Answers are grounded in this paper and cite their sources.
Economists discuss differences in capital intensity or labor intensity very often, but they consider differences in space intensity only relatively rarely. Yet, differences in the spatial intensity of economic activities are fully as important. Indeed, I will argue here that the main features that distinguish farming from manufacturing and most other urban economic activity are due to differences in space intensity. (For some parts of my argument, it might seem more familiar to use the phrase "land intensity" rather than "space intensity." But other parts of the argument hinge on transportation, communication, and surveillance costs that depend on the distance between points, whether or not the land between these points is farmed. Thus the phrase "land intensity" would not be generally appropriate in this paper.) The difference in space intensity between most types of farming and most types of urban industry is obviously colossal. In the United States, farming accounts for decidedly less than 5% of the national income, but farming, at least if we define it broadly enough to include ranching and forestry, uses decidedly more than 95% of the area that is used by the entire economy. Though there are a few types of farming, such as huge feeding lot operations, that are not space intensive, and also some types of urban industry that are fairly space intensive, these are clearly the exceptions. It would be commonplace today to find farms of, say, more than 500 acres, that would utilize only one or two or three workers and produce a total output valued at a few hundred thousand dollars. It would also be easy to find factories that occupy less than five acres yet employ several thousands of workers and produce a total output valued in the tens or hundreds of millions of dollars. Thus, farming activities easily can be many thousands of times as space intensive as many types of urban industry. The generalization that most farming is vastly more space intensive than most urban industry appears to hold for all countries and historical periods in which meaningful urbanrural comparisions can be made. We cannot explain some of the most important economic realities without taking account of these differences in space intensity. And when we do take these differences into account, a number of important puzzles can be solved. One of these puzzles has been attracting attention for quite some time. For example, in 1945 T. W. Schultz observed, in Agriculture in an Unstable Economy, that agriculture and industry tended to react quite differently to general economic fluctuations. Schultz noted that "the prices of farm products reflect quickly a bullish or bearish turn" (p. 131), but the prices of nonfarm products generally do not. Conversely, the quantity of agricultural output in the aggregate "responds slowly, or not at all, to the changes in economic expectations associated with business fluctuations" (p. 131), whereas the "output of industry is highly responsive to changes in economic expectations, with production rising sharply when the outlook turns bright and falling abruptly when it becomes gloomy" (p. 131). This observation was not, so far as I know, very controversial, for many other economists had observed exactly the same thing, and the pattern in the data is quite clear. The data Schultz presented showed that industrial production in the United States fell by nearly half between 1929 and 1932, whereas agricultural production fell by little more than 3%; over the same period the prices received by farmers fell by more than one-half, whereas the prices farmers paid fell by only about a fifth (Schultz 1945, tables 15 and 16). The statistics probably understate the reduction in industrial prices to some extent because of secret discounting and possible changes in Mancur Olson is a professor of economics, University of Maryland.
No takes yet. Share an insight, caveat, or question.
Mancur Olson (1985) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: