James E. Wilen
Dept of Agricultural and Resource Economics
3102 Social Science Humanities Bldg
University of California
One Shields Avenue
Davis, CA 95616
Email:

POSITION STATEMENT

At UC Davis, I have been working over the past several years at incorporating more realistic and explicit notions of space into my research and graduate teaching. On the teaching front, in addition to incorporating new spatial material into my lectures in graduate natural resource economics, I have also organized and conducted formal seminar courses for graduate students focused on spatial econometric methods. These seminars were oriented around Anselin's books and economic journal articles by researchers including Ann Case and Nancy Bockstael. On the research front, I have supervised three major Ph.D. thesis efforts to completion recently (with two more in progress), all aimed at exploring the spatial behavior of resource users, principally in fisheries. These projects have been motivated by new questions that are fundamentally spatial in nature, and not informed by conventional non-spatial conceptual and empirical methods.

My motivations for moving in spatial directions are several. First, newly available information derived from remote sensing and GIS technologies presents opportunities to do empirical work with data sets that may contain unexploited sources of variation associated with spatial heterogeneity. This is important because good empirical economics relies on "natural experiments" to identify behavioral parameters of interest. For the most part, economists have had to find natural experiments in time series data, and time series data present their own difficulties that inhibit unbiased and efficient parameter identification. Economic decision makers surely make decisions that depend upon space as well as time. Ignoring the spatial dimension eliminates the opportunity to exploit additional natural experiments that are spatial as well as temporal. At minimum, ignoring spatial heterogeneity foregoes statistical efficiency gains, and it may also wash out important processes by aggregating behavior that is fundamentally spatial. A second reason for studying spatial economic processes is that, for the most part, economics is essentially a non-spatial discipline. While some subdisciplines have had to incorporate spatial dimensions into their conceptual foundations, most of the core of economics has not done so. This may be because adding the spatial dimension complicates deductive theorizing exponentially rather than proportionately. But the difficulties of exploring alternative spatial theories are rapidly diminishing as computational power increases. Even simple computational methods allow numerical exploration of decision making in experimental environments that incorporate spatial interaction and spatial externalities, different degrees of spatial coordination and myopia, and different kinds of spatial equilibration processes. Full exploitation of these new tools will probably require that economists overcome their reverence for "pure" deductive model-based thinking and come to appreciate the essential usefulness of computational methods for generating and testing behavioral hypotheses in more complex settings. In addition, we will probably have to develop new intuition about spatial processes, by learning how computational methods complement rather than substitute for analytical deductive modeling.

As a natural resource economist, I see particularly exciting new conceptual and empirical opportunities associated with spatial modeling. In many ways, the need to address spatial dimensions of resource utilization problems is perhaps more pressing than in some of the other subdisciplines in economics. Natural resource economics has always been a field that is closely tied to supporting science disciplines, including biology and ecology, soil science, hydrology, etc. While most of these hard science disciplines have begun to transform themselves in ways that explicitly incorporate space, natural resource economics itself seems stuck in a paradigm that focuses on time in a mostly spatially heterogeneous setting. Thus continued multidisciplinary collaboration calls for an upgrading of the foundation of the field, in ways that parallel developments in supporting disciplines. In addition, there are some interesting natural resource policy options that are beginning to come under consideration that require an explicitly spatial framework. For example, many terrestrial mammals are managed with zoned policy instruments that take into account (generally in an ad hoc manner) the fact that the resource is migratory. What is often ignored is that resource users in this setting are also acting strategically to account for the fugitive nature of the resource. Thus managers are bound to be surprised unless they are able to forecast the spatial behavioral implications of policies. This is probably best done with integrated bioeconomic models that not only account for fundamental dynamics of resource and resource user behavior, but also fundamental spatial processes of both components. Another important example of a new resource management policy that requires spatial analysis in order to understand the potential implications is marine reserves. Understanding the implications of marine reserves requires, at minimum, a biological modeling approach that accounts for the important spatial processes at work, but also an understanding of the spatial behavior of resource users.

I see the study of spatial processes in resource economics as a "new frontier" of sorts, and one that will transform the field in a way similar to the manner in which Pontryagin's optimal control theory ushered in the "dynamics revolution" of the 1970s. There are several ways in which an institution like CSISS might accelerate progress. The first is to continue to promote, offer, and even subsidize exposure of students and faculty to spatial data handling methods, spatial statistics, and spatial modeling procedures. There is a considerable demonstration effect achieved by sprinkling students who have had some exposure to these workshops around the country in various programs. A second need is for sponsorship of additional workshops, proceedings, and special sections at ASSA, AEA, and AAEA meetings to promote spatial analysis. Most important is the need to demonstrate how spatial analysis matters. That is, what kinds of issues absolutely require spatial conceptual and empirical methods? What kinds of circumstances give rise to tangible payoffs to the investments necessary to master spatial methods? What happens when you ignore spatial heterogeneity and spatial processes? When are big mistakes in understanding most likely?