measurement, and the question of the appropriate period over which to compute secular change, the productivity standard appeared to offer a framework within which wage movements could be judged in terms of their impact on costs and prices. The argument, simply stated, is approximately as follows. If monetary and fiscal policy, given reasonably full employment, can prevent pressure on prices from the side of demand, then the annual gain in real output per hour can be realized in the form of higher money wages, salaries, and profits without a secular rise in the general level of prices. This will be possible if, from the wage side, the average gain in money wages conforms to the secular rise in productivity, and provided also that price decreases in high productivity industries, where unit costs are falling, roughly balance price increases in low productivity industries.

In a more or less explicit form, the elements of a productivity-based wage and price policy may be found in the Economic Reports of the President throughout the postwar period. In the 1947 midyear Economic Report, for example, note was made (p. 3) of certain special circumstances justifying wage increases. Aside from these special circumstances, the report continued, "wage increases should be related to general trends in productivity and not made on a basis which forces price increases or prevents price reductions needed to assure sales of increasing supplies." The 1953 Report of the Council of Economic Advisers (p. 114) suggests that "since wages constitute the bulk of personal income, the preferable general formula-once wages, prices, and profits are in a workable relationship-is for money wages to increase with productivity trends in the whole economy." The 1957 Economic Report (pp. iii-iv) pointed out that business managements must "recognize the broad public interest in the prices set on their products and services." Both management and labor were urged "to reach agreements on wages and other labor benefits that are consistent with productivity prospects and with the maintenance of a stable dollar." The 1960 Economic Report (p. iv) argued that "the achievement of inflationfree economic growth. . . requires a blending of suitable private actions and public policies."

22 William Fellner and Others, The Problem of Rising Prices (Paris, Organi zation for European Economic Cooperation, 1961).

The Report (p. 8) pointed out that wage increases "are the major means in our free economy by which labor shares in the fruits of industrial progress. But improvements in compensation rates must, on the average, remain within the limits of general productivity gains if reasonable stability of prices is to be achieved and maintained."

Impetus to the more explicit formulation of wage-price policy in terms of the productivity criterion may have been given by the publication in the spring of 1961, under the auspices of the Organization for European Economic Cooperation (OEEC), of a study of the problem of inflation by six distinguished economists.22 The report contended that, "from an economic standpoint, wage increases can be considered an independent causal factor of price increases only when they are determined by negotiations between organized labor groups and employers (or raised by government decree). In such circumstances, the average increase in wages is not necessarily that which would have resulted in a freely competitive market."23 For the period under review (1953-60), the study concluded that "excessive wage increase constituted both an important and independent inflationary force" in the United States, the United Kingdom, and certain other countries.24 On these points, there was agreement among the six authors.

On the question of a solution to the problem, however, the group split. The majority position was that governments should adopt a wage policy, the elements of which were stated as follows:

By having a wages policy, we mean, first of all, that the authorities themselves must have a reasonably precise view, estimated by the best means which they can devise, of the average increase in wages that is appropriate to the economic situation and consistent with stability of the price level. This view will necessarily depend primarily on experience and expectations regarding the longer run rate of productivity in the economy. But it may also have to take into account various aspects of the current economic situation, such as the phase of the trade cycle and the state of the balance of payments. Having such a view, getting it known by the interested parties, and mobilizing support for it as an objective towards which to work, is the essence of having a wages policy."

With respect to the implementation of policy, the majority view at first seemed unexceptional, with statements to the effect that the problem was one involving the art of the National Government in reconciling divergent economic interests; that recognition of the problem and the enunciation of a wage policy would in itself constitute an

enormous step forward; and that the establishment of a norm for wage increases should create a situation in which the burden of proof would be upon any labor or management group that wished to deviate from it. But beyond these homiletic prescriptions, the majority suggested an institutional role for government in the following words:

As the aim of wages policy is to set a norm for the increase in wages, there must be a center of authority in the public services charged with this function and machinery for appropriate consultation with representatives of the broad parties at interest. Furthermore, it might be of benefit to arrange for the governmental authorities to be represented at important wage negotiations-not simply as mediators but as a party at interest to represent the general public's stake in the outcome. It is not reasonable that this function should be neglected, or at best left vaguely to the press and public opinion.26

The majority also urged that governments develop wage policy for their own employees, especially to avoid wage and salary lags that subsequently must be corrected by sharp upward adjustments. Finally, it was suggested, without specific recommendations, that the machinery for wage negotiation in the private sector of the economy might need review and change.

Clearly the guidelines set forth in the 1962 Economic Report of the President, and summarized in the first section of this article, meet the criterion of a wage policy as defined in the OEEC study. Notably in the basic steel situation, the governmental authorities made the intent of this policy clear to the parties at interest,27 and it appears to be the intention of the Government to continue to do so in other situations.28 What follows represents an effort briefly to appraise certain aspects of the policy and the circumstances under which it must operate.

Aspects of Wage Policy

1. A Federal Reserve Board study in 1946 concluded: "The fact is that collective bargaining with strong unions, price stability, and full employment are incompatible. We can have any two of these but not all three. So long as union power is not dampened down by unemployment, there is no apparent power in the State strong enough to check a parallel upward sweep of wages and prices." 29 The basic aim of wage policy is to resolve this dilemma by asserting, in a variety of ways, a public interest in the magnitude of

wage settlements. In essence, the aim of policy is to stabilize the level of efficiency wages, not the level of money wages or real wages. It should be observed that the Federal Reserve Board conclusion was drawn at the height of the postwar demand inflation. Conditions are now considerably different. Judgment must be reserved with respect to the "power of the State," under present-day economic circumstances, to influence the terms of wage bargains.

2. The achievement of a stable price level would prevent the erosion of real income among those groups (e.g., annuitants) whose incomes are fixed in money terms; it would mean, however, that such groups would be precluded from sharing in the gains from technical progress. This latter result could be avoided only if the aim of policy were a long-term decline in the price level (perhaps creeping deflation), which would involve a decline in efficiency wages and other costs. Such a policy has its advocates.30 But for both technical and institutional reasons, price level stability rather than decline appears to represent a more feasible goal of policy. It should be recognized, however, that such a policy implies some shift in income (in both money and real terms) in the direction of those groups, primarily wage and salaried workers, equity shareholders, and independent entrepreneurs, whose incomes have upward flexibility.

3. A wage policy of the type here under discussion would not alter the relative share of wages and salaries in national income. Both labor and nonlabor income would grow, but at the same relative rates. It is sometimes thought that labor would obtain the entire gain in output under

20 Ibid., p. 59.

...we

27 At his press conference on April 18, 1962, the President stated: worked very closely with the steel union in an attempt to persuade them that it was in their interest and the country's interest to meet the standards set by the Council of Economic Advisers, and it was done. That is why this matter [of steel prices] came particularly in sharp focus last week."

28 The fact must be emphasized, however, that general responsibility for wage decisions will continue to rest with the parties. In his address to the convention of the United Automobile Workers, May 8, 1962, the President declared: "We are neither able nor willing to substitute our judgment for the judgment of those who sit at the local bargaining tables across the country. We can suggest guidelines for the economy, but we cannot fix a single pattern for every plant and every industry."

29 Charles O. Hardy, "Prospects of Inflation in the Transition Period," in Prices, Wages, and Employment (Washington Board of Governors, Federal Reserve System, 1946), p. 24.

