Systematic differences are found in the average earnings obtained in different regions, industries, and occupations. The average earnings prevailing in different regions of a country show a considerable range between the highest and the lowest, even when the same procedures for fixing rates apply everywhere. Much of the dispersion is due to differences in the localization of industry: if the relatively high-paying iron and steel industry, for instance, is concentrated in a particular region, then the average pay of the region will be raised to that extent. But regional differences also exist because work of the same kind frequently commands different rates of pay in different regions. Such differences are sometimes necessary to maintain the balance of payments between regions; they may also be in some measure a legacy of history and are likely to be reduced as communications improve and labour becomes more mobile. As noted above, this process of reduction has been expedited by trade union pressures.

Average earnings also vary from industry to industry, and the considerable range that appears can again be attributed largely to differences in the composition of the labour force: an industry such as printing, which by the nature of its processes employs a high proportion of skilled workers, will on that account alone show higher average earnings than, say, the textile industry, which employs a higher proportion of the semiskilled. The similarity between the structure of earnings by industry in different countries—with printing, iron and steel, and engineering near the top, and textiles and food processing low down—is thus attributable to common processes requiring similar compositions of the labour force.

This occupational structure, therefore, presents the main object of economic analysis. International comparisons show that the ranking order of the rates of pay prevailing in different occupations is similar in different countries, but that the range, whether between professional and manual occupations or, within the manual, between the skilled and unskilled, is much wider in economies at an early stage of development and diminishes in the course of development. These are the principal observations to be accounted for by any theory of the differences in the rates of pay that different kinds of work command. Several such theories have been propounded.

One theory stresses the link between occupations and their status in the community, some having higher status than others. The community believes, according to this theory, that pay should correspond to status; and the rate of pay for each occupation is assigned to it by common consent, reflecting the place it occupies in the hierarchy of esteem. The implications are that the community’s discretion is not as a rule subject to other factors such as the market forces of supply and demand and that, if people came to make less distinction of status between occupations, then the rates of pay for different jobs could be more nearly equal. Doubtless, many people do think in the way the theory supposes, feeling it anomalous, for example, when an occupation that is commonly accorded a higher status than another ceases to command a higher rate of pay. But the correspondence between status and pay is ambiguous; it is not clear to what extent pay is made to fit the status and to what extent status follows from pay. Moreover, the occupational pay structures of different countries show more similarity than do their social values. Nor does the theory explain why the pay structure has generally become compressed in the course of development, and the ranking order sometimes inverted, as when some clerical occupations drop below some manual ones—changes that can otherwise be explained as the effect of extended education in increasing the relative supply of more qualified labour.

If the theory is not acceptable as an explanation of the pay structure as a whole, it does call attention to a factor that appears to affect parts of that structure. One of these parts is that of the higher administrative posts. It is generally accepted that any such post must carry a higher salary than any post below it in the chain of command; and when this chain is long, as it is in a big corporation, the salaries set for the posts in it, and the high level reached at the top, are to be accounted for by this principle.

The same theory also suggests a cause of prevailing differences between men’s and women’s rates of pay. Some women’s work is different in kind from men’s irrespective of the fact that it is done by women; and, where men and women both do work of the same description, some disabilities attaching to women as employees, in particular the likelihood that they will not stay in the job as long as men, may make them worth less to the employer. There are many jobs, however, in which these considerations do not apply and in which there is no difference in the productivity of men and women adequate to account for the actual difference between men’s and women’s rates. The difference seems attributable rather to customary attitudes and valuations—in particular, the assumption that women’s productivity is lower in all jobs and the belief that pay should be proportioned to need (women workers generally needing less than the man who has a family to support). If such factors as these account for differences in pay between men and women where there is no corresponding difference in the work they do or the efficiency with which they do it, one may speculate that the same factors account for some part of the pay differential where the two kinds of job are distinct.

A second theory lays its stress on power: the ways in which organized groups can protect and advance the pay of their members. Any group that restricts entry into its occupation can keep its labour relatively scarce and thereby support the rate of pay that that kind of labour commands. The discussion above of the economic effects of the trade union indicated the circumstances in which a trade union would be able to raise the relative pay of its members by the exercise of monopoly and bargaining power. The general increase in the pay of the less skilled relative to that of the skilled manual worker has been attributed in part at least to the increased unionization of the unskilled. Evidently, policies of organized groups will account for some part at least of the position of particular occupations in the pay structure. Power can also be endowed by management in the form of managerial authority, and at least a part of a hierarchy of pay within an organization is likely to be an imposed means of distinguishing and upholding the strata of the organization’s internal power structure.

A third theory treats the differences in pay for different jobs as corresponding to differences in their content or requirements. The simplest form of this theory was embodied in the labour theory of value, whether in the system of Adam Smith or of Karl Marx, by the assumption that different kinds of labour can be reduced to different quantities of “homogeneous labour time,” and that rates of pay are then simply proportional to those quantities. Job evaluation, discussed above, purports to condense the varied requirements of each job to a single figure in a common scale in order that the ranking order of the rates of pay of the jobs may be brought into conformity with that of those figures. But the assumption that, if two articles are priced in the same currency, they must contain quantities of a common substance is gratuitous. The impossibility of establishing the existence of such a substance and measuring the amount of it in any article drives both the labour theory of value and job evaluation into the circular argument of inferring the job content from the rate of pay and then explaining the rate of pay by the job content.