Linking Topical and Categorical Preferences

Southeastern Geographer Vol. XX, No. 1, May 1980, pp. 42-57 LINKING TOPICAL AND CATEGORICAL PREFERENCES Robert E. Lloyd One of the fundamental propositions of Kelvin Lancaster's theory of consumer demand is that people "possess preferences for collections of characteristics, and that preferences for goods are indirect or derived in the sense that goods are required only in order to produce the characteristic ." (J ) Although it is difficult to separate in one's mind the contents (characteristics) from their containers (goods), it is really the contents that are preferred. In a migration context the important characteristics may be related to economic conditions, social relationships, cultural values , and the physical environment. (2) If a complete explanation of migration preferences and behavior is to be obtained, measuring people's preferences for environmental characteristics would seem to be more fundamental than measuring preferences for political units. A recent study by Davis and Casetti classified locational preferences into two types, topical and categorical. (3) They suggested that topical preferences relate to real world geographic entities, such as residential sites, cities, regions, or countries, and that categorical preferences are concerned with the attributes of such geographical units. They further argued that, "Perhaps topical preferences can be explained to a large extent in terms of categorical preferences, and these, in a sense, are the logical starting point for systematic investigations of the structure of preferences for real world geographical entities." (4) This argument is not new to geography. Rushton's "spatial behavior/ behavior in space" argument makes a similar case for observed behavior. He contended that geographers will gain more from studying "spatial behavior" or "the rules by which alternate locations are evaluated and choices consequently made" than they will by studying "behavior in space" or the "description of the actual spatial choices made in a particular system." (5) "In arguing that the rules for behavior in space are more fundamental than descriptions of observed behavior in space, we are implying that the rules by which we make choices among alternatives in space are not modifiable by the form of the system itself." (6) Dr. Lloyd is Associate Professor of Geography at the University of South Carolina in Columbia, SC 29208. Vol. XX, No. 1 43 A corresponding argument for migration would be that, although preferences for particular places in a system may change, preferences for characteristics, which are more fundamental, should be relatively stable. As a place is preferred only because of the characteristics associated with it, it will gain or lose favor as it gains and loses the preferred characteristics. (7) The basic preference for the characteristics should, however, remain relatively constant. This is not to say that general rules used by most people to make migration decisions cannot change over time, but it is argued that such fundamental changes occur slowly. The author has applied the revealed space preference methodology to census migration flows at the state level and found that trade-off rules for size and distance were relatively stable over the last four published censuses . (8) Size and distance, however, are only surrogate measures for the many characteristics that are potentially important to a migrant's decision -making process. If environment is interpreted in its broadest sense to include economic , social, and cultural factors as well as physical conditions, the most basic question raised is which characteristics of die environment are generally preferred and which ones are not. If these environmental conditions were expressed as binary alternatives, some would offer equally feasible choices, e.g., rugged or flat terrain. In other cases, all other things being equal, the choice would be obvious for most decisionmakers , e.g., clean or polluted air. (9) Although one characteristic may be selected over another by an individual , the preference for one characteristic over another may or may not be an important part of the aggregate decision-making process. The ability of individuals to express a preference for a warm or cool climate does not necessarily mean that relative temperature is an important consideration at the aggregate level unless a significant proportion of the population considers the characteristic an important criteria for selecting potential migration destinations. If there is consistency at the aggregate level between...