Dissertation
My Ph.D. Dissertation is titled:
“POVERTY, ECONOMIC NEED, AND LANDMINE/UXO INCIDENTS IN CAMBODIA“
I took this picture in Pailin - an area of Cambodia still heavily laden with landmine-related ordnance. As this boy walks to fill his canisters with water, he crosses through one of the many landmine fields. This picture captures the emotion that provided the motivation behind my dissertation.
Dissertation Committee (left to right): Norman Waitzman, Ph.D., Thomas Maloney, Ph.D., Richard Fowles, Ph.D., Cihan Bilginsoy, Ph.D. (Chair), Zachary Zimmer, Ph.D.
ABSTRACT
This dissertation analyzes the relationship between poverty, economic need and landmine/UXO incidents in Cambodia. In the aftermath of decades of military and political conflict, an estimated 23 million landmines and 324,000 tons of unexploded ordnance litters the Cambodian landscape. In total, 58,000 Cambodian’s suffered landmine/UXO incidents between 1980 and 2007. Various governmental and non-governmental organizations combating the landmine problem focus primarily on landmine removal and mine risk education, but scant attention is paid to the underlying socioeconomic factors that precede incident occurrence. The objective of this dissertation is to determine the impact of economic conditions on the number of incidents. In this effort, I distinguish between accidents and tampering - intentional handling of ordnance - on grounds that each incident type may be driven by distinct factors.
I first employ several socioeconomic indicators of poverty to test, at the district level, the hypothesis that higher economic need leads to more landmine/UXO accidents and tampering. The dependent variable is the number of incidents (per 10,000) pooled over the 1996-2000 period. Independent variables are drawn from the 1998 General Census of Cambodia. Regressions are performed with techniques that control for spatial dependence. I find that incidents increase with higher levels of single parenting, use of firewood for cooking, migration, population density, male-female sex-ratio, the K5 mine belt, and with a lack of education. The magnitudes of these effects however, vary by incident type. I suggest that policymakers track various measures of poverty and provide relief in times of economic need in an effort to reduce future incidents.
Next, I perform a panel analysis of 26 provinces in Cambodia between 1998 and 2006 to test the hypothesis that agricultural vulnerability increases landmine/UXO incidents. I find that landmine/UXO accidents and tampering incidents decline with improved agricultural performance as measured by net rice output, supply of water, rice yields, crop diversification, floods/droughts, and non-rice agricultural production. I suggest creating a sustainable irrigation-based infrastructure, implementing high-yielding rice varieties, providing initial seed-stock supplies, and diversifying crops as policy instruments that can reduce landmine/UXO incidents.
Finally, I test the response of tampering to the price of scrap metal at the Cambodia-Thailand border. Testing this hypothesis became possible following the attainment of an original price record which contains data from April 1996 to December 2007. I regressed the number of monthly tampering incidents on the price level, with specific controls for month fixed effects. I find that tampering activities respond directly and more than proportionately to a change in the price of scrap metal. I suggest a simultaneous effort of metal price-fixing in conjunction with on-site ordnance payment programs as a solution to tampering; the most prevalent form of incident in Cambodia.