Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → États-Unis → 2011 → Measuring Drought Resistance : A State-Contingent Evaluation of In Situ Drought Resistant Wheat and Maize in Northern China

University of California, Davis (2011)

Measuring Drought Resistance : A State-Contingent Evaluation of In Situ Drought Resistant Wheat and Maize in Northern China

Blanke, Amelia Hughart

Titre : Measuring Drought Resistance : A State-Contingent Evaluation of In Situ Drought Resistant Wheat and Maize in Northern China

Auteur : Blanke, Amelia Hughart

Université de soutenance : University of California, Davis

Grade : Doctor of Philosophy (PhD) 2011

Résumé partiel
Drought, a perennial threat to agricultural production, may become increasingly frequent and increasingly harmful due to global climate change and the widespread depletion of groundwater resources. In response to severe droughts and the threat of more to come, policy makers, scientists and corporations have all hailed drought resistant varieties (DRVs) as a potentially valuable part of the solution. Research into a new generation of DRVs, based on advances in genetic modification and renewed focus in conventional breeding programs, promises to provide a new set of varieties with higher yields than conventional in a drought, perhaps at the cost of production in non-drought conditions. Such a drought production boosting technology complements other drought mitigation income smoothing mechanisms available to households (e.g., irrigation, drought insurance, crop choice, informal borrowing and asset sales) and promises to do so at relatively low cost to both farmers and policy makers. As a new generation of DRVs hits the market over the next decade, rural households, corporations and governments will demand evaluations of efficacy. Carefully controlled horticultural research studies may not be adequate to characterize the technology under the wide range of growing conditions used in the field. This dissertation demonstrates one solution to this problem. Varieties with a reputation for drought resistance have been available to farmers for decades, but have received little attention from economists. I use data collected from rural households in Northern China growing either wheat or maize to evaluate the methodology. Conventional econometric evaluation techniques fail to capture the state-contingent benefits of this technology. DRVs are, in essence, a state-contingent technology ; their impact on production is crucially dependent on the state of nature : drought or non-drought. A conventional analysis that ignores state, analyzes differences in yield variance alone, or simply includes the state as one of many explanatory variables will produce an underestimate of the benefit of using DRVs in a drought state that is the product of the analyst‘s choice of model. State-contingent production offers a theoretically attractive method for modeling, but has proven notoriously difficult to implement empirically. States of nature may be too numerous or unidentifiable. Production is only observed in the one state which occurred and state allocations of inputs are rarely observed. DRV evaluation offers a relatively simple case that is substantially easier to implement. By using a discrete two state framework and setting up DRV adoption as the primary way to shift production between drought and non-drought states, I am able to estimate reduced form profit and production equations consistent with state-contingent production theory. With sufficient data, these estimates allow me to answer an increasingly pertinent question : are DRVs truly drought resistant ? This paper uses original data on varieties in use in Northern China to demonstrate a methodology to evaluate DRVs, or any other similar stochastic production technology, within a state-contingent production framework. Household panel data collected in the provinces of Hebei, Henan and Ningxia in 2001, 2004 and 2007 includes detailed data on agricultural production and technology adoption decisions. Within the sample of 280 households growing wheat, adoption of varieties identified by farmers as drought resistant increased from 23 % of selected plots in 2001 to 48 % in 2007. Adoption rates are higher in the sample of 338 maize growing households, increasing from 41 to 59 % of selected plots. Because I have data from the same households across multiple years, I am able to control for unobserved time-invariant differences between households that may bias results. In addition, the data set provides several instruments that allow me to control for unobserved time-varying differences between adopters and non-adopters that may change over time. I supplement these data with a set of rainfall data and a model that allows me to calculate a discrete drought index based on estimated plot soil moisture.

Search Oxford Libraries Online (SOLO)

ProQuest Dissertations Publishing

Page publiée le 16 avril 2013, mise à jour le 5 décembre 2018