point of origin if the trip is to be repeated. Only by coincidence will the demand for transport in both directions balance Ordinarily one direction or the other will have excess vehicle capacity that could accommodate more goods or people at an extremely low out-of-pocket cost. A rational rate-making policy will then quote lower back-haul rates in the underutilized direction That direction can sometimes change rather often-for example, in intraurban travel there is a morning inbound and an afternoon outbound rush hour, and in some instances lesser reversals around the noon hour and in the evening. On weekends there is a reverse pattern of recreational travel from and to the main urban area In this particular case, highway and bridge tolls and transit fares do not embrace the back-haul pricing principle, but they easily could, and it might be persuasively argued that they should Differentiation of charges on passenger travel according to direction is likewise not applied to intercit interregional travel within a country. We might wonder why not, in view of the frequency of the practice in commodity transport. The essential difference between people and goods in this context is that people want to return home eventually and goods do not. Accordingly, "people flows"have a natural tendency to balance out over any substantial time interval. On certain internatinnal travel routes however the seasonal imbalance of travel demand is enough to induce airlines and shipping firms to re have been at times special one-way bargain rates to entice transport that use no durable vehicles and for which there is are one-way routes for the transport of logs or for primitive end, and pipelines normally operate in similar one-way fashion ewise have no back-haul problem. Nothing is moved, so nothing sion gives some idea of the many"dimensions"in which transfer ection, specific origin and destination, quality of service, size of red. Clearly, there is some point at which detailed proliferation of omplexity, and various simplifications and groupings commend mmodities, for example, is held within bounds by assigning most ng a single schedule of rates apply to that class as a whole. The points served by a transfer system is analogously simplified by For example, rail freight rates for some commodities between Length of haul o Pittsburgh proper but to a much larger area embracing the major type is particularly prevalent when competitive pressures do not -actual costs and the prices that are charged. An illustration of the FIGURE 3-3: Typical Stepwise Progression ute distance is shown in Figure 3-3. which gives us a still more of Transfer Rates by Distance Zones d3-2 3.3.4 Time Costs in Transfer We have already indicated one way in which the time consumed in transfer is felt in costs: Both the labor and the capital sed in the transfer operation are hired on a time basis, so the labor cost and the capital cost of a trip will be less if the trip is faster. It is the high speed of aircraft, particularly jets, that enables them to transport passengers and certain kinds of freight at costs per mile comparable to those of ground transport. The capital and labor costs per hour are spread over at least ten times as many miles Quite apart from this, speed cheaper transfer for users because they bear "inventory costs"associated with th length of time that the trip takes. 2 In goods shipments, there is the cost of interest on the capital tied up in shipments in transit insurance premiums, and the risks of delay--considerations obviously more weighty when interest rates are high. Moreover many kinds of goods deteriorate so rapidly with the passage of time that it is well worth paying more for their fast delivery There are the obvious physical perishables such as fresh meat, fish, fruit, or vegetables, and also a further class of perishables such as fashion clothing, magazines, and newspapers, which lose value as they become out of date. In the transmission of information, the very word"news"suggests quick perishability, and the more quickly perishable forms of information provide a rapidly rising demand for a variety of telecommunication services Finally, in the transfer of human beings, the time of the user of the service is even more highly valued than are the rather high costs of transporting this delicate type of freight. The basis for the high valuation placed on travel time is primarily that of opportunity cost. People begrudge the time spent in traveling because they could be using that time pleasantly or profitably in some other way The value each of us imputes to the time spent on travel can vary greatly according to circumstances, length and purpose of the trip, and the characteristics of the person. Recreational travel is supposed to be a pleasure in itself. For such obligatory journeys as commuting to work, it is sometimes suggested that the commuter's hourly earnings rate while working should be applied to the travel time also. However, such a basis may well be too high. 3 In order to suggest the magnitude of time costs of human travel, let us consider the case of an individual who values his travel time at $7.50 an hour. If he travels, say, at 30 miles an hour, his time costs are 25 cents a mile, comparable to the money costs of driving a standard car. Decisions by commuters concerning the use of alternative transfer modes can easily be influenced by costs of this size 3.4 LOCATIONAL SIGNIFICANCE OF CHARACTERISTICS OF TRANSFER RATES We have seen that the structure of transfer rates departs markedly in a number of ways from the straightforward
21 point of origin if the trip is to be repeated. Only by coincidence will the demand for transport in both directions balance. Ordinarily one direction or the other will have excess vehicle capacity that could accommodate more goods or people at an extremely low out-of-pocket cost. A rational rate-making policy will then quote lower back-haul rates in the underutilized direction. That direction can sometimes change rather often—for example, in intraurban travel there is a morning inbound and an afternoon outbound rush hour, and in some instances lesser reversals around the noon hour and in the evening. On weekends there is a reverse pattern of recreational travel from and to the main urban area. In this particular case, highway and bridge tolls and transit fares do not embrace the back-haul pricing principle, but they easily could, and it might be persuasively argued that they should. Differentiation of charges on passenger travel according to direction is likewise not applied to intercity or other interregional travel within a country. We might wonder why not, in view of the frequency of the practice in commodity transport. The essential difference between people and goods in this context is that people want to return home eventually and goods do not. Accordingly, "people flows" have a natural tendency to balance out over any substantial time interval. On certain international travel routes, however, the seasonal imbalance of travel demand is enough to induce airlines and shipping firms to vary their rates seasonally according to direction, and there have been at times special one-way bargain rates to entice permanent migrants to areas considered underpopulated. Interestingly enough, there are a few kinds of goods transport that use no durable vehicles and for which there is consequently no question of back-haul rates. Some rivers are one-way routes for the transport of logs or for primitive goods-carrying rafts that are broken up at the down-stream end, and pipelines normally operate in similar one-way fashion. Telecommunications media and power transmission lines likewise have no back-haul problem. Nothing is moved, so nothing needs to be brought back. Simplification of Rate Structures. The foregoing discussion gives some idea of the many "dimensions" in which transfer rates can logically be differentiated: according to mode, direction, specific origin and destination, quality of service, size of consignment, and nature of the commodity or service transferred. Clearly, there is some point at which detailed proliferation of individual rates produces a tariff schedule of impractical complexity, and various simplifications and groupings commend themselves. The variety of rates charged for transport of different commodities, for example, is held within bounds by assigning most commodities to one of a limited number of classes and letting a single schedule of rates apply to that class as a whole. The determination of individual rates for each and every pair of points served by a transfer system is analogously simplified by grouping some of these points into zones or rate blocks. For example, rail freight rates for some commodities between Pittsburgh and other parts of the country are applied not just to Pittsburgh proper but to a much larger area embracing the major part of six contiguous counties. Rate setting behavior of this type is particularly prevalent when competitive pressures do not force a close correspondence between the transfer agency’s actual costs and the prices that are charged. An illustration of the application of the rate block principle to rates graded by route distance is shown in Figure 3-3, which gives us a still more realistic picture of rate patterns than we had in Figures 3-1 and 3-2. 3.3.4 Time Costs in Transfer We have already indicated one way in which the time consumed in transfer is felt in costs: Both the labor and the capital used in the transfer operation are hired on a time basis, so the labor cost and the capital cost of a trip will be less if the trip is faster. It is the high speed of aircraft, particularly jets, that enables them to transport passengers and certain kinds of freight at costs per mile comparable to those of ground transport. The capital and labor costs per hour are spread over at least ten times as many miles. Quite apart from this, speed means cheaper transfer for users because they bear "inventory costs" associated with the length of time that the trip takes.2 In goods shipments, there is the cost of interest on the capital tied up in shipments in transit, insurance premiums, and the risks of delay—considerations obviously more weighty when interest rates are high. Moreover, many kinds of goods deteriorate so rapidly with the passage of time that it is well worth paying more for their fast delivery. There are the obvious physical perishables such as fresh meat, fish, fruit, or vegetables, and also a further class of perishables such as fashion clothing, magazines, and newspapers, which lose value as they become out of date. In the transmission of information, the very word "news" suggests quick perishability, and the more quickly perishable forms of information provide a rapidly rising demand for a variety of telecommunication services. Finally, in the transfer of human beings, the time of the user of the service is even more highly valued than are the rather high costs of transporting this delicate type of freight. The basis for the high valuation placed on travel time is primarily that of opportunity cost. People begrudge the time spent in traveling because they could be using that time pleasantly or profitably in some other way. The value each of us imputes to the time spent on travel can vary greatly according to circumstances, length and purpose of the trip, and the characteristics of the person. Recreational travel is supposed to be a pleasure in itself. For such obligatory journeys as commuting to work, it is sometimes suggested that the commuter’s hourly earnings rate while working should be applied to the travel time also. However, such a basis may well be too high.3 In order to suggest the magnitude of time costs of human travel, let us consider the case of an individual who values his travel time at $7.50 an hour. If he travels, say, at 30 miles an hour, his time costs are 25 cents a mile, comparable to the money costs of driving a standard car. Decisions by commuters concerning the use of alternative transfer modes can easily be influenced by costs of this size. 3.4 LOCATIONAL SIGNIFICANCE OF CHARACTERISTICS OF TRANSFER RATES We have seen that the structure of transfer rates departs markedly in a number of ways from the straightforward
