Spatial Location Choice

Spatial choice is an important topic in my research. I have studied both home and establishment location choice based on the work of William Alonso, John Douglas Hunt (MSc advisor), Kara Kockelman (postdoc advisor), Steve Lerman, Francisco Martinez, Eric Miller (PhD committee member), Rolf Moeckel, Paul Waddell (PhD committee member), and many others. I was therefore pleasantly surprised to be directed by my colleague Byron Miller to a 1980 paper by Eric Sheppard entitled “The Ideology of Spatial Choice”. Eric provides many arguments that parallel my own thoughts on the topic, which I will share here.

A first element to highlight is that location choice comprises two elements. First, following Manski and others in the econometrics literature, choice set formation that defines the available alternatives. Sheppard notes that the choice set is a function of current social structures - the system of values and economics that lead to a particular set of dwelling and neighbourhood typologies - which contrasts with the standard econometrics assumption that all alternatives are available to the decision-maker and they simply make a decision as to which alternative they prefer. Second, the spatial location choice is made conditional upon the available choice set. At this stage, it is often assumed that the decision-maker is a utility maximizer - i.e., makes a fully compensatory choice after weighing all the attributes of all available alternatives in their choice set. Sheppard notes that such a decision process is both infeasible and incomplete in its assumption of the means of decision-making.

The choice set problem is fundamental because it “grounds verification in comparing our theories to actual observed events” (Sheppard, 1980). In other words, the empirical approach characteristic of engineering and economics implicitly assumes that, so called, revealed preferences based on observed choice represent the decision-maker’s preferences unconstrained by the decisions of other actors, including other residents, property developers, and government regulations. This characterization of choice as synonymous with preference is not founded in immutable law. Rather, it is based on a theoretical framework put forward by the economist Paul Samuelson in the 1940s, and subsequent perpetuation by other scholars leading to it becoming de facto axiomatic.

A further point made by Sheppard that I wholeheartedly endorse is that rational choice is typically conflated with the egoistic choice of homo economicus. It is argued by many scholars that deviations from purely egoistic rational choice is a function of imperfect information and inferior ability, rather than a mischaracterization of the choice process. Rarely does the discussion touch on the goals of behaviour, a topic that I have been putting considerable thought towards addressing in my work. Forty years on from Sheppard, it still rings true that few utility-based models of spatial choice exist that consider the utility of others or collective utility. As argued by Sheppard, such assumptions about human nature are also statements on current social relations. He notes that there is an inductive justification of such assumptions - i.e., in a classical liberal society, individuals often do act egoistically and it is therefore valid to model behaviour thus.

Critiques of conventional economic theory, such as that outlined above with application to spatial location choice, are examples of what Schumacher termed meta-economics for their questioning of neoclassical economic theory and its promotion as a universal law of human nature. Applying the concept of meta-economics to non-market valuation, Schumacher notes that the valuation of ecosystem services typical of environmental economics “is a procedure by which the higher is reduced to the level of the lower and the priceless is given a price”. Both Schumacher and Sheppard note that economic analysis is usefully applied within a given framework that as Schumacher says “lies altogether outside the economic calculus”. Sheppard argues that any validation of economic assumptions must necessarily be independent of the theory it seeks to validate. An assessment of choice behaviour requires a method of evaluating the decision-maker’s level of information about, and preferences for, the available alternatives. The stated preference approach pioneered by Louviere, Hensher, Street, and others provides such a framework for evaluating alternatives. However, as noted by Hollis and Nell, one enters an infinite recursion in that the stated preference approach then requires its assumptions to be evaluated by a third theory, etc. In the same way, Sheppard argues that revealed preference is tautological in its assumption that people choose what they prefer.

We now come to the question of what drives the discrepency between choice and preference. One explanation is that choices are constrained by the available alternatives - i.e., choice set. Revealed preference analysis assumes that the most frequently selected choices from historical empirical data are indicative of preferences. However, it may well be the case that these data represent choice from a non-exhaustive set of alternatives and, as stated by Sheppard, “represent no preference at all”. Sheppard provides an excellent example that again rings true to my own thoughts on this topic. In the standard bid-rent model of Alonso, suburbanization arises from the preferences of high income households for space. The implication of the model according to Sheppard, and I support with him on this point, is that the crowding of low income households onto “higher priced inner-city land” is an outcome of their preferences. However, higher income households can clearly outbid lower income households regardless of the preferences of each group. Harvey also supports the argument that revealed preference observations are more indicative of the capitalist social structure than household preferences for housing. This discussion does not even touch on the sociospatial structures and exclusionary zoning constraints on home choice. Let me conclude with a quote by the great location theorist, August Lösch, himself referencing Schiller: ” ‘Live with thy century, but be not its creature; give to thy contemporaries, but give what they need, not what they laud’? No! The real duty of the economist is not to explain our sorry reality, but to improve it. The question of the best location is far more dignified than determination of the actual one”.