Does Economic Space Matter?

The chapters in this book are presented in honour of our esteemed colleague and friend, Mel Greenhut. Professor Greenhut has been a leading figure (see George Norman’s account, Chapter 2 in this volume) in the development of modern spatial microeconomics, on which this book is focused. However, we honour him in this book not only for his leadership in this field of specialization, but also in respect and thanks for his important contributions to the study of economics in general.

The purpose of this compact essay is to explore, within a specific time-frame, two related issues regarding spatial economics. The first issue is whether and to what extent elements of space mattered to economists of the past. The second issue is why did a particular interest in spatial economics, as measured by an individual’s ‘peak performance’, surface when it did. By asking the questions why and when, we raise a more basic issue about the relative import of endogenous and exogenous influences on economic theory.

Spatial economics has as its remit the study of the use of a finite resource — space. It takes explicit account of the twin facts that economic activities both consume space and are separated by distance. Two principal sets of questions are addressed: how economic agents of various types choose their locations in a spatially extensive economy, and how the market areas of these agents are determined.

If lands are available to all, people under push-shove equilibrium will allocate themselves among different acres to equalize average rather than marginal labour productivities. By contrast, if the land is owned by numerous decentralized proprietors, the auction rents they charge will generally increase total social output by shifting workers from places of lower marginal productivity towards those of higher mar­ginal productivity. (See H. Scott Gordon, 1954; Paul Samuelson, 1974, 1990; Martin Weitzman and Jon Cohen, 1975.)

Geographical space matters for economics in at least two fundamental ways. First, space is a factor of production and a consumption good. Second, it separates economic agents, thereby imposing constraints on their interaction. For example, an urban economic model contains both: residential land is a good while urban transportation must be used to overcome the friction of distance. A third way in which space matters arises in the context of public economics. There, introducing geographical space generates interaction between private and local public sectors which can be affected by the way political jurisdictions are drawn. For example, in economies with public goods, the location of a political boundary determines the very nature of spillovers and thus affects public policy. At a more basic level, drawing political boundaries over a resource-diversified territory can determine jurisdictions with unequal resource endowments and production capabilities. It follows that a political boundary creates public wealth inequalities which may affect the distribution of welfare over the land. For example, the recent split of the former Soviet Union into smaller political entities will sharpen wealth inequalities. In the opposite direction, Western Europe has moved a long way towards removing political boundaries.

The Japanese are the great children lovers, ever since the era of Manyoh (700s AD) when the poet Okura Yamanoue declared that silver, gold and precious stones were nothing compared to his beloved children. If the wife is the master of the household, the children as its core are the kings of a typical Japanese family. Nevertheless the present Japan may be characterized by gross underproduction of children per family. One of the many reasons cited casually is poor housing conditions with cramped spaces often referred to as ‘rabbit hutches’. It goes without saying that the greater are the costs of anything, the less is the quantity demanded, and supplied accord-ingly, in the market. The higher cost of housing, however, does not ipso facto imply underproduction of it, nor does it imply the underproduction of children as an output(!?) of the housing input.

Many commentators criticize centralized government production of public goods for not providing sufficient variety. Citizens with widely varying tastes all consume the same type and level of services. With fixed costs of production and variations in tastes and incomes, an efficient solution, familiar in the public finance literature, can be generated by a single government producer that supplies the public good and charges citizens different tax prices to take account of the variations in their tastes and incomes.¹ This solution, while efficient, is not feasible. Citizens have, in general, no incentive to reveal their willingness to pay, and recent advances in demand-revealing processes have not produced entirely satisfactory solutions.² In addition, if citizens’ tastes diverge sharply enough and if production costs are low enough, efficient solutions may also exist with several suppliers each producing a different variety of output for a subset of citizens.^3,4

Location rent arises from differential access to pubic services and to privately supplied goods and services. Under some conditions, the net value provided by a service is equal to the increase in location rent that results from its provision. This chapter discusses these conditions and the implications of departures from them. If the conditions were generally met, one might propose that public services be evaluated and financed by the increases in rent that they generate. However, the conditions are generally not met. To accommodate this fact, the chapter develops a method of evaluating and financing public services that combines rent information with decisions made by the demand-revealing process. While any application of the demand-revealing process determines whether a proposed service satisfies a benefit-cost criterion, the incorporation of rent collection into the financing increases the likelihood of finding an acceptable way of providing an efficient service when a cost is attached to unintended redistribution. The chapter also discusses the ways in which the conclusions apply to privately provided goods and services.

Does space matter in economics? Of course it does. Mel Greenhut devoted the better part of his career to the analysis of economic space and greatly improved our understanding of the subject. The important issue is how space matters, and we still have much to learn about that issue.

The central business district (CBD) of big cities like Tokyo is constantly suffering from congestion phenomenon for the consumption of services generated by the urban infrastructures (transport, communication, water, sewage, and so on). A typical example is dangerously crowded peak-time commuting trips into the CBD of Tokyo.

It is, of course, an honour and, even more, a personal pleasure to be among those invited to participate in this Festschrift volume for Mel Greenhut. Many years of pleasant encounters, both direct and through the mail, as well as admiration for his contributions, made this invitation an opportunity not to be missed. Yet, having done no work in location theory, the main arena of Professor Greenhut’s research interest, I am forced by lack of competence to turn instead to a subject that can be interpreted only with some degree of generosity to overlap or at least to border this area.

The negative impact of oil supply shocks has led those countries that are both producers and net importers of crude oil to consider policies that would increase the degree to which such countries are self-sufficient in crude oil. It is understood by all that such self-sufficiency comes at a price. Either internal production must be subsidized or imported crude oil must be taxed, both of which lead to a reduction in welfare. Those arguing in favour of these insulation policies claim that the costs incurred during the periods of supply shocks when a country is at the free market level of self-sufficiency outweigh the welfare cost of whatever programme is required to reduce exposure to external shocks. A prominent policy designed to increase crude oil self-sufficiency is a fee applied to imported crude oil. In this chapter we examine the welfare implications of import fees when used as a device to increase the degree that a country is self-sufficient in any critical raw material. The analysis applies to the current energy debate but its generality extends to any raw material of a country that, at current world prices, is a net importer of that raw material.

The birth of industrial location theory is generally dated 1909, when Weber published his book entitled Uber den Standort der Industrien. This theory is later extended by Isard (1956), Moses (1958), Sakashita (1967), Bradfield (1971), Emerson (1973), Woodward (1973), Khalili, Mathur and Bodenhorn (1974), Miller and Jensen (1978), Mathur (1979), Mai (1981, 1984), Eswaran, Kanemoto and Ryan (1981), Shieh and Mai (1984), Hurter and Martinich (1989) and others. All these papers consider location in a heterogeneous cost space and define the optimum location in terms of the search for cost minimization or profit maximization under conditions where the demand factor is held constant and where locational interdependence of firms is disregarded.¹

Following the seminal work of Hotelling (1929), many researchers have studied firms’ strategic choices over price and/or location within spatial markets. These studies on spatial competition are not mere extensions of non-spatial theories but add a new aspect of their own: the introduction of space relaxes price competition between firms because firms scattered over the geographical space acquire some degree of monopoly power. Thus price competition does not necessarily yield a Bertrand outcome (marginal cost pricing).

M. L. Greenhut has long maintained that ‘One policy area which not only lends itself to [spatial price] theory, but actually requires consideration of spatial relationships is that of antitrust’ (Greenhut and Benson, 1989, p. 3). Market delineation is a fundamental issue in antitrust practice (Marcus, 1980, p. 294) which cannot be fully appreciated without direct application of spatial price theory (Greenhut and Benson, 1989, p. 3).¹ This is a required first step in the application of the US Clayton Act merger section, in Sherman Act monopoly cases, and under Canada’s Competition Act, where market share and market concentration data are offered as evidence of market power. Thus, a large literature has developed on the subject of market delineation (surprisingly little of it draws from spatial price theory, however, the exceptions being Benson, 1980; Greenhut and Benson, 1989; Benson and Faminow, 1990; and Benson et al., 1991)² Indeed, a distinction between the ‘economic market’ and the ‘antitrust market’ has been drawn (Scheffman and Spiller, 1987). The economic market in which a firm operates encompasses all of the supply and demand forces that influence that firm’s price, while an antitrust market presumably consists of a ‘group of producers and geographic area, that could, if cartelized, profitably exercise market power’ (Scheffman and Spiller, 1987, p. 126).

Firms typically invest in productive capacity before output is earmarked to particular markets. This paper studies the sequential decision choice of production and allocation to spatially separated markets. The resulting equilibrium will generally differ from equilibrium with simultaneous choice. Whether total output is higher or lower in the sequential model depends on an allocation-shifting effect: firms wish to rearrange their rivals’ allocations out of the most profitable market (desire-to-shift effect), and changes in capacity are a strategic device enabling them to do this (ability-to-shift effect). If firms have both the desire and the ability, there will be allocation shifting. We examine allocation shifting for a duopoly serving two spatially separated markets. Numerical simulations for a cross-hauling model show that the allocation-shifting effect can be substantial. We also consider a case with firms at either end of a linear segment over which markets are evenly distributed. It is shown that the direction of allocation shifting depends on the convexity or concavity of demand: for linear demand the two-stage equilibrium is identical to that of the simultaneous game.

Today in North America commercial dairies typically transport raw milk from the dairy farmers to their processing plants, while fifty years ago it was the dairy farmers who transported the raw milk. In the salmon fishery in the Pacific north-west, fishers sometimes transport their catch to the fish processor, and processors sometimes buy the fish on the fishing grounds, transporting the catch to their own processing plants. On the Canadian prairies, wheat farmers usually transport the wheat they grow to the grain elevator. In most labour markets, the seller accepts responsibility for getting to work, but this is not universally the case; harvest labour, for example, is often transported to the work site by the buyer. And in some labour markets the responsibility is shared — the company provides the car and the petrol, and the employee the time needed to get to and from work.

Let us begin with a naive story about an imaginary country named Webertopia. This country is situated along a very narrow valley-corridor h kilometres in length, which makes the geographical territory of Webertopia appear as if it is a horizontal line. At one end of the valley-corridor exists city A which sits continguous to the boundary of country AA, while city B sits at the other end of the corridor contiguous to country BB. In city A, a licensed factory M _A produces the fixed amount q _z of commodity z every day for export to country AA. In city B, we have a licensed factory M _B producing daily the amount q _z of z for export to country BB. The conditions of the licence issued to M _A require that the commodity z produced by M _A should be sold at a fixed unit price p _z to an importer I _A from country AA who comes to city A every day to purchase the product. The similar conditions must be met by factory M _B with respect to importer I _B from country BB. The average cost to produce commodity z is c₁ for both factories when they are operated separately. If the two factories were to go into joint production at a single location within the national boundary of Webertopia, the average production cost would turn out to be c₂.

M. L. Greenhut has been claiming for many years now that economic space inherently involves oligopolistic relationships among firms. In turn, he has maintained that contrary to conventional economic thought, oligopolistic competition can yield a long-run equilibrium that is unique, want-satisfying, and efficiently stable; in fact, he claims it contains features isomorphic to those of perfect competition. This is a surprising argument which has to this time drawn insufficient attention from the professional circle.