that, due to initial conditions it is more difficult for poorer states to achieve inclusive growth. After equalizing initial conditions poorer states are much more able to achieve high rates of poverty reduction given their original growth rates. This research was subsequently presented to the Government of Brazil.
The research from this position was then extended to a cross-country analysis in “The Unfairness of (Poverty) Targets”. This paper presented and used the same framework to calculate counterfactual poverty reduction when all countries are given the same initial conditions but still maintain their original growth rates. Results showed that, due to initial conditions it is more difficult for poorer states to achieve inclusive growth. After equalizing initial conditions poorer countries are much more able to achieve high rates of poverty reduction given their original growth rates. This results holds for different measures of poverty as well as different poverty lines. This research was subsequently presented to the Government of Brazil.
The application of this framework goes beyond poverty benchmarking, as it is also common in the health and education sectors to set benchmarks on outcomes. My future research agenda is to use this framework to measure policy performance towards health and education outcomes.
III. Economic Education as it Pertains to Students’ Knowledge Attainment and Retention of Economics
Anecdotal evidence from Principles of Economics faculty suggests that many students fail to comprehend foundational material in economics due to gaps in their understanding of basic Algebra I concepts. To address this issue, Principles faculty at George Washington University administer a common Algebra I assessment at the start of each fall semester. “The Role of Algebra I Assessment in Improving Student Performance in Economics” presents results from data collected on 1361 students registered for a Principles of Economics class for which the prerequisite is Algebra I. Our results show that student performance on the Algebra I assessment is a good predictor of cumulative final exam scores in the principles class; however SAT Math scores cannot be used as an equally good predictor. We suggest that use of a calculator and/or multiple-choice format on the SAT Math test are masking students’ weaknesses in Algebra I.
“Does Calculator Use and Test Format Mask Weakness in Basic Math Ability? Experimental Evidence of Impact on Comprehension in Principles of Economics” further explores whether test format and/or calculator use were driving the difference in explanatory power of the SAT Math and Algebra I assessment on students’ performance in Principles of Economics using an experiment in the Fall 2013 semester. In this experiment 971 incoming students at George Washington University in a Principles of Economic course are reported. In this experiment students were given an Algebra I assessment (to ensure they had the necessary Algebra I skills to take a Principles of Economics course), and immediately following were randomly allocated to a treatment or control group conditional on SAT math scores to test if there was a significant impact of test format (multiple choice vs. open ended), calculator use and type, and the interaction of calculator use/ type and test format on students scores on the assessment. The results from this experiment do suggest that each treatment had a significant impact on students’ scores, with much variation depending on the type of question asked.
Abstract:
A new approach to the measurement of service delivery is introduced. The Service Delivery Underperformance Index (SDUI) adapts the Alkire and Foster (2011) methodology used for poverty measurement to measure the underperformance, or multiple inadequacies, in service delivery. The index satisfies numerous properties. It focuses on underperformance, satisfies dimensional monotonicity, subgroup decomposability, and decomposability by dimensions and indicators. Dimensions and indicators that could be used to populate the index are discussed, referencing past work that has shown the negative impact of the poor delivery of services on health and education of individuals. It is demonstrated how the Alkire and Foster (2011) methodology is applied to calculate the index using facilities as the unit of analysis. Significance and robustness of the resulting rankings are discussed.
Abstract:
The Service Delivery Underperformance Index (SDUI) is used to analyze healthcare delivery in Rwanda, Uganda and Tanzania using data from Demographic and Health Surveys Service Provision Assessment. A cross-country ranking of healthcare delivery is done for these countries, where it is found that Rwanda has the best performing healthcare delivery despite being ranked below Tanzania and Uganda in terms of some health outcomes. A more extensive analysis of healthcare delivery in Rwanda shows that there are significant disparities in the performance of different types of healthcare facilities. An analysis is done of healthcare delivery in facilities that did and did not participate in policies intended to improve healthcare delivery. It is found that facilities that participated in community involvement performed significantly better than facilities that did not. This observation calls for future work to be done using the SDUI as an impact evaluation tool to analyze how policies impact underperforming healthcare delivery.
Abstract:
Adopted on September 8 of 2000, the United Nations Millennium Declaration stated as its first goal that countries “...[further] resolve to halve, by the year 2015, the proportion of the world’s people whose income is less than one dollar a day and the proportion of people who suffer from hunger...” Each country committed to achieve the stated goal, regardless of their initial conditions in terms of poverty and inequality levels. This paper presents a framework to quantify how much initial conditions affect poverty reduction, given a level of “effort” (growth). The framework used in the analysis allows for the growth elasticity of poverty to vary according to changes in the income distribution along the dynamic path of growth and redistribution, unlike previous examples in the literature where this is assumed to be constant. While wealthier countries did perform better in reducing poverty in the last decade and the half (1995-2008), assuming equal initial conditions, the situation reverses: we find a statistically significant negative relation between initial average income and poverty reduction performance, with the poorest countries in the sample going from the worse to the best performers in poverty reduction. The analysis also quantifies how much poorer countries would have scored better, had they had the same level of initial average income as wealthier countries. The results suggest a remarkable change in poverty reduction performance, in addition to the reversal of ranks from worse to best performers. The application of this framework goes beyond poverty targets and the Millennium Development Goals. Given the widespread use of targets to determine resource allocation, in education, health, or decentralized social expenditures, it constitutes a helpful tool to measure policy performance towards all kinds of goals. The proposed framework can be useful to evaluate the importance of initial conditions on outcomes, for a wide array of policies.
Abstract:
Anecdotal evidence from Principles of Economics faculty suggests that many students fail to comprehend foundational material in economics due to gaps in their understanding of basic Algebra I concepts. To address this issue, Principles faculty at George Washington University administer a common Algebra I assessment at the start of each fall semester. This paper presents results from primary data collected on 1361 students registered for a Principles of Economics class for which the prerequisite is Algebra I. The results show that student performance on the Algebra I assessment is a good predictor of cumulative final exam scores in the Principles class. However, SAT Math scores cannot be used as an equally good predictor. We suggest that use of a calculator and/or multiple-choice format on the SAT Math test are masking students weaknesses in Algebra I.
Abstract:
Results from an experiment in Fall 2013 of 971 incoming students at George Washington University in a Principles of Economic course are reported. In this experiment students were given an Algebra I assessment (to ensure they had the necessary Algebra I skills to take a Principles of Economics course), and immediately following were randomly allocated to a treatment or control group conditional on SAT math scores to test if there was a significant impact of test format (multiple choice vs. open ended), calculator use and type, and the interaction of calculator use/ type and test format on students scores on the assessment. The results from this experiment do suggest that each treatment had a significant impact on students’ scores, with much variation depending on the type of question asked.