proportionality to distance Chapter 2. What does this mean in terms of modified 3.4.1 Effects of Limite System 3 In our initial discussic One end, one fork rets and input sources were envisaged as conflicting dimensional surface where between any two points. A e the shortest possible route System 4 a coarse route network, the No ends, no forks ong the routes. Does this B, and C, which we might nomic activity. Figure 3-4 shows four different config FIGURE 3-4: Some Possible Route Systems Connecting Three Points: A, B, and C Let us now assign ideal weights to A, B, and C. It is easy to see that if any of these ideal weights is predominant(exceeds the sum of the other two), there is no contest: That point is the optimum location so far as transfer costs are concerned, A, B, and C respecti yout. But what if the ideal weln in parenore evenly balanced, with none predominant-Say 2, 3, and 4 for ively? These are the weights shown in parentheses at the A, B, and C points on System on the left side of Figure 3-4 In System 1, we see that the optim le locations between A and B, there B and C, or 3+ ed ideal weights of by the small cir nit of output, then there will be net tra ernative location to the left of B. Sim t output (3+2 mum location has been found route shift along each of the linear d to 4, 2, and imum location, being i advantage in int ways to the dominant), we fir gths of the r route note, then, that in ne length of the quite immaterial ately avoided nal unit simply on the able from at o the marginal cost of transfer with d this marginal cost as range short range. If that of a stretched likelih the sense ocation for a unit, then, is a little lik such a s and minor and major peaks. In ive to some direct comparisor termining the ideal location of a transfer-oriented activity unit generally cannot rely entirely on gradients of transfer cost or measurements of ideal weights but
22 proportionality to distance that was assumed in our simplified discussion of individual locations in Chapter 2. What does this mean in terms of modified conclusions or new insights? 3.4.1 Effects of Limited Route Systems and Service Points In our initial discussion of transfer orientation, the economic advantages of proximity to markets and input sources were envisaged as conflicting forces, and the most profitable location appeared as the point on a two-dimensional surface where these forces just balanced. Some route networks are so dense that transfer can be effected in an almost straight path between any two points. A relatively close approximation to a uniform transfer surface is a city street system; though even here the shortest possible route and the fastest possible route may both be substantially longer than crow-flight distance. But on a coarse route network, the locational pulls toward input sources and markets are exerted in a one-dimensional way, along the routes. Does this significantly affect orientations of specific units of activity? The best way to visualize the effect is to consider a route system connecting three points, A, B, and C, which we might identify as the market and the sources for two transferable inputs for a unit of some type of economic activity. Figure 3-4 shows four different configurations that this route system might take.4 Let us now assign ideal weights to A, B, and C. It is easy to see that if any of these ideal weights is predominant (exceeds the sum of the other two), there is no contest: That point is the optimum location so far as transfer costs are concerned, regardless of route layout. But what if the ideal weights are more evenly balanced, with none predominant-say, 2, 3, and 4 for A, B, and C respectively? These are the weights shown in parentheses at the A, B, and C points on System 1 on the left side of Figure 3-4. In System 1, we see that the optimum location now turns out to be B. For all possible locations between A and B, there would be a net gain in moving toward B, since in that direction we have a pull corresponding to the combined ideal weights of B and C, or 3 + 4 =7, whereas there is a counterpull toward A of only 2. The strengths and directions of these pulls are shown by the small circled numerals with arrows attached. If the ideal weights represent, say, cents per mile per unit of output, then there will be a net transfer cost saving of 5 cents per unit of output in moving 1 mile closer to B from any alternative location to the left of B. Similarly, we find that for any location between B and C, there is a net gain of 1 cent per unit of output (3 + 2 — 4) from shifting the location 1 mile nearer B. Once we are at B, there is no incentive to shift farther; the optimum location has been found. This device of totaling the forces in each direction and thus finding the favorable direction of location shift along each route segment is a handy technique for analyzing network location in simple cases and is the conceptual basis of the linear programming approach for determining the optimum point.5 Let us now apply this procedure again to System 1 of Figure 3-4, changing the ideal weights from 2, 3, and 4 to 4, 2, and 3, as shown in the map at top right in the figure. Again we come out with the intermediate point B as the optimum location, despite the fact that it has the smallest ideal weight of the three! We begin to suspect that there is some special advantage in being in the middle; and this is, in fact, the "principle of median location," mentioned in Chapter 2. If we have three points arranged along a route as shown, and if none of their ideal weights is predominant, then the transfer orientation is always to the middle point.6 Applying the same procedure to System 2 of Figure 3-4 (and still assuming that none of the ideal weights is predominant), we find that the optimum point is the junction J. In System 3 it is A, J, or B, depending on the relative lengths of the rout e segments AB, BJ, and AJ and the ideal weights of A, B, and C. And in System 4 it could be A, B, or C. We note, then, that in every one of the four systems the optimum location is always at an intermediate point (one from which routes lead in at least two directions) and never at an end point. This holds true regardless of the ideal weights so long as none is predominant, and regardless of the length of the dead-end route segments (AB and BC in System 1; AJ, BJ, and CJ in System 2; CJ in System 3). Finally, it is quite immaterial which of the points are markets and which are input sources. In these illustrations, such identification was deliberately avoided. It is clear that when none of the ideal weights predominates, we cannot predict the orientation of a locational unit simply on the basis of its inputs and outputs; we can say, however, that it will locate not at dead ends but at points reachable from at least two directions—whether these be input sources, markets, or junctions. 