PRIME COSTS AND OUTPUT
The statement commonly made that average prime costs in a steel- works tend to rise as output falls (or in other words that marginal are below average prime costs) is not true in all circumstances. This note attempts--rather tentatively--to show within what limits it is true.
There are, of course, two familiar factors which should tend to raise average prime costs as the output of a works rises. Not all workmen are of equal efficiency; and it sometimes happens that of the parallel units at one stage of production (e.g. the blast-furnaces) some are less efficient than others. Obviously if the plant has to be worked below capacity, other things being equal, the management will be disposed to lay off the least efficient workers and units. Differences of efficiency between parallel units--to take the second factor first--are only likely to occur in an old plant which has been expanded; they are more likely among furnaces than mills (since mills doing the same job side by side are rare); and they will certainly not always be the determining factor in selecting a furnace for stopping; more often perhaps the condition of the furnace linings is likely to be decisive. Hence the second factor can only be spasmodically effective, and not in all works. If anything the first factor seems even less significant. Its scope is narrowed by the extent to which wages are paid by results; 70 or 80 per cent of the workers have at least part of their wages on this basis. There is, too, a strong tendency to "ca-canny" in a works when work is short.1 Moreover the effect of variations in workers' efficiencies must be spasmodic, for most of the work is done by large groups of complementary workers, and it is only when a whole group is taken on or off that changes of output can change the make-up of the labour force much. In the____________________