For example, see Council on Prices, Productivity, and Incomes First Report (London, H.M. Stationery Office, 1958), pp. 32-33. The council reported that "we have been impressed by the apparently widespread revival of interest in the idea of falling prices" and raised the question of whether "a gently falling price level" should not be considered as a goal of policy.

a productivity-based wage policy, but this would not be the case. If wages, salaries, and supplementary compensation represented 70 percent of costs in the economy, then roughly 70 percent of the gain in man-hour output would accrue to labor in the form of higher wages and benefits. The remaining 30 percent would augment nonlabor income. There is, indeed, little evidence to suggest that collective bargaining is a powerful instrument for altering income shares in a direction favorable to the working class as a whole. A broad scale effort in this direction would almost inevitably produce a round of price increases, as business sought to protect its income position. Conversely, if business, in response to a favorable demand situation, sought to improve its income share through general upward price adjustment, a reaction from the side of wages would undoubtedly occur. A wage-price or pricewage spiral is essentially a struggle over income distribution.

4. A wage policy involving an announced permissible average annual increase in wages might have an inflationary effect of its own. Thus, the British Council on Prices, Productivity, and Incomes has suggested that there would be "a real danger that the prescribed average would always become a minimum, and the process of wage inflation therefore [would be] built into the system."31 In Machlup's model of the mechanism of cost inflation, this would occur through "responsive" reactions to "aggressive" wage increases in some sectors of the economy. The guideline approach does tend to reenforce expectations of annual wage advances that have grown up during the postwar period. On the other hand, even in the immediate postwar years, when demand inflation was at its height, there was by no means uniformity of response in terms of wage

31 Ibid., p. 45.

32 See BLS press release, "Increases Granted in Basic Rates Since V-J Day," September 20, 1946.

"An interesting analysis of the nature of power in the labor market may be found in J. Pen, The Wage Rate Under Collective Bargaining (Cambridge, Mass., Harvard University Press, 1959), especially pp. 107-112.

At its 1961 convention, the AFL-CIO, in its resolution on collective bargaining, decried "efforts to bind bargaining by national formulas, by ceilings, and other constraints." The resolution continued, in part: "The 'public interest' is the announced justification for such efforts. The public interest is multifaceted. The overall and longrun public interest cannot be served in a democratic society by throttling genuine and unshackled collective bargaining. Responsible management, we believe, will resent any alteration of the fundamentally free character of collective bargaining because the disadvantages of such change would far outweigh any seeming or short-run advantages of 'orderliness' or 'industrial peace,'” See AFL-CIO Resolutions Committee, Report No. 2, pp. 65-67.

settlements, 32 and the notion of pervasive "patterns" of adjustment, except within the orbits of certain industries or employments, has little current substance. Conditions affecting particular firms and industries should continue to play an important role in differentiating wage settlements.

5. Nevertheless, the preceding consideration does lead to one of the critical questions relating to the guideline type of wage policy. A policy prescribing that annual wage increases, on the average, shall not exceed the long-term gain in man-hour output inevitably means that some increases should fall short, and others should exceed, the average. The principal reasons for divergence from the average are clearly indicated in the statement of the Council of Economic Advisers. In industries or employments where labor is in excess supply, the increase should fall below the average or perhaps, in a given year, be skipped altogether. But given a union capable of exercising market power, 33 how is this result to be achieved? Would not such a union strive to attain at least the average increase sanctioned by the wage policy? There is, after all, no central power on the trade union side that might assist in the allocation of the permissible annual wage increase among constituent unions. Indeed, the practice and tradition of autonomy in the American trade union. movement is so strong that relinquishment by individual unions of control over collective bargaining policy and tactics would represent an extraordinary development. It is perfectly natural that any organized group should consider itself the best interpreter of the needs of its constituency.34

This attitude does not mean, of course, that unions or employers can overlook the existence of a wage policy or of governmental efforts to implement it. These factors add new dimensions to the bargaining process. They may well have to be taken into account, at least in key situations, in the formulation and rationalization of demands, and in bargaining strategy. The crucial question, in fact, revolves about the role of government, and it was fundamentally this issue that gave rise to dissent to the majority OEEC position on wage policy. The minority among the six OEEC economists stated their position as follows:

