THE LARGE AND PROLONGED INFLATION which has been experienced in many countries has led to the introduction of various techniques designed to protect the incomes of some sectors of the economy from the adverse effects of inflation. Thus, in many countries wages have been linked to prices, so that a rise in the cost of living necessitates an offsetting rise in wages—at least, if the rise in the cost of living is more than some critical figure.1 A fixed tie between wages and prices must have significant effects in any economy. For a free economy, in which economic adjustments are induced by changes in prices and wages, the imposition of the degree of rigidity implied by this association is of far-reaching importance. It can be justified, if at all, only by the need to deal with a critical situation, such as that presented by a large and prolonged inflation.


THE LARGE AND PROLONGED INFLATION which has been experienced in many countries has led to the introduction of various techniques designed to protect the incomes of some sectors of the economy from the adverse effects of inflation. Thus, in many countries wages have been linked to prices, so that a rise in the cost of living necessitates an offsetting rise in wages—at least, if the rise in the cost of living is more than some critical figure.1 A fixed tie between wages and prices must have significant effects in any economy. For a free economy, in which economic adjustments are induced by changes in prices and wages, the imposition of the degree of rigidity implied by this association is of far-reaching importance. It can be justified, if at all, only by the need to deal with a critical situation, such as that presented by a large and prolonged inflation.

THE LARGE AND PROLONGED INFLATION which has been experienced in many countries has led to the introduction of various techniques designed to protect the incomes of some sectors of the economy from the adverse effects of inflation. Thus, in many countries wages have been linked to prices, so that a rise in the cost of living necessitates an offsetting rise in wages—at least, if the rise in the cost of living is more than some critical figure.1 A fixed tie between wages and prices must have significant effects in any economy. For a free economy, in which economic adjustments are induced by changes in prices and wages, the imposition of the degree of rigidity implied by this association is of far-reaching importance. It can be justified, if at all, only by the need to deal with a critical situation, such as that presented by a large and prolonged inflation.

Although the basic causes of inflation vary widely in different countries and at different times, the process of inflation always shows strikingly similar characteristics. Essentially, it is caused by the excessive expectations of government, business, or labor in the use of the national product and is associated with the excessive creation of credit. If the causes of inflation are not eliminated, the mere linking of wages to prices will not of itself bring the inflation to a halt. On the other hand, if the causes of inflation are eliminated, the linking of wages to prices need not prevent the restoration of stability, provided the wage base which is thus linked is not too high in relation to the productivity of the economy. As a practical matter, in a number of countries the use of wage-price links is a consequence of the fear of labor that real wages will be adversely affected by inflation.

It cannot be emphasized too strongly that inflation is a monetary phenomenon. While the level of prices will rise when there is increased demand, even if the money supply is restricted, the rise must in that case be limited, and it must in a short time come to an end. A large and prolonged inflation is possible only if the monetary authorities permit an excessive and continuous expansion of credit. Whatever the initiating factors may be, they can act on the price level only if they are made effective by the provision of additional credit, although they may be intensified by an accompanying increase in the velocity of money. For this reason, any measures taken to halt inflation must include the restriction of credit. Devices such as the linking of wages to prices are of significance only as part of the environment for making monetary policy. This is the proper framework for any discussion of the economic aspects of wage-price links in a large and prolonged inflation.

Excessive Demand and Rising Prices

In an economy which is functioning properly, the distribution and use of the gross national product should result in an aggregate demand for goods and services that tends to equal the available supply of goods and services at approximately stable prices. At times, however, the government, business, or labor may attempt to secure more goods and services than would ordinarily be available to them from the distribution of current output. In a relatively fully employed economy, such an attempt must result in a rise in prices. The rise in prices would, however, be quite limited, if expectations of increased investment and consumption, private or public, were not supported by the creation of credit by the monetary system.