3.4.2 General Locational Effect of Transfer Rates Rising Less than Proportionally with Distance Ideal weight expresses extra cost imposed per unit of added distance— in other words, the marginal cost of transfer with respect to distance. Our initial image of the relation of transfer cost to distance (Figure 3-1) showed this marginal cost as almost uniform, corresponding to a constant ideal weight regardless of distance. The more realistic transfer rate gradient in Figure 3-3, flattening off at longer distances, implies that ideal weights and the locational pulls of transfer cost factors are not constant but systematically weaker at long range and stronger at short range. If we seek a physical analogy, then, it should not be that of a weight on a string as in the Varignon Frame, nor that of a stretched spring, but that of a force more like gravitation or magnetism. This feature of transfer rates tends to enhance the advantages of location at input sources and markets and to reduce the likelihood of location at intermediate points. Each input source and market point, in fact, becomes a local optimum location, in the sense that it is better than any location in the immediately adjacent area. The search for the most profitable location for a unit, then, is a little like the search for the highest altitude in a landscape studded with hillocks and minor and major peaks. In such a landscape, we could not rely on getting to the highest point by simply continuing to walk uphill but would have to make some direct comparisons of the heights of various peaks. Analogously, a program for determining the ideal location of a transfer-oriented activity unit generally cannot rely entirely on gradients of transfer cost or measurements of ideal weights but
at Sol FIGURE 3-6: Map of Isoprofit Lines for a Transfer-Oriented Activity and market locations Profits are positive for locations within the yellow areas) This principle is illustrated graphically in Figures 3-5 and 3-6. In Figure 3-5, we have the transfer charges per unit of output as they would be at various points along a route running through the input source and the mar ket point. The two black lines show how the input transfer and output transfer charges per unit of output vary with location of the facility. The white line at the top of the figure shows total transfer charges on a unit of output plus the amount of input required to produce it It will be observed that there are local minima of total transfer charges at the input source and at the market. In this case he total costs for a location at the market would be slightly lower than for a location at the source, but both are much lower than those at surrounding locations Figure 3-6 shows a two-dimensional pattern of profits with three transfer points involved They could be, say, two input sources and a market. Here the profits per unit of output7 are shown by contour (iso-profit) lines connecting points of equal advantage. A local peak appears at each of the three points, with that at S2 the highest 3.4 3 Modal Interchange locations Ro It has been suggested above that the long-haul discount characteristic of transfer costs and rates lessens the transfer advantages of locations that are neither sources nor markets for transferable inputs and ou Some kinds of intermediate points, however, are relatively attractive in terms of transfer costs Most transfers involve one or more changes of mode or other terminal type of operation en route rather than proceeding right through from initial origin to final destination. This situation becomes more frequent as the variety of available transport modes increases, each with its special advantages for longer or shorter hauls, larger or smaller shipments, high speed, low money cost, and so on Textbooks often tell us that points of transshipment or modal interchange, such as ports, are particularly strategic locations because location of a processing facility at such a point"eliminates transshipment costs Such a statement may be misleading. Let us take a simple hypothetical case involving a flour mill. Grain is collected at an inland point connected by rail to a port(transshipment point), from which ships go to a market for flour. We want to choose among three possible locations for the mill:(1)at the grain-collection point,(2)at the port, or(3)at the market. To focus ectly on the question of the transshipment points possible advantage, we assume that the handling and transfer costs(per barrel of flour) are the same for flour as for grain, which makes the grain-collection point and the flour market equal in locational advantage. The question, then, is whether location at the transshipment point (port) is superior or inferior to the mill Let us denote the elements of cost as follows, per barrel of flour Milling cost Cost of each loading of grain or flour Cost of each unloading of grain or flour R Cost of shipping grain or flour from the collection point to the port Cost of shipping grain or flour from the port to the market The costs involved for each of the three mill locations are as itemized in Table 3-2 We notice that for each of the three possible mill locations, the total cost is the same: M+B+ W+2(L U). althoug the transshipment point location is apparently just as good as either of the others, it does not show any special advantage Indeed, we might surmise that more realistically it would be under some handicap. With either of the other two mill locations it might be possible to achieve some savings by direct transference of the grain or flour from rail to ship( the U and L operations at the port)at less cost than is involved in the two separate port transfers(grain from rail to mill, and flour from mill to ship) that are involved if the mill is located at the port. This possible saving is suggested by the square brackets in Table 3-2 If we modify the preceding case by assuming that flour is more costly to ship, unload, or load than is grain, then the most economical location is at the market; location at the grain- collection point would be less advantageous, and location at the port would be intermediate in terms of cost Clearly, then, we must explain the observed concentrations of activity at ports and other modal interchange locations on the basis of other factors. Some(the transport advantages of junction points with converging or diverging routes) have already een mentioned. A modal interchange point is likely to have such nodal characteristics, if only because different transfer modes have route networks of different degrees of fineness, so that where they come in contact, there is likely to be more than one route of the mode with the finer network The focusing of transfer routes upon points of modal interchange reflects scale economies in transfer and terminal operations, and sometimes also the lie of the land. Thus along a coastline, suitable natural harbors are limited in any event, and scale economies tend to restrict the development of major ports to an even smaller selection of points. The same applies to crossings of a mountain range or a large river A further characteristic advantage of modal interchange points is that they are likely to be better provided with specialized facilities for goods handling and storage than are most other points 3.5 SOME RECENT DEVELOPMENTS CONCERNING THE STRUCTURE OF TRANSFER COSTS 3.5.1 Introduction
