919-966-1733 Carolina Population Center
Boshamer Distinguished Professor
David Guilkey is an applied econometrician with a microeconomics focus. He has been on the faculty of the University of North Carolina at Chapel Hill since 1977 and a Full Professor since 1983. He has been a Fellow of the Carolina Population Center since 1982 and was the Chair of the Economics Department for the five year period 1995-2000. He is also the current Senior Technical Advisor to the MEASURE Evaluation Project, a large USAID funded project that is designed to develop improved methods for use in the monitoring and evaluation of USAID's world wide population and health projects. A summary of his current research interests follows.
Survey Data Sets
Much of Guilkey's research involves the use of large survey data sets -- either cross sectional or longitudinal. The main focus of his interest is the development of estimation methods that can be used to analyze large survey data sets with limited dependent variables especially when endogenous right-hand-side variables are present. In collaboration with Tom Mroz, he has developed methods to estimate structural equations models with discrete dependent variables that rely on fewer assumptions that standard methods. They have also studied the sensitivity of parametric estimators for these types of models to the level of identification of the structural model. With Ken Bollen, they have made many of these methods accessible to demographers by demonstrating their use in the evaluation of the impact of Tunisia's family planning program on fertility intentions and contraceptive use.
Health and Nutrition
Guilkey has published over 40 articles dealing with health and nutrition using both domestic and international data sets. Most of his work with US data sets has involved the analysis of USDA data. The questions of interest have included determination of how individuals choose to participate in Federal programs such as the Food Stamp program and the school lunch program and an evaluation of the impact of the program participation on the individual's nutritional status. His international work has concentrated on the health related issues such as the determinants and consequences of health care usage and the determinants of breast-feeding by mothers. Along with Barry Popkin and John Akin, he helped initiate the Cebu Longitudinal Health and Nutrition Survey that followed 3,000 mother and infant pairs from birth through adolescence. With various coauthors, he has examined a wide range of substantive questions using this data set. Topics include the effects of infant formula company's marketing activities on the initiation and duration of breast feeding, the effects of breast feeding and other inputs into child health on child morbidity and mortality, and the effects of feeding patterns on child stunting.
Guilkey has also used the Cebu data to examine the effects of contraceptive use and breast feeding on birth spacing. Recent work involves collaborative work with Susan Cochrane of the World Bank to evaluate the impact of family planning programs on contraceptive method choice in Tunisia, Zimbabwe and Columbia using DHS data. His current work involves the estimation for reduced form equations that evaluate the impact of family planning programs on fertility intentions, contraceptive use, and fertility in 13 countries. He has also worked closely with the USAID Mission in Tanzania to evaluate the impact of the Family Planning Support Services project on contraceptive use in Tanzania. He has directed the EVALUATION Project's facility level data collection efforts that have allowed the Mission to not only monitor the logistics, IEC, and training components of the program but to also link the data with DHS data for multivariate analysis. Finally with Angeles and Mroz, he developed methods for the evaluation of family planning programs when they are targeted and no experimental data is available. In an application of their methods to Tanzania, they find substantial bias in the estimated program impact when simple methods are used relative to methods that correct for program targeting.
Real Estate Research
For over ten years, Guilkey has collaborated with Mike Miles and several of his students to study various topics in real estate. Much of their early work involved a comparison of appraisal based methods to transactions based methods for valuing properties and portfolios of properties. Most of their current work has involved the use of pooled time series/ cross sectional data sets to forecast MSA level returns for various classes of commercial real estate.