Assume, for example, that the government wants to use more resources for public purposes than it can acquire through tax revenues, loans from the public, and part of the normal growth in credit money.2 If the government is insistent on securing additional resources, it will be able to get them by borrowing from the central bank and commercial banks, or by issuing its own currency. Similarly, if business wants to invest more than it can acquire through retained business profits, new securities purchased by the public, and part of the normal growth in credit money, it will be able to do so if additional funds are supplied by the banking system. The continued creation of credit on an excessive scale will have far-reaching economic effects—including a rise in prices.

The effect on prices of a given increase in credit, say, to finance additional investment, will not be the same for all countries, nor will it be the same in any given country at different times. If there is a substantial pool of unemployed resources, as in a period of cyclical recession, the increase in demand will manifest itself to a large extent in an increase in output and to a small extent in a rise in prices. The typical experience in the United States is that over the whole period of a cyclical expansion the increase in industrial production is of the order of two to three times the rise in wholesale prices. Of course, in the latter part of a cyclical expansion, the increase in production tends to be much smaller relative to the rise in prices. In fact, the expansion takes on the characteristics of inflation, a functionless rise in prices. In a large and prolonged inflation, there is not likely to be any substantial pool of unused resources, so that a continued increase in demand will soon result in excessive aggregate demand and will manifest itself either in a rise in prices or in a balance of payments deficit.

In a country in which foreign trade is large relative to the national product, a substantial proportion of the excessive demand will be directed toward import and export goods. The remainder will be spent on domestic goods and will induce a rise in their prices. As prices rise, however, there will be a growing tendency to substitute import and export goods for domestic goods. If the amount of excessive demand (in real terms) does not become progressively larger, a stage will be reached where all of the excess in aggregate demand will be directed toward import goods, supplemented by some absorption of export goods. At such a stage, the entire effect of the excessive demand will be manifested as a deficit in the balance of payments; and although home prices have risen, they may rise no more.3

The extent to which excessive demand can be absorbed by an increase in imports and a decrease in exports, rather than by a rise in prices, depends on the willingness and the ability of the monetary authorities to finance a balance of payments deficit. In a large and prolonged inflation, it is impossible for the balance of payments to absorb all of the excessive demand. Sooner or later, in order to protect the monetary reserves, the monetary authorities must restrict the flow of imports into an inflated economy. At such a stage, the excessive demand can be manifested only by a rise in home prices—including the prices of the restricted import goods. It is this case that is of special interest for analyzing the price effects of excessive demand resulting from a large expansion of bank credit for business or the government.

Suppose that inflation is initiated by the demand of business for additional resources for investment. Suppose further that the banking system supplies credit to finance additional investment. Prices will rise and real income will be shifted from those with a higher propensity to consume to those with a lower propensity to consume. The rise in prices will also limit the increase in investment that business expected to undertake. The rise in prices must be sufficient to induce a net increase in savings equal to the actual increase in investment. For this reason, the price rise will be smaller, the greater the rigidity of money incomes—i.e., the greater the ease with which real income can be shifted to profits from other sectors of the economy.

If wages do not respond to higher prices, the rise in prices will be relatively small—not much more than in proportion to the excess aggregate demand. In fact, however, labor has become very sensitive to changes in the cost of living, and even a moderate rise in prices now leads to expectations of offsetting increases in wages. Nevertheless, there is always some sector that remains, at least for a time, unresponsive to the loss of real income involved in rising prices. Pensioners, bondholders, and other recipients of fixed incomes usually cannot or do not secure an adjustment to offset fully the rise in prices. The larger the group whose money incomes are unresponsive, inadequately responsive, or delayed in their response to the rise in prices, the less the price effects of any given excess demand will be.

When the excess demand is large and continues for an extended time, its price effects become relatively greater. The unresponsive sector of the economy is gradually narrowed because the real incomes of the fixed income group have been squeezed to a minimum and new savings no longer go into fixed-income assets. People with wages and salaries which are fixed by custom may accept a small and once-for-all reduction in their real incomes; they will not, however, concur in a more or less permanent fall in their standard of living. As inflation continues, labor expects prompt action in adjusting wages to the rise in prices, and it is no longer possible to transfer a significant amount of real income from an inflation-penalized to an inflation-favored sector. If the excess demand is allowed to continue, inflation will degenerate into a hyperinflation in which prices and wages rise rapidly in an attempt to force additional savings out of an economy determined to protect its consumption.