23 at some stage must incorporate direct comparison of specific source and market locations. This principle is illustrated graphically in Figures 3-5 and 3-6. In Figure 3-5, we have the transfer charges per unit of output as they would be at various points along a route running through the input source and the market point. The two black lines show how the input transfer and output transfer charges per unit of output vary with location of the facility. The white line at the top of the figure shows total transfer charges on a unit of output plus the amount of input required to produce it. It will be observed that there are local minima of total transfer charges at the input source and at the market. In this case, the total costs for a location at the market would be slightly lower than for a location at the source, but both are much lower than those at surrounding locations. Figure 3-6 shows a two-dimensional pattern of profits with three transfer points involved: They could be, say, two input sources and a market. Here the profits per unit of output7 are shown by contour (iso-profit) lines connecting points of equal advantage. A local peak appears at each of the three points, with that at S2 the highest. 3.4.3 Modal Interchange Locations It has been suggested above that the long-haul discount characteristic of transfer costs and rates lessens the transfer advantages of locations that are neither sources nor markets for transferable inputs and outputs. Some kinds of intermediate points, however, are relatively attractive in terms of transfer costs. Most transfers involve one or more changes of mode or other terminal type of operation en route rather than proceeding right through from initial origin to final destination. This situation becomes more frequent as the variety of available transport modes increases, each with its special advantages for longer or shorter hauls, larger or smaller shipments, high speed, low money cost, and so on. Textbooks often tell us that points of transshipment or modal interchange, such as ports, are particularly strategic locations because location of a processing facility at such a point "eliminates transshipment costs." Such a statement may be misleading. Let us take a simple hypothetical case involving a flour mill. Grain is collected at an inland point connected by rail to a port (transshipment point), from which ships go to a market for flour. We want to choose among three possible locations for the mill: (1) at the grain-collection point, (2) at the port, or (3) at the market. To focus directly on the question of the transshipment point’s possible advantage, we assume that the handling and transfer costs (per barrel of flour) are the same for flour as for grain, which makes the grain-collection point and the flour market equal in locational advantage. The question, then, is whether location at the transshipment point (port) is superior or inferior to the grain-source and flour-market locations for the mill. Let us denote the elements of cost as follows, per barrel of flour: M Milling cost L Cost of each loading of grain or flour U Cost of each unloading of grain or flour R Cost of shipping grain or flour from the collection point to the port W Cost of shipping grain or flour from the port to the market The costs involved for each of the three mill locations are as itemized in Table 3-2. We notice that for each of the three possible mill locations, the total cost is the same: M + B + W + 2(L + U). Although the transshipment point location is apparently just as good as either of the others, it does not show any special advantage. Indeed, we might surmise that more realistically it would be under some handicap. With either of the other two mill locations, it might be possible to achieve some savings by direct transference of the grain or flour from rail to ship (the U and L operations at the port) at less cost than is involved in the two separate port transfers (grain from rail to mill, and flour from mill to ship) that are involved if the mill is located at the port. This possible saving is suggested by the square brackets in Table 3-2. If we modify the preceding case by assuming that flour is more costly to ship, unload, or load than is grain, then the most economical location is at the market; location at the grain-collection point would be less advantageous, and location at the port would be intermediate in terms of cost. Clearly, then, we must explain the observed concentrations of activity at ports and other modal interchange locations on the basis of other factors. Some (the transport advantages of junction points with converging or diverging routes) have already been mentioned. A modal interchange point is likely to have such nodal characteristics, if only because different transfer modes have route networks of different degrees of fineness, so that where they come in contact, there is likely to be more than one route of the mode with the finer network. The focusing of transfer routes upon points of modal interchange reflects scale economies in transfer and terminal operations, and sometimes also the lie of the land. Thus along a coastline, suitable natural harbors are limited in any event, and scale economies tend to restrict the development of major ports to an even smaller selection of points. The same applies to crossings of a mountain range or a large river. A further characteristic advantage of modal interchange points is that they are likely to be better provided with specialized facilities for goods handling and storage than are most other points. 3.5 SOME RECENT DEVELOPMENTS CONCERNING THE STRUCTURE OF TRANSFER COSTS 3.5.1 Introduction
