Integration of Economic Geography and Social Psychological Models of Patronage Behavior

Robert F. Lusch, University of Oklahoma
ABSTRACT - This article develops a model of patronage behavior which incorporates key concepts from geography, social psychology and economics. The model is shown to be rich in explanatory power but its predictive power is not assessed.
[ to cite ]:
Robert F. Lusch (1981) ,"Integration of Economic Geography and Social Psychological Models of Patronage Behavior", in NA - Advances in Consumer Research Volume 08, eds. Kent B. Monroe, Ann Abor, MI : Association for Consumer Research, Pages: 644-647.

Advances in Consumer Research Volume 8, 1981      Pages 644-647


Robert F. Lusch, University of Oklahoma


This article develops a model of patronage behavior which incorporates key concepts from geography, social psychology and economics. The model is shown to be rich in explanatory power but its predictive power is not assessed.


The study of retail patronage behavior is not new. Over fifty years ago W. J. Reilly (1929) in his now classic monograph developed a gravity based model of retail patronage behavior. This model postulated that a household's attraction to a community for shopping purposes was positively related to the size of the community and inversely related to the distance to the community. Some thirty years later David L. Huff (1962) transformed Reilly's law into an intra-urban patronage model. Huff's model suggested that a household's attraction to a store was positively related to the size of the store (i.e., square feet) and inversely related to the distance from the household's place of origin to the store. At about the same point in time Bob R. Holdren (1960) in an elegant model of the structure of retail markets postulated that the buyer evaluates the store offering on three dimensions: (l) the distance the buyer is from the store; (2) the level of prices at the store; and (3) the level of non-price offer variation. All three of these authors viewed patronage behavior from an economic geographic perspective. That is patronage behavior was viewed as being determined by the structure of physical space. This spatial orientation to patronage behavior continues to be pursued. For example, recent work by David MacKay (1973), Dennis Lord (1975) and William Young (1975) is reflective of the spatial research orientation toward patronage behavior.

On the other hand some early pioneers in patronage behavior research attempted to explain shopping behavior in terse of social-psychological concepts. Utilization of these concepts resulted in researchers treating shopping behavior as being unaffected by physical space. Notable early examples of this approach include Gregory Stone's (1954) work on categorizing shopping orientations used to evaluate stores. Stone's shopper orientations fell into five categories: (1) economic, (2) personalizing, (3) ethical, (4) apathetic, and (5) a residual category of unique or indeterminate criteria. Another early example includes the work of George Fisk (1961-1962) in which the consumer's image of the store was postulated as the major latent guiding force in determining patronage behavior. The social psychological approach to explaining patronage behavior continues to receive attention. Some relatively recent examples include the work of Darden (1980), Grenada and Painter (1976), Monroe and Guiltinan (1975), James et al. (1976), Miller (1976), and Marks (1976).

As far as the author is aware the only attempt to integrate the geography and social psychology approaches to studying patronage behavior is in the area of what has been labeled cognitive mapping. In this stream of research the consumer's perceptions or cognitions of store size and/or distance are utilized in traditional geographic patronage models rather than real distances or actual store sizes. This approach is relatively recent and includes the research of Maze (1974), Cadwallader (1975) and MacKay and Olshavsky (1975). Findings from this stream of research suggest that cognitive or perceptual measures are better predictors of behavior than physical or real measures.

The purpose of this paper is to provide further integration of the geography and social psychological approaches to the study and explanation of patronage behavior. In addition useful concepts from microeconomics are integrated into the proposed patronage model. Portions of the modal are taken from a theory of retail structure recently developed by Ingene and Lunch (forthcoming).


As in microeconomic theory we begin by assuming that the consumer (i.e., shopper) is rational. It is however recognized that not all consumers have the same rationality. What may be rational from one shopper's perspective may not be rational from someone else's perspective, but, nonetheless rationality as the individual defines it does exist.

In addition it is assumed that shoppers don't have perfect information. Therefore shoppers must make decisions in the face of uncertainty.

The fundamental underlying proposition upon which the model is based is that shoppers will make rational shopping decisions so as to maximize their net consumer surplus. Consumer surplus is a concept borrowed from economics. Consumer surplus occurs when goods are sold at a price lower than the maximal demand price a consumer would be willing to pay for them. It is because goods are offered by retailers below the consumer's maximal demand price that consumers are enticed into expending effort to shop for goods.

When the price the consumer must pay for the product at the store plus the costs associated with acquiring the product, such as the cost of traveling to and from the store, exceeds the consumer's maximal demand price shopping will become uneconomic. The preceding implies that there are certain costs incurred in shopping. When these costs are subtracted from consumer surplus one arrives at net consumer surplus. It is net consumer surplus that the consumer attempts to maximize.