The Price-Wage Spiral

Even when excess aggregate demand arises from an increase in investment (or a budget deficit), labor occupies a key role in the inflation process. As prices rise, real wages fall. If labor is aware of this reduction in real wages, it may insist on an equivalent rise in money wages. Business will be in a position to meet the wage demands because prices and profits are higher. But the rise in wages will necessitate some reduction in investment below the excessive level to which it is being pressed, unless business can meet the claims for higher wages through an expansion of bank credit. If the banking system is able and willing to provide additional funds to finance the excessive investment, there will be an unresolved struggle between labor and business over the distribution of the national output between consumption and investment. Each rise in prices will be followed by a rise in wages, and each rise in wages will be followed by a new rise in prices.

After a time, the rise in prices is certain to be accelerated. Labor becomes increasingly alert to protect its interests, and the period between successive wage adjustments is gradually shortened. In order to carry out the additional investment, in the interval of the wage lag, business will require a much larger expansion of credit. The price effects of the excessive investment will become relatively larger. In time, wage demands begin to take account of the loss in real income resulting from the lag between the rise in prices and the adjustment of wages. Under such conditions, the demand for wage increases will begin to exceed the amount of the price rise. The economy will be confronted with massive wage adjustments of 20, 30, or 40 per cent at a time, applicable to almost the entire labor force.

Even if a rise in wages is economically justified and is financed by a shift of money income from profits to wages, it can be undertaken without inducing a rise in consumer prices only if the supply of consumer goods is increased to the same extent as the increase in consumer expenditure. This may involve no great difficulty with a moderate wage increase—say, not more than 5 per cent in the course of a year. If the rise in wages is accompanied by an equivalent reduction of the expenditure of other sectors, and if allowance is made for additional savings by the wage earners, it will be possible to meet the increased demand by a diversion of consumption goods from nonwage earners to wage earners, by drawing down inventories, and by changes in the pattern of imports. In the longer run, apart from the effects of productivity on supply, the restoration of real wages to an appropriate level will require some shift of productive resources to consumer goods industries. In contrast to this adjustment that an economy can make to a moderate wage increase, there is no way by which the supply of goods to be consumed by wage earners can be increased rapidly to match the increased consumption expenditure that accompanies a massive increase in wages.

When wages and prices both rise rapidly, there is a statistical illusion that tends to exaggerate the real change in the standard of living of workers. In fact, during a large and prolonged inflation, short-period changes in the real income of labor cannot be measured through a comparison of price and wage movements. This is especially true in countries in which the custom of granting massive wage increases prevails. For example, suppose that a 35 per cent increase in wage rates is put into effect by decree. It may be assumed that wage payments rise promptly thereafter. Consumer prices, however, will not rise correspondingly at once. Instead, they will rise gradually over a period of 12 months or even longer. Obviously, the actual consumption of the wage earners could not rise by 35 per cent immediately, nor is it reasonable to believe that consumption later will decline by 25 per cent when there has been a further rise in consumer prices.4

Apparently, during a period of rising prices the consumption of wage earners is maintained by drawing down cash balances, by running into debt, and by using durable goods purchased during a previous period of relatively high wages. After a massive wage increase, the higher wages are used to restore cash balances, to repay debt, and to buy durable consumer goods to be used for a considerable period in the future. The actual level of consumption varies much less than the level of wages relative to prices. The massive wage increases are probably part of a pattern of wage payments that make aggregate wages over a period of time reasonably well suited to the economic value of the work. This is not a satisfactory system of wage payments from the point of view either of the workers or of the employers. While it is a device that enables them to make a tolerable adjustment to the disruptions imposed by a chronic inflation, reliance on massive wage increases to offset the effects of inflation on real wages adds to the difficulty of restoring monetary stability.