The preceding sections have focused on some important aspects of the structure of transport rates and characteristics of route systems. As has been demonstrated, they provide information that can be used in conjunction with the theoretical insights gained from Chapter 2 in order to appreciate more fully the role that transfer factors may play in location decisions. In some instances, changes that take place in the markets of important commodities in a national or international context or changes in basic technological relations can have direct effects on the spatial distribution of economic activity. These effects often, but certainly not al ways, manifest themselves as a result of changes in transfer costs In this section, attention is directed to two such changes, both much in evidence at this time. We attempt to use the location principles that have been developed in order to understand some of the spatial consequences of higher energy prices anges concerning the processing and transmission of information. It should be emphasized that our treatment of issues related to these phenomena is speculative and illustrative. There is a very slim factual basis on which to location theory can help us to speculate constructively e, auge any of the effects that will be mentioned. However, it is hoped that this analysis will demonstrate how even elementary 3.5.2 Higher Energy Prices and the Pattern of Industrial Location The rapid increase in energy prices during the decade of the seventies affected our economy in many ways. We are acutely aware of the impact of this phenomenon on the rate of economic growth as well as on the distribution of income. However, little attention has been paid to the effect of higher energy prices on the spatial distribution of economic activity. It is important to recognize these spatial effects as well as the mechanics by which they are transmitted The effect of higher energy prices since the 1970s on locational choice might be considered from several perspectives. It would be possible, for example, to examine the nature of commuting or shopping behavior when people are confronted with higher motor fuel prices. Alternatively, we might recognize that higher energy pries have affected production decisions as well as the transport costs on material and finished products. This being so, our previous analysis of transfer-oriented industries would imply that, for at least some locational units, the spatial consequences of higher energy prices will depend on the nature of responses in production and the kind of changes in the structure of transport costs that take place. Much of the preceding discussion in this text has pointed to the result that the orientation of industry toward particular inputs or toward the mar ket can be influenced by these locational determinants. We are well equipped to understand many issues related to the effects of higher energy prices if we examine the systematically in this context It has been pointed out(see Figure 3-2) that intermodal competition among transfer agencies leads to a gradient of transfer rates with respect to distance that is much more curved than that of any single-mode cost gradient. For long hauls, customers will find that the decrease in transfer rates with increased distance is accentuated by competition of this sort. The locational significance of this characteristic of transfer rates is that it puts intermediate locations(places that are not markets or sources of transferable inputs) at some disadvantage One channel by which higher energy prices might affect location decisions is through their effect on the structure of intermodal transfer costs. 8 As shown in Table 3-3, transfer modes differ in their intensity of energy use. Specifically, shorter-haul transport by motor carriers(trucks) is most energy intensive, whereas rail and barge transport, which generall involve longer distances, are much more energy efficient. The most direct consequence of this is that we might expect the tapering off of transport rates with distance to become yet more accentuated as a result of higher energy prices; short-haul (truck) rates will increase relative to long-haul (rail and barge)rates. By our earlier arguments, the attractiveness of end-point locations is enhanced as a result of this effect TABLE 3.3: Domestic Intercity Freight Movement: Energy Intensity and Average Length Haul by major Transport Modes, 1979* Energy Intensity Average Length of Haul(miles) /(Btu/ton-mile) rucl 2380 Waterborne commerce r *Data on certified route air carriers are also presented in this source. They indicate that while air transport is very energy intensive(7780 Btu/ton-mile), relatively little tonnage is involved. Air carriers accounted for only 1/10 of 1% of total tonnage shipped in 1979 Source: G. Kulp, D B. Shonka, M. C. Holcomb, Transportation Energy Conservation Data Book: Edition 5(Oak Ridg Tenn: Oak Ridge National Laboratory, 1981), Table 1.13, p 1-26 The differential impact of higher energy prices on alternative modes of transport can be expected to ha subtle effects, however. Modes differ not only in their competitiveness by length of haul, but also in the kinds of commodities that they can most effectively transport. For example, not only is trucking particularly suited for the transfer of commodities over short distances, but it is also best suited to commodities that have a high ratio of value to weight and to commodities that must be shipped in small lots. 2 Both of these characteristics encourage the use of trucks to deliver finished and other highly processed goods to market. Conversely, because of the high fixed costs and relatively low line-haul costs associated with rail and barge modes, they not only have an advantage on longer hauls but also are particularly suited to the transfer of bulk commodities with low value-to-weight ratios, a category that often includes raw materials These considerations imply that the changes in relative freight rates(truck versus rail or barge) that are the result of hig her energy prices may have some significant effect on material versus market orientation. The energy intensity of truck transport will be reflected in higher line-haul rates for this mode as compared to other modes. Additionally, because of the relatively inelastic demand for transport services associated with high value-to-weight commodities, more for the energy price increases
24 The preceding sections have focused on some important aspects of the structure of transport rates and characteristics of route systems. As has been demonstrated, they provide information that can be used in conjunction with the theoretical insights gained from Chapter 2 in order to appreciate more fully the role that transfer factors may play in location decisions. In some instances, changes that take place in the markets of important commodities in a national or international context or changes in basic technological relations can have direct effects on the spatial distribution of economic activity. These effects often, but certainly not always, manifest themselves as a result of changes in transfer costs. In this section, attention is directed to two such changes, both much in evidence at this time. We attempt to use the location principles that have been developed in order to understand some of the spatial consequences of higher energy prices and technological changes concerning the processing and transmission of information. It should be emphasized that our treatment of issues related to these phenomena is speculative and illustrative. There is a very slim factual basis on which to gauge any of the effects that will be mentioned. However, it is hoped that this analysis will demonstrate how even elementary location theory can help us to speculate constructively. 