The majority of the group assumes that governmental guidelines (or "norms") relating to the supportable average

rate of money wage increases could contribute importantly to the solution of the problem. We have at best moderate hopes in this regard. Hence we believe that the rather ineffective steps, against which we have no particular misgivings as long as governments do not go beyond them, would have to be followed by more specific measures if governments were to make a real effort to put into effect a national wages policy. Wage negotiations and pricing practices relate not to average wage rates and prices but to wages and prices in specific industries. Where wage-push inflation tends to develop, governments committed to a national wages policy would, in our opinion, have to become involved in rather extensive regulations relating to individual wages and prices. In such circumstances, political considerations and the usual group pressures would exert a very significant influence on the wage and price structure and also on the general wage and price level. We are opposed to moving in this

direction.35

It is, of course, much too early to gage the extent to which Government in the United States might have to intervene in the labor market (or the product market) to make a guideline wage (and price) policy effective. If, again to use Machlup's terminology, potentially "aggressive" wage actions were confined to a limited number of sectors in the economy, then the power of Government admonition and public opinion might produce the desired result. Even in such circumstances, however, the question arises of whether, with time, admonition would lose its effectiveness.36

6. Finally, there is the intractable fact of the labor market, that immensely complex mechanism through which the wage level is determined by thousands upon thousands of collective and individual bargains and employer personnel actions. On the institutional side, there are in our system few bargains that have a national impact, in the sense that they tend to determine wages and conditions of work for the employees of entire industries. We operate, indeed, under a comparatively decentralized system in which the locus of bargaining for most industries and employments is the individual plant or company, associations of similar employers within particular labor markets, or, more rarely, industry associations on a broader geographic basis. For large sectors of employment, including some in which wages and salaries are advancing rapidly, compensation is not determined through formal collective bargaining procedures at all. This general system has many advantages. For one thing, it permits account to be taken of the economic circumstances of individual firms and of the interests of particular

groups of workers; it tends to encourage novelty and innovation in the terms of employment.

However, the system may not be well adapted to the application of a national wage policy of the guideline type. To use some obvious examples: Average annual employment in the contract construction industry, where bargaining is predominantly on a local basis, is more than four times as great as employment in basic steel where, in essence, national bargaining occurs; employment in food stores, a single segment of retail trade, is almost twice as great as in the manufacture of automobiles and parts; and a union such as the International Association of Machinists has bargaining agreements in numerous industries, ranging from automobile repair to aircraft-missiles, but its bargaining is primarily with individual companies or local employer associations. Some bargains admittedly are vastly more important than others; it may be argued that these "key" bargains set the tone for wage determination in broad areas of the economy and that the crucial problem is to keep these in line with policy. But the relationship between the "key" bargains and the tens of thousands of "nonkey" bargains may not be as great as one might suppose, and the cumulative effect of the latter may be substantial.

Underlying the institutional mechanisms for wage determination are demand-supply conditions in the labor market at any given time and over periods of time. These fundamental factors help powerfully to determine the pattern of wage change; in many situations, the impulses of the market are simply transmitted through the collective bargaining process. It is bound to be difficult in many cases to judge the extent to which, if at all, unions have exerted market power, in the sense that settlements are greater than they would otherwise have been.

Within the manufacturing sector, a great deal of information is available on the structure of wage settlements during 1960. In that year, which was one of recession, approximately 11.4 million factory workers were employed in establishments with a policy of making general wage

35 William Fellner and Others, op. cit., p. 63.

6 The additional Council of Economic Advisers guideline relating to wage in creases above or below the average, depending on the level of wage rates in relation to bargaining strength, is not considered here. It would obviously be difficult, if not impossible, to implement in the absence of effective Government decisions as to the appropriateness of rate levels in particular situations. It might also run counter to the criteria relating to labor supply.

adjustments.37 General wage increases were received by about four-fifths (9.1 million) of these workers, with the increases ranging from less than 1 percent to 10 percent or more. Increases of 3 percent or more were received by three-fourths of the workers for whom upward adjustments were made. The average (median) increase, including in the calculation those who received no increases, was between 3 and 3.5 percent. Fringe benefit improvements were made for a substantial proportion of the workers whose wage rates were advanced. However, it is not possible to tell how the distribution of general wage increases would have been altered if increases in expenditures on fringe benefits could have been included. with the wage rate changes.38

These data throw some light on the pervasiveness of wage pressures under conditions of less than full employment.3 But the main point for this discussion is the diversity of the adjustments suggested by the data. Clearly there would be immense difficulties in external judgment of the validity of these adjustments within the framework of a wage policy. Adjustments of, say, 5 or 6 percent may be completely justifiable on labor market grounds in some instances, and adjustments of 2 percent not justifiable in others. Put differently, an increase of 6 percent in certain circumstances, may simply reflect labor market pressures rather than union power. It may be observed, by way of illustration, that increases in the average annual salaries of college and university teachers, who are largely untouched

27 The remaining workers, about 1.3 million, were in establishments where pay adjustments were made on an individual rather than on a general basis. "See "Wage Developments in Manufacturing During 1960,❞ Monthly Labor Review, August 1961, pp. 846-850.

Of the 9.1 million factory workers obtaining general wage increases in 1960, 7.4 million were in union establishments and 1.7 million in nonunion establishments. Most of the 1.1 million union workers who did not receive an adjustment were in establishments in which there was either no bargaining on wages or bargaining was not concluded in 1960. About 1.2 million workers in nonunion establishments with general wage change policies did not receive an increase in 1960.

40 Salaries Paid and Salary Practices in Universities, Colleges, and Junior Colleges, 1961-62 (Washington, D.C., National Education Association, 1962), p. 25, table 22.

"Statement of the Chancellor of the Exchequer in the House of Commons, July 25, 1961. The pay pause ended on March 31, 1962.

42 Incomes Policy: The Next Step (London, H. M. Stationery Office, Cmnd. 1626, February 1962), p. 4.

43 The 20-member council consists of representatives of the government, private and nationalized industry, and the trade unions, together with several independent experts.

"It should be noted that the question of wage policy is being considered, together with other questions, by the President's Advisory Committee on Labor-Management Policy (established by Executive Order No. 10918, Feb. 16, 1961).

by unionism, have equaled or exceeded 6 percent in each of the past 6 academic years.40 These increases may well reflect excess demand for academic personnel; they do not reflect institutional pressure in the labor market.

Conclusion

The United States is by no means alone in grappling with the problem of peacetime wage policy in the context of policies relating to full employment, price stability, and economic growth. In Great Britain, for example, the Government in July 1961 called for a "pay pause" to permit productivity to catch up with incomes and to "mark the beginning of a new long-term policy."41 A subsequent White Paper declared that the "objective must be to keep the rate of increase of incomes within the long-term rate of growth of national production."42 A National Economic Development Council, with a permanent staff, has been established to provide guidance on national economic development.43 Even West Germany, with its postwar miracle of economic growth, has not escaped the problem. In a speech on March 21, 1962, Minister of Economics Erhard proposed setting up a council, which he thought should be "more scientific and technical than administrative," to advise industry on the adjustment of wages to costs and prices.

This article has attempted to indicate the circumstances giving rise to the quest for a wage policy, the nature of the policy that appears to be emerging, and some of its implications and difficulties. The problem is real, the objectives of policy are important, and the implementation of any type of policy is immensely difficult. The objective, in brief, is to keep the rise of national money income, including wages and salaries as its major component, in line with the rise in national output. Real incomes will thus advance within a framework of price stability. To achieve this end without sacrificing the great advantages of freedom in the labor market (or, on the price side, in the product market) is one of the challenges of our time. The aim is for responsibility and restraint within the field of private decisionmaking, and not for state planning of private actions. The more deeply and widely the nature of the problem is understood, the more likely we are to work out a viable approach.“