No doubt, if wage earners would consent to their real wages being reduced during a period of inflation, the rise in prices resulting from excessive demand would be more moderate. As a practical matter, however, it is too much to expect wage earners to bear the whole cost of an investment or budgetary inflation. If the monetary authorities do not succeed in halting the inflation, the workers will, sooner or later, insist on wage increases to offset the rise in prices. Nor should it be assumed that the tendency for wage increases to follow the rise in prices is entirely the result of a forceful trade union movement determined to protect the interests of the workers. Generally speaking, during an inflation the demand for labor exceeds the supply at going rates of wages. The inflated profits enable producers to bid up wage rates, and those most determined to expand will take the lead in offering higher wages.

Wage-Induced Inflation

An inflation may have its origin in excessive wage expectations rather than in excessive expenditure by the government or excessive investment by business. A strong trade union movement may demand a higher level of wages than can be paid by business if the current level of employment is to be maintained. Employers may feel compelled to meet the wage demands, particularly if the government indicates that social difficulties might arise from strikes and layoffs. To maintain the same level of production and employment at higher money costs, business then needs additional working capital. If the banks are willing to expand credit, and the monetary authorities provide the means to enable them to do so, employers may raise wages, mark up their prices, and maintain the prevailing level of employment. The lag in consumer prices will, for the time being, result in higher real wages; but as consumer prices rise and resume their normal relation to wholesale prices, the expectations of labor for higher real wages will be defeated. If workers are insistent on higher real wages, the repeated attempt to secure wages in excess of the economic value of their work will inevitably lead to a spiral of rising wages and rising prices.

In a world that gives increasing consideration to the objective of full employment, the inflation that arises from excessive expectations of labor concerning real wages is the most difficult to deal with. The unemployment that would result from resisting an uneconomic level of wages is an obvious and measurable loss to the economy. On the other hand, the rise in prices that accompanies an uneconomic level of wages can easily be dismissed as part of the cost of maintaining industrial peace. Unfortunately, the price effects of an uneconomic level of wages are likely to stimulate labor unrest and to become the justification for new wage demands. An uneconomic level of wages thus results not only in higher prices but in a tendency toward continuously rising prices. As in all other cases of inflation, the price-wage spiral depends upon the expansion of bank credit to maintain full employment at higher costs. If the monetary authorities were to restrain the expansion of credit, an excessive rise in wages would lead to some unemployment; and unemployment would act as a deterrent to renewed demands for higher wages to offset the rise in prices. Unless an uneconomic wage level is resisted, there can be no escape from continuous inflation.

There is a great temptation for the monetary authorities to impute to the pressure for higher wages much of the blame for an inflation that has its origin in excessive investment and government expenditure. Wage and employment conditions have acquired political significance in all countries; and in some countries, unfortunately, governments tend to take the view that monetary policy must not impede full employment or rapid development even if it results in inflation. When there is pressure for wages to rise more rapidly than productivity, such wage behavior is to a large extent dependent on an environment of excessive public and private investment financed by bank credit. As a matter of fact, the wage-price spiral is itself conducive to excessive investment by creating expectations of rising prices and costs. Nothing could be more disastrous for monetary stability than wide acceptance of the view that the monetary authorities can do nothing about inflation on the ground that it is caused by pressure for higher wages.

In the confusion and disorder that accompany a large and prolonged inflation, it may be difficult to determine to what extent the price-wage spiral is specifically caused by excessive government expenditure, excessive investment, or excessive expectations of labor concerning real wages. Even though the inflation may have been initiated by huge budget deficits, or by very large investment, it may in the course of time have been converted into a wage inflation. After it has resulted in the expropriation of the fixed income group, a wage-price spiral may have little further effect on the distribution of real income, despite the disorder that it causes. In a large and prolonged inflation, the rise in prices and wages, and the accompanying expansion of the money supply, are part of a race to stay in the same place. Even with a very large rise in prices, the diversion of resources from consumption to investment, or from private to public use, soon becomes quite small.

In a large and prolonged inflation, the concern of labor, after a time, is not so much to increase real wages further as it is to protect the existing level of real wages from erosion through inflation. Although the inflation may have been caused initially by an uneconomic level of wages, the continued rise in prices will affect different wage groups quite unevenly. That is one reason why labor will generally support the linking of wages and prices in a large and prolonged inflation. Business may also support a wage-price link under such conditions in order to avoid the extreme and haphazard wage demands with which it may otherwise be confronted. It may be regrettable that the degree of flexibility in the economy is reduced by the linking of wages and prices. The only hope of restoring greater flexibility is to terminate the inflation. In an uninflated economy, it is doubtful whether labor would favor a formal linking of wages and prices, particularly if this would hamper wage increases on the basis of productivity.

Linking Wages and Prices

There is a good deal of evidence, as shown in Appendix II, that in all countries any persistent rise in the cost of living will induce demands for an offsetting rise in wages. Even in a country like the United States, in which the supply of and demand for labor in particular industries remains the primary factor affecting wage rates, the process of collective bargaining must take account of the behavior of consumer prices. In an article on wages and employment, the Federal Reserve Bulletin states:

The sustained rise in consumer prices since early 1956 has been an additional important factor in wage determination. The number of labor-management contracts containing cost-of-living escalator clauses has increased sharply. Relatively large wage increases negotiated in recent months in the rubber, petroleum, construction, and other major industries were also influenced by higher living costs. Moreover, in activities largely outside the sphere of direct union bargaining, wages have advanced to new highs under the pressure of rising living costs as well as expanding demand for labor. Although in some industries wages have increased more sharply than prices since early 1956, the rise in average earnings for the economy as a whole, as well as for production workers in manufacturing, has been largely matched by the increase in consumer prices.5

When the rise in prices becomes accelerated, an informal system of adjusting wages to prices inevitably arises. This is an indication of the loss of confidence in the functioning of the monetary system. This loss of confidence, in turn, adds to the difficulty of restoring economic stability. Under such conditions, attempts may be made to impose a link between wages and prices, either through legislation or through collective bargaining. There is no reason for believing that the mere adoption of such a link will slow down the wage-price spiral, much less bring it to a halt. Nor can the wage-price link be regarded as an automatic device for escaping the evils of inflation. Rather, it is an indication that the monetary system has failed to perform its function of regulating the production, distribution, and use of the national product.

The primary purpose of linking wages to prices is to give workers confidence that their real income will not be seriously reduced through continued inflation. With such assurance, it may be possible to avoid demands for an uneconomic level of wages based on a wide variety of claims. These demands may arise out of expectations for a rise in wages greater than the growth in the productivity of the economy would justify. They may arise from the assumption that the extraordinary level of profits in the inflation-favored industries provides a basis for determining what the general level of wages should be. The linking of wages to prices may provide a check on competition among trade unions to see which can secure the highest wage increase under the favorable bargaining conditions created by inflation. Once wage rates behave in an orderly way, it may be possible to take other measures, first to slow down and then to terminate the rise in prices.

There is another way in which the linking of wages to prices may contribute to the success of a comprehensive stabilization program. Partly because of its uneven incidence and partly because of attempts by the government to prevent a rise in prices in certain sensitive fields of consumer expenditure, inflation is likely to produce distortions. In various sectors of the economy, price controls may have been imposed “to hold down the cost of living.” An overvalued exchange rate may be supported by severe controls on payments for fear of the effect of devaluation on the cost of essential import goods. Ordinarily, labor is likely to object to the adjustment of such distortions through a rise in some prices, on the grounds that it is “inflationary.” With assurance that real wages will not be seriously affected by such adjustments, labor may not oppose measures taken to restore proper price relationships. Because aggregate real income is not reduced by eliminating such distortions, it may be possible to make the price adjustments and maintain the level of real wages without setting off a wage-price spiral.