3.5.2 Higher Energy Prices and the Pattern of Industrial Location The rapid increase in energy prices during the decade of the seventies affected our economy in many ways. We are acutely aware of the impact of this phenomenon on the rate of economic growth as well as on the distribution of income. However, little attention has been paid to the effect of higher energy prices on the spatial distribution of economic activity. It is important to recognize these spatial effects as well as the mechanics by which they are transmitted. The effect of higher energy prices since the 1970s on locational choice might be considered from several perspectives. It would be possible, for example, to examine the nature of commuting or shopping behavior when people are confronted with higher motor fuel prices. Alternatively, we might recognize that higher energy pries have affected production decisions as well as the transport costs on material and finished products. This being so, our previous analysis of transfer-oriented industries would imply that, for at least some locational units, the spatial consequences of higher energy prices will depend on the nature of responses in production and the kind of changes in the structure of transport costs that take place. Much of the preceding discussion in this text has pointed to the result that the orientation of industry toward particular inputs or toward the market can be influenced by these locational determinants. We are well equipped to understand many issues related to the effects of higher energy prices if we examine the systematically in this context. It has been pointed out (see Figure 3-2) that intermodal competition among transfer agencies leads to a gradient of transfer rates with respect to distance that is much more curved than that of any single-mode cost gradient. For long hauls, customers will find that the decrease in transfer rates with increased distance is accentuated by competition of this sort. The locational significance of this characteristic of transfer rates is that it puts intermediate locations (places that are not markets or sources of transferable inputs) at some disadvantage. One channel by which higher energy prices might affect location decisions is through their effect on the structure of intermodal transfer costs.8 As shown in Table 3-3, transfer modes differ in their intensity of energy use. Specifically, shorter-haul transport by motor carriers (trucks) is most energy intensive, whereas rail and barge transport, which generally involve longer distances, are much more energy efficient. The most direct consequence of this is that we might expect the tapering off of transport rates with distance to become yet more accentuated as a result of higher energy prices; short-haul (truck) rates will increase relative to long-haul (rail and barge) rates. By our earlier arguments, the attractiveness of end-point locations is enhanced as a result of this effect. TABLE 3.3: Domestic Intercity Freight Movement: Energy Intensity and Average Length Haul by Major Transport Modes, 1979* Energy Intensity / (Btu / ton-mile) Average Length of Haul (miles) Truck 2380 270 Rail 625 595 Waterborne commerce 440 770 *Data on certified route air carriers are also presented in this source. They indicate that while air transport is very energy intensive (7780 Btu / ton-mile), relatively little tonnage is involved. Air carriers accounted for only 1/10 of 1% of total tonnage shipped in 1979. Source: G. Kulp, D. B. Shonka, M. C. Holcomb, Transportation Energy Conservation Data Book: Edition 5 (Oak Ridge, Tenn.: Oak Ridge National Laboratory, 1981), Table 1.13, p. 1-26. The differential impact of higher energy prices on alternative modes of transport can be expected to have more subtle effects, however. Modes differ not only in their competitiveness by length of haul, but also in the kinds of commodities that they can most effectively transport. For example, not only is trucking particularly suited for the transfer of commodities over short distances, but it is also best suited to commodities that have a high ratio of value to weight and to commodities that must be shipped in small lots.9 Both of these characteristics encourage the use of trucks to deliver finished and other highly processed goods to market. Conversely, because of the high fixed costs and relatively low line-haul costs associated with rail and barge modes, they not only have an advantage on longer hauls but also are particularly suited to the transfer of bulk commodities with low value-to-weight ratios, a category that often includes raw materials. These considerations imply that the changes in relative freight rates (truck versus rail or barge) that are the result of higher energy prices may have some significant effect on material versus market orientation. The energy intensity of truck transport will be reflected in higher line-haul rates for this mode as compared to other modes. Additionally, because of the relatively inelastic demand for transport services associated with high value-to-weight commodities, more for the energy price increases
can be expected to be passed on by agencies serving this class of goods. Smaller portions of energy price increases will be passed along by those modes that service low value-to-weight commodities because of the sensitivity of their demand to price transport rates on finished P t e g y be neede L 0 service to firms. While ma proximity to the suppliers of the service has meant s down time, more regularity of production, an therefore lower operating costs. There is eve fewer machines are needed to ensure a given rate o production and as goods spend less time in the production proce But in recent years some highly sophisticated "smart" machines that incorporate computer systems to monitor
