A collection of articles which provide examples that demonstrate the application of dynamic programming to a wide variety of decision problems in agriculture.

Humans interact with and are part of the mysterious processes of nature. Inevitably they have to discover how to manage the environment for their long-term survival and benefit. To do this successfully means learning something about the dynamics of natural processes, and then using the knowledge to work with the forces of nature for some desired outcome. These are intriguing and challenging tasks. This book describes a technique which has much to offer in attempting to achieve the latter task. A knowledge of dynamic programming is useful for anyone interested in the optimal management of agricultural and natural resources for two reasons. First, resource management problems are often problems of dynamic optimization. The dynamic programming approach offers insights into the economics of dynamic optimization which can be explained much more simply than can other approaches. Conditions for the optimal management of a resource can be derived using the logic of dynamic programming, taking as a starting point the usual economic definition of the value of a resource which is optimally managed through time. This is set out in Chapter I for a general resource problem with the minimum of mathematics. The results are related to the discrete maximum principle of control theory. In subsequent chapters dynamic programming arguments are used to derive optimality conditions for particular resources.

Humans interact with and are part of the mysterious processes of nature. Inevitably they have to discover how to manage the environment for their long-term survival and benefit. To do this successfully means learning something about the dynamics of natural processes, and then using the knowledge to work with the forces of nature for some desired outcome. These are intriguing and challenging tasks. This book describes a technique which has much to offer in attempting to achieve the latter task. A knowledge of dynamic programming is useful for anyone interested in the optimal management of agricultural and natural resources for two reasons. First, resource management problems are often problems of dynamic optimization. The dynamic programming approach offers insights into the economics of dynamic optimization which can be explained much more simply than can other approaches. Conditions for the optimal management of a resource can be derived using the logic of dynamic programming, taking as a starting point the usual economic definition of the value of a resource which is optimally managed through time. This is set out in Chapter I for a general resource problem with the minimum of mathematics. The results are related to the discrete maximum principle of control theory. In subsequent chapters dynamic programming arguments are used to derive optimality conditions for particular resources.

The agriculture industry is unique in the high level of influence from randomness in the environment due to the many ways in which an agricultural production system interacts with the environment. Static models fail to accurately characterize the sequential decision process of the farmer. Capturing the temporal dynamics of farm management decisions necessitates a sequential decision model. The demands of conceptualizing and computing sequential decision-making models are especially salient in agriculture, where substantial influence from countless exogenous random variables has made useful formulations of many farm-level decisions intractable. I begin with an assessment of decision making in agriculture, focusing on the potential of sequential decision models, and provide a brief history of their applications in farm-level decision making. I then present a generalized model of in-season crop production that ties all field level decisions to the temporal constraint inherent to a single production season. I discuss the considerations necessary to effectively model farm-level decision making with a dynamic model, and additional considerations for possible extensions of the model. I then present three applications of the model. The first application is to a simplified crop decision problem where a farmer is deciding the allocation of acres between corn and soybeans. The second application is to the prevented planting policy of Revenue Protection crop insurance. The third application is to a short-horizon alfalfa harvest optimization problem.

After all the research on agricultural risk to date, the treatment of risk in agricultural research is far from harmonious. Many competing risk models have been proposed. Some new methodologies are largely untested. Some of the leading empirical methodologies in agricultural economic research are poorly suited for problems with aggregate data where risk averse behavior is less likely to be important. This book is intended to (i) define the current state of the literature on agricultural risk research, (ii) provide a critical evaluation of economic risk research on agriculture to date and (iii) set a research agenda that will meet future needs and prospects. This type of research promises to become of increasing importance because agricultural policy in the United States and elsewhere has decidedly shifted from explicit income support objectives to risk-related motivations of helping farmers deal with risk. Beginning with the 1996 Farm Bill, the primary set of policy instruments from U.S. agriculture has shifted from target prices and set aside acreage to agricultural crop insurance. Because this book is intended to have specific implications for U.S. agricultural policy, it has a decidedly domestic scope, but clearly many of the issues have application abroad. For each of the papers and topics included in this volume, individuals have been selected to give the strongest and broadest possible treatment of each facet of the problem. The result is this comprehensive reference book on the economics of agricultural risk.

Dynamic Programming and Its Applications provides information pertinent to the theory and application of dynamic programming. This book presents the development and future directions for dynamic programming. Organized into four parts encompassing 23 chapters, this book begins with an overview of recurrence conditions for countable state Markov decision problems, which ensure that the optimal average reward exists and satisfies the functional equation of dynamic programming. This text then provides an extensive analysis of the theory of successive approximation for Markov decision problems. Other chapters consider the computational methods for deterministic, finite horizon problems, and present a unified and insightful presentation of several foundational questions. This book discusses as well the relationship between policy iteration and Newton's method. The final chapter deals with the main factors severely limiting the application of dynamic programming in practice. This book is a valuable resource for growth theorists, economists, biologists, mathematicians, and applied management scientists.

This book has the purpose of providing the "state of the arts" concerning bio-economic modelling dealing with agricultural systems. In most cases, the contributions use a methodology combining the use of biophysical and economic models, in all cases, an engineering production function approach is totally or partially applied. This practice is being developed in the last years as a response to concrete policy matters: agricultural policies are increasingly combined with environmental and natural resources policies, and this reality involves the need of an integrated assessment, that current economic models are not able to provide.