If excessive demand is permitted to persist, however, prices will inevitably continue to rise. Whether they will rise more or less rapidly with than without a wage-price link depends upon the conditions that previously prevailed for adjusting wages. In an economy in which the adjustment of wages to higher consumer prices lagged and was incomplete, the introduction of a wage-price link could result in a tendency for prices to rise more rapidly as a consequence of any given excess in aggregate demand. On the other hand, in any economy in which wages were periodically overtaking prices through massive wage increases, and thus were establishing an uneconomic level toward which wage demands tended to gravitate, the introduction of a wage-price link could result in a tendency for prices to rise less rapidly as a consequence of any given excess in aggregate demand.

It is generally true that, wherever the linking of wages to prices has been introduced, this action has followed a period of large and prolonged inflation. Under such circumstances, some formal linking of wage rates with the cost of living may have become unavoidable. Nevertheless, it is important to recognize that such measures tend to perpetuate the distorted wage relationships of the past. Furthermore, they may limit the flexibility of the economy in meeting the problems of the future. If the making of economic policy is not to be unduly hampered by linking wages to prices, it is important to avoid protecting an uneconomic wage base that may prevent the government from meeting the problems with which the country is confronted. Furthermore, the government must retain sufficient flexibility in adjusting wages to prices to enable it to deal with any unexpected economic difficulties.

Some Dangers of Linking Wages to Prices

The principal danger in linking wages to prices is that the link will be based on an uneconomic level of real wages. Under such conditions, the economy would be confronted with the danger of a renewal of the wage-price spiral. If the monetary authorities were to take forceful action to restrain credit, in order to prevent a continued rise in prices, some unemployment would undoubtedly appear. If the basic wage to which prices are linked is not too far above the economic value of the work, the normal increase in productivity (if not accompanied by a corresponding rise in wages) may in time restore a proper wage level. That is the best way to restore price stability after wages have been linked to prices at an uneconomic wage level.

The process of stabilization involves adjusting relative prices, and not merely maintaining stability of the average price level. Unless wages in different industries are properly adjusted, some of the distortions generated by inflation will tend to be perpetuated. For example, wage differentials tend to be reduced in a period of inflation. If these differentials are frozen by the application of a fixed wage-price link, the economy may become unable to adjust the supply of labor to the demand for particular skills. Here again, the best hope of restoring an appropriate wage structure is through gradual adjustment of relative wages as productivity increases.

Another danger that may arise from the linking of wages to prices is the possibility that wage earners, who receive the greater part of the national income, may be exempted from bearing any of the hardships imposed by economic fluctuations at home or abroad. The burden of absorbing the impact of adverse conditions is thus thrown on other sectors of the economy. This may be inequitable and, in any case, it may not always be feasible. If a serious deterioration in the economic situation should occur, even though of a temporary nature, the linking of wages to prices may induce a resumption of the wage-price spiral. This can be avoided only if labor will assume its share of the burden of a reduction in the national income instead of relying on the wage-price link as a shelter from adversity in periods of economic stress.

The problem can be put more clearly in concrete terms. Suppose that a country which supplies a considerable part of its own consumption of agricultural products were to have a crop failure. Agricultural prices would rise. The real income of the country would be reduced by the amount by which agricultural production fell short of its previous level. Consumption and investment would have to be reduced, unless foreign exchange reserves were used to maintain them at a level which exceeded current output plus any net inflow of capital from abroad. A rigid wage-price link would in this situation protect the real wages of workers; however, this would not only add to the burden placed upon other sectors, but it might result in an uneconomic level of wages.

A similar situation would occur if the prices of import goods were to rise more than the prices of export goods. The prices of imported foodstuffs and raw materials may rise for any of several reasons which have their origin in world markets. An adverse movement in the terms of trade would reduce the real income of the country. If the rise in prices of import goods were relatively large and imports constituted a considerable proportion of total consumption and investment, the decline in real income could be significant. Nevertheless, if a rigid link between wages and prices were maintained, workers would not share in the burden being borne by the economy as a whole. An increase in wages to offset a rise in the prices of import goods, the demand for the country’s exports remaining unchanged, would result in an uneconomic level of wages and add further to the economic difficulties of the country.

