again the choice of location will depend in part on economies of scale. brought to light some of the less obvious complexities of the problem of measuring the ng the choice of location for a specific business establishment or other unit. It n be assigned to the various factors only in certain cases(to be argued that the relative influence of the various factors e geographical patterns of variation of the peaks rise above face look very egularities. i itern is quite choosing whethe metropol ation(say, within a erent regio aggregation. The choice is a , or among countries y the level of comparison outputs. After all, th only really comparison o hood), we must recognize t r, electric energy, tr ling requires travel to th to h d at any to the location rger than on are generally ey may even be the ide). Many the sense of equall that the postal example unit is locate utputs that are not transferable b nined b size of place; but there are Some activities cater to local markets and cannot operate at a minimum efficient scale except in places of at least a certain minimum size. In selecting a8 again the choice of location will depend in part on economies of scale. The foregoing discussion has brought to light some of the less obvious complexities of the problem of measuring the relative importance of the various factors affecting the choice of location for a specific business establishment or other unit. It should now be clear that definite quantitative "weights" can be assigned to the various factors only in certain cases (to be discussed later in this chapter) involving transfer cost. It has also been argued that the relative influence of the various factors upon location depends on the amounts and kinds of inputs and outputs and on the geographical patterns of variation of the respective input supplies and output demands. 2.4 SPATIAL PATTERNS OF DIFFERENTIAL ADVANTAGE IN SPECIFIC LOCATION FACTORS If one views the earth's surface from space, it looks completely smooth—after all, the highest mountain peaks rise above sea level by only about 1/13 of 1 percent of the planet's radius. A closer view makes many parts of the earth's surface look very rough indeed. Again, if one looks at a table-top, it appears smooth, but a microscope will disclose mountainous irregularities. The same principle applies to spatial differentials in a location factor: The interregional (macrogeographic) pattern is quite different from the local (microgeographic) pattern. For example, we should not expect land cost to be relevant in choosing whether to locate in Ohio or in Minnesota; but if the choice is narrowed down to alternative sites within a particular metropolitan area, land cost will indeed be important. Large differences may appear even within one city block. Labor supply and climate, in contrast, are examples of location factors where there is little microgeographic variation (say, within a single county or metropolitan area), but wide differences prevail on a macrogeographic scale involving different regions. Locational alternatives and choices are generally posed in terms of some specific level of spatial disaggregation. The choice is among sites in a neighborhood, among neighborhoods in an urban area, among urban areas, among regions, or among countries. No useful statements about location factors, preferences, or patterns can be made until we first specify the level of comparison or the "grain" of the pattern we are concerned with. This principle was in fact implicit in our earlier distinction between local and transferable inputs and outputs. After all, the only really non-transferable inputs are natural resources or land, including topography and climate. In a very fine-grained comparison of locational advantages (say, the selection of a site for a residence or retail store within a neighborhood), we must recognize that all other inputs and all outputs are really transferred, though perhaps only for short distances. Water, electric energy, trash, and sewage all require transfer to or from the specific site. Selling one's labor or acquiring schooling requires travel to the work place or school; selling goods at a retail store requires travel by customers. Accordingly, our distinction between local and transferable inputs is a flexible one: It will vary according to how microgeographic or macrogeographic a view of location we are taking for the situation at hand. Thus if we are concerned with choices of location among cities, "local" means not transferable between cities. Some inputs or outputs properly regarded as local in such a context are properly regarded as transferable between sites or neighborhoods within a city. What, then, are the possible kinds of spatial differential patterns for a location factor as among various locations at any prescribed level of geographic detail? The simplest pattern, of course, is uniformity: All the locations being compared rate equally with respect to the location factor in question. For example, utility services are commonly provided at uniform rates over service areas far larger than neighborhoods, often encompassing whole cities or counties. Wage rates in an organized industry or occupation are generally uniform throughout the district of a particular union local, and in industries using national labor bargaining they may even be uniform all over the country. Tax rates are in general uniform over the whole jurisdiction of the governmental unit levying the tax (for example, city property taxes throughout a city, state taxes throughout a state, and national taxes nationwide). Many commodities are sold at a uniform delivered price over large areas or even over the whole country. Climate may be, for all practical purposes, the same over considerable areas. The special term ubiquity is applied to inputs that are available in whatever quantity necessary at the same price at all locations under consideration. Air is a ubiquity, if we are indifferent about its quality. Federal tax stamps for tobacco or alcohol are a ubiquity over the entire country. If an input is ubiquitous, then its supply cannot be a location factor—being equally available everywhere, it has no influence on location preferences. The demand-side counterpart of a ubiquity is of course an output for which there is the same demand (in the sense of equally good access to markets) at all locations under consideration. There does not seem to be any special technical term for this, and it is in fact a much rarer case than that of an input ubiquity. Perhaps we could illustrate it. Imagine some type of business that distributes its product by letter mail, but with speedy delivery not being a consideration. In such a case, proximity to customers is inconsequential; demand for the output is in effect ubiquitous. The reason, in this special case, is that the postal service makes no extra charge for additional miles of transportation of letters. A different pattern of advantage for a location factor can be illustrated by market access for wheat growers. The demand for their wheat is perfectly elastic, and what they receive per bushel is the price set at a key market, such as Chicago, minus the handling and transportation Charges. The net price they receive will vary geographically along a rather smooth gradient reflecting distance from Chicago. The locational effect of the output demand factor can be envisaged as a continuous economic pull in the direction of Chicago. Similar pull effects reflecting access advantage operate within individual urbanized areas. For example, workers' residence preferences are affected by the factor of time and cost of commutation to places of employment. Another kind of systematic pattern involves differential advantage according to the size of the town or city in which the unit is located. This might apply to certain location factors involving the supply of or the demand for inputs or outputs that are not transferable between cities. It would be surprising to find any kind of differential advantage that is precisely determined by size of place; but there are many location factors that in fact show roughly this kind of pattern. Some activities cater to local markets and cannot operate at a minimum efficient scale except in places of at least a certain minimum size. In selecting a