Plants of different sizes can be built in the long run, so a unique set of short run cost curves exists for each possible plant size. Possible changes in the short run marginal costs and average total costs of a garment manufacturer with (a) 100, (b) 200, or (c) 300 sewing machines are highlighted in Figure 7. Under these short run cost curves, we have placed an envelope curve, which reflects the plant sizes associated with the average costs of producing each level of output. This envelope is the long run average cost (LRATC) curve for the firm.

 

A long run average total cost (LRATC) curve reflects the plant size that allows the minimum possible short run average costs to produce each possible level of output.

 

            Notice that this envelope curve is not tangent to the minimum point on each SRATC curve. Only a plant in which 200 machines are used to produce 4,000 garments per day (point a) yields the absolute minimum long run average cost of production. At this point, the envelope curve is tangent to the minimum point of the SRATC curve. To the left of this point all short run cost curves are tangent to the LRATC curve on the left side of their respective minimums. To the right of point a, tangencies with the envelope curve are at the right sides of the short run curves.

 

Measuring Long Run Average Costs

Any firm that fails to exploit economies of scale will have higher average costs than those of competing firms that do; firms that are too small for efficient operation must either grow or fail. Many people think that bigger firms can almost always produce at lower costs than smaller firms. While it is true that a firm must be large enough to exploit all feasible economies of scale, bigger plants may encounter diseconomies of scale and be forced to reduce the scope of their operations or sink.

 

            Studies of Portugal's decline as a world power from 1400 to 1600 indicate that its wooden sailing ships were too large for then-current technology. A large part of the Portuguese fleet sank in bad weather, in part because of huge cargos. In the 1970s, the U.K./French Concorde passenger jets suffered such great cost disadvantages that their government sponsors took financial baths. How large is the optimal convenience store or gas station? How about atomic power plants or oil refineries? In the 1930s, the world's largest auto assembly plant was Ford's River Rouge plant. It was never fully utilized, and today much of it has been torn down. It was simply too large to be efficient.

 

            The ranges where economies or diseconomies of scale are actually encountered vary substantially among industries. Engineering estimates and the few statistical studies of cost functions that are available indicate that there typically are substantial ranges of output for which average costs are roughly constant, as depicted in the middle of the LRATC curve in Figure 8.

    An idea known as the survival principle suggests that clustering within an industry of firms or plants of a particular size is conclusive evidence about the efficient scale of operations. Some economists have tried to apply this principle to specific industries as a way of measuring the minimum points of long run average cost curves. Critics, however, argue that survival depends on a multitude of factors (luck, monopoly power, business acumen, growth or decline of an industry, and so on); therefore, some inefficient firms may survive, while some efficient firms fail.

Minimum efficient scale (MES) plants are the smallest that will produce output at minimum average total cost.

Minimum efficient scale (point a at the beginning of the flat portion in Figure 8) has been estimated for various industries using accounting data, engineering estimates, and the survival technique. Typically, MES is reported as a percent of the total market. Figure 9 presents some estimates of MES for selected industries here and abroad. Measuring long run cost curves is unavoidably imprecise, but the concept is still useful in analyzing industry adjustments to changes in demands, resource prices, or other events.

 

            In summary, economies of scale encourage size. If the minimum efficient scale of production in an industry requires huge firms, then fewer firms will inhabit that industry. (This raises questions about government policies to control excessive market power---issues that are addressed in later chapters.) On the other hand, significant diseconomies of scale tend to reward compact firms so that many competitors inhabit an industry.

 

            Firms and industries grow in response to widespread perceptions of profit opportunities, or wither when economic losses are expected to persist. Long run adjustments allow firms to enter an industry and grow infinitely, or to shrink to zero and leave an industry; a firm can perfectly adjust its size by purchasing more or fewer resources, but, by assumption, we hold technology constant. We need to consider the possibility, however, that research and development may respond strongly to profit opportunities.