The relation of taxes to the wage-price link is of special importance. Additional resources required by the government can be acquired—apart from foreign credits—only by reducing private consumption and investment. If inflation is to be avoided, this must be done largely through taxation. The taxes that may be levied for this purpose depend upon economic, social, and administrative considerations. To the extent that there is an increase in the personal income tax, the increased taxation will ordinarily have little effect on prices, although it will reduce the disposable income of people in all fields of economic activity. To the extent that additional taxes take the form of excises or customs duties, market prices will be higher than they would be if the same amount of revenue were raised through the income tax. If workers receive an increase in wages to compensate them for the part of the rise in market prices that reflects higher indirect taxes, they will escape paying their share of the taxes, and the anti-inflationary effect of the taxes will be reduced. In order to avoid this, the cost of living index used for adjusting wages to prices should be calculated net of indirect taxes.

As all these instances indicate, a system of linking wages to prices may impose a high degree of rigidity of real wages on the economy. Under favorable circumstances, the economy may be able, more or less, to adjust itself to such rigidity. Under unfavorable circumstances, the rigidity of average and relative wages, in real terms, may have serious consequences. It may impair the country’s competitive position in world trade. It may impose unemployment on the economy. It may even induce a resumption of the wage-price spiral. These dangers are inherent in the linking of wages to prices; they may be overcome to some extent by providing flexibility in the practical application of the wage-price link.

Administrative Safeguards

The greatest danger inherent in linking wages to prices is the risk of freezing an uneconomic wage base. If real wages substantially exceed the economic value of the wage earner’s work, the only remedy is to let prices rise gradually, with only a fractional offset in wages. If the real wages exceed the economic value of the wage earner’s work to only a small extent, it should be possible to correct this by a gradual rise in productivity which is not offset by a rise in wages. To prevent the continuance of an uneconomic level of wages, through claims for higher wages based on greater productivity, no wage adjustments should be made, except for the purpose of offsetting a rise in the cost of living, unless the index number of wholesale prices of domestically produced industrial goods has not risen during the preceding year.6 This may require a degree of wage control that is difficult to impose except under emergency conditions.

Even if an uneconomic wage base can be avoided, it is still important to keep the average rise in wages within the limits imposed by the actual increase in productivity in the economy as a whole. It would probably be futile to attempt statistical measurement of so complex a concept as a change in the productivity of the economy. Nor is such a measurement necessary for this purpose. On the whole, it would be better to provide that the increase in average wage rates to allow for the productivity factor shall not exceed some assumed standard rate of technical progress, say, 2 per cent per year. If such an increase in wages were to result in higher wholesale prices, as measured by the index defined above, a further rise in wages to allow for the productivity factor would be suspended until stability in the wholesale price level was restored.

Some means must be found to adjust relative wages in order to eliminate the distortions generated by inflation. Under free wage bargaining, this would ultimately come about by itself. Where the wage-price link is established by legislation and is administered as a check on money wages, such adjustments have to be made without destroying the wage restraint necessary for stability. Some fraction of the productivity allowance might be specifically allocated for the adjustment of wage rates in industries or occupations in which the supply of labor is markedly inadequate. This would mean that adjustments in relative wages would take place only during periods when wholesale prices are kept stable and efficiency wages are allowed to rise. These are difficult provisions to incorporate either in legislation or in collective bargaining agreements.

Every system linking wages with prices is open to the danger of a renewal of the wage-price spiral as a consequence of a temporary or fortuitous rise in the cost of living. To minimize this danger, the collection of goods and services covered by the cost of living index to which the wage adjustment is linked should be broad and representative. In many countries, the cost of living index has too narrow a base to serve as a suitable means of linking wages to prices. Furthermore, the index should use actual prices, rather than so-called official prices, even though the latter may for a time serve to hold down the cost of living index. When prices are decontrolled or official prices are adjusted upward, an index number that uses official prices rather than actual market prices will show an artificially large increase in the cost of living.