25 can be expected to be passed on by agencies serving this class of goods. Smaller portions of energy price increases will be passed along by those modes that service low value-to-weight commodities because of the sensitivity of their demand to price increases. Therefore, in the tug-of-war governing location decisions for industries that are sensitive to transport costs, we should expect that the pull of the market will be enhanced relative to that of transferable inputs as transport rates on finished goods increase relative to those associated with materials. We should recognize that this analysis concentrates on only one component of the "ideal weight" measures defining locational pulls. It has been argued that energy price increases will be reflected in transport rates. The other component of ideal weight is, of course, the physical weight of the transferable input or output. There are some evidence that the materials and energy are substitute inputs in the production process associated with U.S. manufacturing as a whole.10 This would imply that an increase in energy prices may increase the weight of materials transferred for output of a given weight. Such a change would tend to increase the ideal weight of materials and may serve to counteract any tendency toward market orientation due to changes in relative transport rates. The highly aggregative nature of empirical evidence concerning this matter pr ecludes any definitive judgment, however. Higher domestic energy prices not only affect transport and production costs, they also imply substantial shifts in the spatial distribution of income. Energy-producing regions have gained for at least two reasons. Greater local production at higher prices obviously has meant greater income to workers as well as to the owners of capital in these regions. 11 Further, while price controls on domestic petroleum and natural gas production are being phased out, the presence of these restrictions has meant at least a short-run advantage to energy consumers in energy-producing regions. They have faced relatively lower energy prices than they would in regions that must rely exclusively on higher-priced, imported energy. Therefore, recognition of the concept of "market access potential" developed in Chapter 2 would indicate that for some locational units higher energy prices mean that the median location of the market will shift in the direction of those regions with substantial existing or developing capacity in energy production.12 While we have been able to identify certain gross tendencies that may be manifest as a result of higher energy prices, this analysis is only suggestive of the kind of forces at work. Individuals who are concerned with the behavior of specific industries could obtain more detailed information on transport modes and on the character of production and markets that are relevant to their interests. They might then be in a position to know whether transport rate, production, or market considerations will be most influential. 3.5.3 Technological Change in Data Processing and Transmission In contrast to the behavior of energy prices, the cost of moving and processing information has fallen dramatically in recent years, and the end is not in sight. Advances in electronics technology have abruptly enhanced the efficiency of computers and our ability to interact with them. At the same time, developments in communications technology have weakened the constraints of distance on some types of location decisions. Significant locational effects are emerging on both the microspatial and the macrospatial levels, foreshadowing still further shifts. As we shall see in Chapter 7, the internal spatial arrangements of urban areas are shaped largely by considerations of access—it might even be said that access is what cities are all about. At this microspatial level, the journey to work and one’s ability to maintain close, flexible contact with customers, suppliers, co-workers, and friends are major determinants of both business and residence location. So if people or firms find that their work and other activities no longer demand close physical contact, locational incentives will change. For example, it is now becoming increasingly practicable to use computer hookups to communicate with other workers or with central data banks. As a result, the valuation of locations with respect to their nearness to long-established foci of urban economic activity is changing considerably. This "communications revolution" has potentially wide implications. Some people have speculated that the "cottage industry" of the near future will comprise people who work at home and maintain business contacts via integrated computation and communications systems. Early evidence of such a trend is already appearing. For some activities, the very nature of outputs or inputs, or both, may change as a result of advances of the sort just mentioned. Banking is an obvious case in point. From one perspective, deposits received by a bank may be regarded as inputs; banks then take those inputs and use them to earn income by "selling" loans and other investments and services. Alternatively, one might view the receiving of deposits as a form of services provided by the bank and thus as one of the bank’s outputs. Until recently, the deposit activity of a bank was essentially non-transferable, and many separate banking offices were needed to service adequately a large urban area full of depositors and borrowers. But the deposit services of a bank may soon become very transferable indeed. We see already more and more banking machines acting as robot tellers; banking by phone is developing, and banking via home computer is in the offing. So depositors who now have to travel to a bank, or use the mails, will soon find that the bank’s services travel instantly to them. With the proliferation of electronic transactions and home computer terminals, we can foresee that the customer service area of a single banking office will no longer be confined to a neighborhood, and that presumably far fewer bank locations will be needed. Locational relations among different activities likewise are subject to important alteration when the transferability of information is greatly enhanced, as is now happening. An example of this is firms that provide troubleshooting and repair service to users of complex equipment. The easy and quick availability of such services has been an important factor to many firms. While maintenance specialists can be dispatched some distance to attend to problems, speed is of the essence. Close proximity to the suppliers of the service has meant speedy attention, less down time, more regularity of production, and therefore lower operating costs. There is even a saving in capital costs as fewer machines are needed to ensure a given rate of production and as goods spend less time in the production process. But in recent years some highly sophisticated "smart" machines that incorporate computer systems to monitor