There are two simple and overarching reasons why economists have pursued little spatial analysis. The first is that spatially explicit economics data is difficult to obtain. The second is that many economists remain unconvinced that space matters in the systems that they study.
While the second problem is fundamental, the first problem is not inconsequential. Satellite imagery, remote sensing technology, and over-flight photography provide a means of registering physical phenomenon, but only rarely are the things that economists study observable in this sense. Economists model human decisions, and only occasionally are these decisions “deducible” from pictures. Land use change is one of the few examples where they are. But even in this case the boundaries of the decision units are unobservable from above, as are many of the explanatory variables that would be needed at a spatially disaggregated level to incorporate into the models of human behavior (e.g. prices, incomes, transactions, etc.)
It is notable that the economic analyses that have succeeded in incorporating spatial data are those where either the human decision is observationally deducible (e.g. deforestation and land use) or where the human activity involves recording location for some other reason (e.g. housing transactions, regulated marine fishing activities). Attempts at collecting data on human actions at a spatially disaggregated level are frustrated, however, by confidentiality regulations that prohibit the dissemination of data that would allow one to deduce the identity of an individual firm. Because geocoding provides location, it inherently violates those regulations. This has presented particular obstacles to spatial analysis of agricultural systems, because agriculture has even more stringent privacy provisions than most industries.
Nonetheless, spatial data of interest to economists is on the increase, and more could be done with spatially explicit data if economists chose to use it in their empirical work. But many economists, even within the environmental and resource economics profession, remain unconvinced of the value of spatial analysis. At best they see geographic information systems as a means of capturing and storing a richer data base – or providing greater accuracy and variability in variables for their otherwise aspatial analysis. The analyses that actually care about space, generally do so because space matters to someone else. For example, the location of human activity matters because of its effects on the environment (such as a subwatershed or wetland) or the stock of natural resources (like fish or forests) – none of which are themselves locationally fungible.
An altruistic motivation is likely to have a limited impact on a profession like economics. Only if it matters inherently to the economics of the problem will spatial analysis reach a suitable level of sophistication in economics research. From a methodological viewpoint, there have been a number of advances, with spatial econometrics methodology for handling spatial data paralleling the advances made decades ago for time series. Yet broad acceptance of the importance of temporally dynamic models of economic processes does not have a parallel in spatially dynamic models. In fact there are few inherently spatial models in economics. The ones that come to mind are the monocentric city model of land price gradients and the new economic geography as typified by Krugman’s work, some of this arising in the regional economics field. And there are even fewer empirical applications. This is despite the fact that there are some inherently spatial processes such as endogenous interactions, contagion, diffusion, dissemination that have interest for economists. Economics will benefit more from efforts (many of which may be underway already) that are devoted to the development of explicitly spatial models of economic processes as well as adaptation of existing spatial models to economics problems from other disciplines. At this point, the development and dissemination of ways of modeling spatial processes will contribute more to spatial analysis in economics than further attempts at improving accessibility to GIS data and software.
Bockstael has advised federal and state agencies, lectured at institutions
in the U.S. and Europe, served on National Academy Committees and NSF panels,
and has consulted for the Justice Department and NOAA in several natural
resource damage assessment cases. Her work is published in JEEM, Land Economics,
American Journal of Agricultural Economics, Marine Resource Economics,
American Economic Review and the Economic Journal. She has served
as an associate editor of the Journal of Environmental Economics and Management,
and is now on the Editorial Council of that journal and Land Economics
as well as the Review Board of Environment and Development Economics. She
has been a member of EPA’s Science Advisory Board since the creation of
its economics subcommittee in 1991 and is President-elect of the Association
of Environmental and Resource Economists.