The Patronage Model

The model that is being proposed of patronage behavior can be formalized as follows:

E(Wj)=Pj(Pm-Pe)q-2rt'-[(2r/m)+s]t"-t"'   (1)


E(Wj) = expected net consumer's surplus from shopping at store j,

Pj = probability that if store j is patronized that the product being sought is in stock,

Pm = maximal demand price per unit given quantity desired,

Pe = expected price per unit,

q = quantity desired,

r = distance in miles from store to household departure point, and therefore 2r is equal to round trip distance,

t' = cost per mile of travel,

m = speed of travel in miles per hour,

s = expected in-store shopping time,

t' = opportunity cost per hour of time,

t"' = psychic costs (benefits) of shopping trip evaluated in dollar terms.

Appreciation and understanding of this store patronage model can be facilitated by an elaboration of its individual components.

The first term (Pj) is included in order to capture the degree of uncertainty that confronts the shopper. Once a consumer recognizes a need for an item it must assess the set of stores which may have the item and estimate the probability that the item is in stock at each of the respective scores. This probability is the Pj term and it captures uncertainty on the part of the shopper. In reality the product the consumer plans to shop for either is or is not in stock at the store. That is if the consumer knew the truth then Pj would be assigned a value of 0.0 or 1.0. But the consumer does not know reality (i.e., certainty) and therefore Pj is its estimate of the likelihood that the item will be in stock at the store.

The term Pj is itself a function of several managerial and behavioral variables that are not explicitly included in the model. Most important from a managerial and behavioral perspective is the information the prospective shopper has available. This information is a function of the quantity and quality of promotion on the part of the retailer and also of the information processing capabilities of the shopper. Theory and concepts from social psychology can therefore be used to empirically study the determinants of Pj.

The second term, the maximal demand price (Pm) captures the consumer's desire, need or want for the item and the consumer's ability to acquire the item. As the consumer's purchasing power increases the maximal demand price will rise. Purchasing power may increase due to rising intone or more readily available credit. Thus the retailer can have some influence over Pm via its credit policies. In addition the retailer could advertise or promote the item to have certain unique and strong traits or benefits which could result in the consumer wanting or desiring the item more and therefore shifting Pm upwards. Behaviorally oriented theories of advertising could be helpful in shedding light on this matter.

The expected price per unit (Pe) reflects what the consumer believes the retailer will charge for the item being sought. Once again this term suggests the shopper faces uncertainty. A shopper seldom knows a priori what the precise price of an item will be. Therefore the shopper must estimate the price. As you might well expect the retailer's promotional efforts--especially its advertising--can aid the shopper in better estimating prices. In addition the process of learning will become important in establishing (Pe). Every time the consumer shops at a store it learns more about its pricing policies.

The quantity (q) the shopper plans to purchase on any given shopping trip is basically an inventory decision if the item is a nondurable rapidly consumable item such as food. A household will tend to purchase nondurables in larger quantities when its income is higher; the cost of travel is high; time pressures are great; storage is available and household size is larger. Most durable items (e.g., care, furniture, appliances) are purchased one at a time. If more than one is purchased on a single trip it is usually due to income or wealth factors.

Combining together the four variables we have discussed (Pj, Pm, Pe, q) will allow us to formalize expected consumer surplus. Expected consumer surplus is equal to [Pj (Pm- Pe)q]. From this the costs of shopping must be subtracted. The major variables influencing the costs of shopping will next be discussed.

In traditional economic geography models the most important variable is (r) or the distance from the store to the shoppers point of departure--usually its residence or place of employment. Naturally this variable is a function of where the retailer locates its store and the location of households or businesses surrounding it. But (r) is also a function of city planning of road networks. The straight line distance between the store and shopper may be short but if travel is to occur on existing roadways the distance could be considerably greater. From a behavioral perspective (r) is a function of learning, information availability and information processing. The more times one travels to the store the more they will lean the shortest route. Or in terns of a first trip the more the information the consumer has on how to get to a store the better it will be able to travel the shortest route and thus minimize shopping costs.

Travel costs per mile are also important in determining patronage behavior. In the model the cost per mile of travel is labeled (t'). This important variable is influenced by the availability and cost of public transportation and if private transportation is used the cost of operating an auto. Costs of operating an automobile will vary by region of the country due to varying rates for insurance, gasoline, and taxes in different parts of the country. The total round trip costs of traveling to the score are represented in the patronage model by 2rt'.

