The notion of returns to scale - and the related notion of division of labor - are among the oldest and most familiar insights in economics, going back at least to Adam Smith. It is not necessary to expound them in detail again. However, it is helpful (for what follows) to remind the reader that the classic 'supply-demand' intersection presupposes a rising supply curve, for the economy as a whole, or for a particular good or service. This implies that the marginal cost of production (of the good or service) rises monotonically with output. A rising marginal cost curve reflects short-term rigidities, namely, a fixed workforce, a fixed physical capital and a fixed technology.
At first sight this equilibrium picture seems incompatible with economies of scale, a point that properly bothers many thoughtful first-year economics students. However,
* This chapter is reprinted in full from Chapter 15 of Robert U. Ayres and Leslie W. Ayres (1996), Industrial Ecology: Towards Closing the Materials Cycle, Cheltenham, UK and Brookfield, US: Edward Elgar. With permission of the publishers.
economies of scale become meaningful when we relax the assumption of a fixed set of factors of production at a static point in time. Absent the (implied) condition of short-term rigidity, it is easier to see that alternative technologies of production can, and do, exist, in principle, at different scales of operation. They are likely to differ in many ways, especially in terms of capital/labor ratio. At larger scales of operation, and especially with longer production runs, more operations can be carried out by machines or other equipment than at small scales. The physical reasons why this is true range from the obvious (longer runs = less set-up time) to the arcane (decreasing surface/volume ratios) and need not be considered further here. The important consequence is that the number of workers (or man-hours) per unit output tends to decline with increasing scale, ceteris paribus.
This fact, in turn, has had a major impact on economic growth in the past. Passing from a static to a dynamic framework, consider what happens when existing productive capacity in an industry is augmented by a new and more efficient plant. The new plant can produce more cheaply. Potential supply in the industry has increased. Demand will absorb the larger supply only at a lower price. If competing suppliers cut prices, demand will continue to rise. This will stimulate further investment in supply, and so on. The cycle of price cuts (permitted by economies of scale) leading to increasing demand (thanks to price elasticity of demand) has been called the 'Salter cycle'.
To maximize economies of scale, manufacturers in the late 19th and early 20th centuries adopted a strategy of product standardization and 'mass production'. These were basically US innovations. Successful exemplars ranged from Waltham watches, Colt 45 revolvers, Remington rifles, Yale locks and Singer sewing machines, to Ford's successful 'Model T' (which was produced continuously from 1908 to 1926). At that time, the keys to success in manufacturing were product standardization, division of labor and volume (Ayres 1991b). Frederick Taylor incorporated these elements, together with some others, into a formal theory of 'scientific management' which strongly influenced (and was influenced by) Henry Ford.
The benefits of scale are diluted in the case of mechanical or electrical products which are evolving and improving over time. Mechanical automation requires large investments in specialized machinery and equipment, which must be depreciated. This tends to discourage technological change, since new models require new and costly production lines to be designed and custom-built (see, for example, Abernathy 1978). The benefits of scale are most obvious (and easiest to analyze econometrically) in the case of homogeneous commodities such as steel, petrochemicals or electric power. Economies of scale tend to encourage industrial gigantism, and oligopolies, at the expense of competition. Economists have argued the relative benefits to consumers of scale economies v. competition in regulated utilities, such as telecommunications, electric power, water and gas distribution, railroads or airlines (see, for example, Christensen and Greene 1976). Evidently, oligopolistic pricing inhibits growth, but it appears that economies of scale were still an important engine of economic growth even for the USA, and much more so for Europe and Japan, in the post-war decades (see, for example, Denison 1962, 1974, 1979). Scale economies were perhaps the only significant growth factor for the Soviet Union and Eastern Europe during that period.
Recently, there have been indications that economies of scale are no longer as important as they once were. Markets for many standardized products have become saturated, at least in the West. Quantity of supply (of final goods and services) is less and less important relative to quality and variety. From the classical Taylorist-Fordist point of view, variety (diversity) of output is incompatible with maximum efficiency. Yet there are other, hitherto neglected, dimensions of production technology and other strategies for cost reduction that are more appropriate for meeting demand in markets where diversity - even 'customization' - is inherently valuable. Newer strategies maximizing 'returns to scope' (or 'economies of scope') have become increasingly important in recent years, thanks to the introduction of new computer-based technologies in manufacturing. In brief, the idea is that a manufacturer who can produce a large number of different products efficiently from a small number of flexible workers or programmable machines will be more able to meet variable demand than a manufacturer with inflexible machines geared to a single standardized product. It appears that advanced forms of computer-controlled production, linked with computer-assisted design and engineering - known as computer-integrated manufacturing, or CIM - offer a feasible path away from traditional mass production (see Goldhar and Jelinek 1983, 1985; also Ayres 1991a).
Another hitherto neglected dimension of strategy (with a few exceptions) is to maximize systems integration. We consider this strategy in more detail next.
Was this article helpful?