The costs of shopping not only include travel costs but also the implicit cost of time. As families are increasingly facing a scarcity of time the opportunity cost of that time rises. In the model the opportunity cost per hour of time is labeled t". This variable is a function of the alternative opportunities an individual has for leisure or work. One would expose that some of the correlates of t" would include: income, lifestyle, and social class.

The total amount of time spent on the shopping trip can be divided into two components: (1) travel time; and (2) in-store shopping time. Travel time is equal to (2r/m) in the patronage model, where m is average speed of travel in miles per hour. In most part speed of travel is a function of the type of transportation used, speed laws, road networks, and road congestion. In-store shopping time is labeled s in the model. This time variable is a function of a variety of factors which include: (1) the amount of information the shopper has on where the item(s) is located in the store; (2) in-store congestion; (3) store layout; (4) degree of sales assistance available; and (5) merchandise availability.

The cost of time per hour (t") when multiplied by the amount of time spent on the shopping trip (2r/m)+s will determine the total cost of time devoted to the shopping trip.

The final parameter in the patronage model is t''' or the psychic costs or benefits of the shopping trip. The shopping trip cannot only cost in terms of direct monetary travel costs and the opportunity cost of time devoted to the shopping trip but also in term of psychic costs. For example let us assume that in order for a shopper to travel from its home in the suburbs to a large downtown department store that it must maneuver its auto over 32 miles of congested freeway and then through 12 blocks of ghetto neighborhoods. In addition to the time and travel involved this trip may be very nerve wrecking and tension creating thus resulting in certain psychic costs.

In principle psychic benefits can also accrue during a shopping trip. That is the shopping trip may be viewed as an enjoyable experience and thus of value in and of itself. For instance assume that it has been a long cold winter and with the arrival of spring the consumer desires to load his family into the car and drive to a regional shopping center forty miles away in order to browse in the shopping mall and look at the new spring fashions. This shopping trip may actually be an enjoyable experience thus creating psychic benefits vs. psychic costs.


The traditional patronage models in economic geography, such as the Huff model, have good predictive power but are lacking in explanatory richness. This is not the case with the proposed patronage model. Not only can the proposed model explain the impact on patronage of changes in behavioral, environmental and geographic variables but it is also capable of explaining well accepted theories such as central place theory, Reily's law and the Huff model.

Behavioral Variables

There are at least several behavioral variables which can fit well into the model and increase its potential explanatory power. Some of these variables include information processing, attitudes and consumer certainty.

Lets briefly look at information processing. As a person obtains sere and better information about a store they will be better able to make shopping decisions which will increase their net consumer surplus. Increased net consumer surplus via better informed decision-making will lead to consumers being more store loyal.

A second behavioral variable is customer attitude toward the store. If attitudes become more favorable then the store should become a more enjoyable place to shop which should reduce the psychic costs (t''') of patronizing the store. This in cum could increase the distance (r) a shopper would be willing to travel to visit a store.

A third behavioral variable regards the degree of certainty the consumer has that the product being shopped for is in stock. As certainty increases the number of patrons a store can potentially attract increases because the distance a consumer would be willing to travel (r) rises. At the same time existing patrons will become more loyal because their net consumer surplus will rise.

Environmental Variables

In terms of environmental variables we could investigate the impact of changes in things such as transportation costs and speed of travel. As transportation costs rise or the average speed of travel declines several things could happen. Consumers may purchase in larger quantities so they will not need to make as many shopping trips; they may try to gather more information to increase their certainty as to whether the product is in stock; they may decide to patronize more stores closer to their residences or places of work; or they may try to reduce the amount of time involved devoted to in-store shopping by more carefully planning their shopping trips.

Geography Concepts and Models

The proposed patronage model is not only capable of explaining the impact of changes in behavioral and environmental variables on patronage behavior it also helps to explain central place theory, Reilly's law and the Huff model.

Central place theory (Christaller 1966) tolls us that there is a hierarchy of communities each of which is progressively larger and performs a larger group of retail functions representing an increased variety of product availability. The smallest communities perform the most basic functions and the larger perform more specialized functions. In addition each successively larger community performs all lower-order basic functions. The larger communities are spaced farther apart and are bordered by smaller communities. In terms of the proposed patronage model this suggests that the larger the community the greater the probability that an item can be found in a store within the community. The larger the community therefore the greater will be the distance (r in the model) people will be willing to travel to the community. This is in principle what central place theory concludes.

Both Reilly's law and the Huff model postulate that patronage is inversely related to distance and positively related to size of store or size of community. In the proposed patronage model distance (r) is also incorporated in such a fashion that an increase in distance reduces net consumer surplus and therefore the likelihood of a patron shopping at the store declines. Size of store is not directly incorporated in the proposed patronage model. Importantly, however, what size of store is a surrogate for is included in the model. Size of store itself is not an attraction force but rather the increased width and depth of merchandise assortments which accompany large stores is the sectarian force. When these assortments increase then the probability that a shopper will find the item(s) it is seeking will increase which will in turn increase consumer surplus. Increased consumer surplus will increase a store's ability to draw patrons. Thus we see that what the size variable is reflective of is more directly incorporated in the proposed patronage model. And importantly it allows for the same end conclusion to be drawn as the size variable in the Reilly and Huff frameworks.


The patronage model as it has been presented is not complete in all respects. Further work and refinement is needed. For example the model assumes that only one product is being shopped for. This is unrealistic. To make the model more realistic multiple product searches should be included. If the shopping trip is a multi-purpose single stop trip then the model can be extended by summing across items to be purchased. However if the trip is a multi-purpose multi-stop trip then extension of the model is considerably more complex.

In addition the model could become richer if it not only explicitly included demand related variables but also supply and competitive variables. This would enable the retail manager to manipulate certain managerial variables in order to determine a profit maximizing strategy.

Finally the model is in need of rigorous testing of both its individual components and its total composition. It is believed that all of the variables in the model can be measured directly except the psychic costs or benefits of shopping. Since many of the correlates of the major constructs in the model are social-psychological in nature, indirect measures may be necessary for them.

To better understand the relationship between patronage behavior and each of the variables in the model several approaches could be used. First, calculus could be used to establish the impact of a change in each variable on net consumer surplus. Second, simulation could be used to investigate the magnitude of these effects. Third, empirical research could be used to test hypotheses derived from the two preceding approaches.


A model of patronage behavior has been presented which incorporates concepts from economics, social-psychology and geography. The model appears to be rich in explanatory power. With additional refinement its explanatory power should be further improved. Its predictive power, however, still remains to be assessed.


Cadwallader, Martin (1975), "A Behavioral Model of Consumer Spatial Decision Making," Economic Geography, 51, 339-349.

Christaller, Walter (1966), Central Places in Southern Germany, Englewood Cliffs, NJ: Prentice-Hall, translated by C. W. Baskin from the 1933 German edition.

Darden, William R. (1980), "A Patronage Model of Consumer Behavior," in Competitive Structure in Retail Markets: The Department Store Perspective, Ronald W. Stampfl and Elizabeth Hirschman eds., Chicago: American Marketing Association, 43-52.

Fisk, George (1961-1962), "A Conceptual Model for Studying Customer Image," Journal of Retailing, 37, 1-8, 51.

Granzin, Kent L. and Painter, John J. (1976), "Store Attributes and Shopper's Attributes As Determinants of Retail Store Patronage," paper presented at the Southwestern Marketing Association Conference.

Huff, David L. (1962), Determination of Intra-Urban Retail Trade Areas, Los Angeles, California: Real Estate Research Program at the University of California-Los Angeles.

Ingene, Charles A. and Lusch, Robert F. (forthcoming), "A Model of Retail Structure," Research in Marketing.

James, Don L., Durand, Richard M. and Dreves, Robert A. (1976), "The Use of a Multi-Attribute Attitude Model in a Store Image Study," Journal of Retailing, 52, 23-32.

Lord, Dennis J. (1975), "Locational Shifts In Supermarket Patronage," Professional Geography, 27, 310-314.

MacKay, David B. (1973), "Spatial Measurement of Retail Store Demand," Journal of Marketing Research, 10, 447-453.

MacKay, David B. and Olshavsky, Richard W. (1975), "Cognitive Maps of Retail Locations: An Investigation of Some Basic Issues," Journal of Consumer Research, 2, 197-205.

Marks, Ronald B. (1976), "Operationalizing the Concept of Store Image," Journal of Retailing, 52, 37-46.

Maze, Edward M. (1974), "Determining Shopper Movement Patterns by Cognitive Maps," Journal of Retailing, 50, 43-48.

Miller, John A. (1976), "Store Satisfaction and Aspiration Theory," Journal of Retailing, 52, 65-84.

Monroe, Kent B. and Guiltiman, Joseph P. (1975), "A Path-Analytic Exploration of Retail Patronage Influences," Journal of Consumer Research, 2, 19-28.

Reilly, W. J. (1929), Methods for Study in Retail Relationships. Austin, TX: Bureau of Business Research at the University of Texas--Austin.

Stone, Gregory P. (1954), "City Shoppers And Urban Identification: Observations On The Social Psychology Of City Life," The American Journal of Sociology, 60, 36-45.

Young, William J. (1975), "Distance Decay Values And Shopping Center Size," Professional Geography, 27, 304-309.