The broader the collection of goods included in the cost of living index, the less the danger that a fortuitous rise in prices will necessitate an adjustment of wages. Even so, there may be seasonal movements or temporary bulges in the prices of certain foodstuffs. Wage adjustments should not be made on the basis of a rise in the cost of living index in a single reporting period. Nor would it be desirable to undertake a formal adjustment of the average level of wages, unless there is a significant rise in prices.

The two tests of the magnitude and persistence of a price rise could be combined in a fairly simple way. Thus, it could be provided that no adjustment in wages would be made for a rise in the cost of living of less than some stated figure—say, 3 per cent—unless the price rise has persisted for some stated period—say, one year. If the rise in prices is somewhat higher—say, in excess of 3 but less than 5 per cent—an adjustment in wages could be made if the price rise persists for some shorter period—say, six months. If the rise in prices exceeds 5 per cent in a relatively short period, the adjustment of wages would have to be made more promptly. In general, however, the principle should be that the monetary authorities must be given time to impose more restrictive financial measures to halt the price rise and thus to minimize the need for wage adjustments.

Even with these safeguards, some provision should be made to suspend wage adjustments in an emergency. In terms of a program to halt inflation, an emergency exists whenever prices again begin to rise rapidly. This may be caused by a failure of the monetary authorities to maintain proper policies; it may be caused by external circumstances of a political character. In order to give adequate time for corrective action, the Ministry of Labor or a Wage Board should be authorized to declare an emergency and to suspend for six months all or part of a wage adjustment to offset the first 10 per cent of any price rise. It should be frankly recognized that the flexibility of wages downward, even under a system linking wages with the cost of living, is considerably less than the flexibility upward. For this reason, hasty adjustment of wages with every small and temporary rise in prices may result in building into the economy a structure of steadily rising wage costs.

Summary and Conclusions

It is important to recognize that the linking of wages to prices is not a remedy for inflation nor can it by itself restore monetary stability. The need for such a far-reaching step is an indication that the monetary authorities have not succeeded in the basic objective of all monetary policy—to give the people confidence in the stability of the value of money. Once an economy can restore a reasonable degree of monetary stability, it is far better off without any formal linking of wages to prices. As a matter of fact, when monetary stability is restored, neither labor nor business nor the government can have any good reason to continue such a tie, which in many respects interferes with the normal process of wage bargaining.

There are good reasons for avoiding any firm link between wages and prices, and more particularly between all wage rates and the index of consumer prices. The greatest degree of flexibility in an economy can be secured if prices and wages are allowed to adjust themselves to market conditions in an environment of monetary stability. Even at best, the linking of wages and prices will increase to some extent the rigidity of the wage structure. If the wage base which is linked to prices is in excess of the economic value of the work, the linking of wages to prices may become a serious obstacle to the restoration of monetary stability. Even if the wage base is properly related to the productivity of labor, it is still necessary to maintain safeguards in administering the wage-price link in order to avoid exempting wage income from sharing the burden of adverse economic forces. Without such safeguards, the wage-price link may become an instrument for generating and perpetuating inflation instead of a means of facilitating the adoption of a comprehensive stabilization program.

Although the linking of wages to prices is not a means of restoring stability to an economy suffering from an investment or budgetary inflation, it would be a mistake to assume that it is the cause of continued inflation. It is quite true that, if labor were willing to concur in a reduction in real wages below the economic value of its work, the price effects of a given excess in aggregate demand would be more limited. Nevertheless, it is unrealistic to expect workers to accept the whittling away of their wage standards by inflation at a time when business is benefiting from an inflated level of profits. Whatever economic objections may properly be made against the linking of wages to prices, the wage-price link will come into use in one form or another, unless the monetary authorities succeed in halting the inflation.

There is apt to be a danger of complacency on the part of labor, business, and the government toward the evils of inflation, once the linking of wages to prices has been established. No formula linking wages and prices can prevent the distortion in investment and production and the disruption of trade and payments which are the inevitable consequences of inflation. The interest of labor, as of the economy as a whole, is not merely in protecting the distribution of the present level of output, but even more in making sure that production and real income continue to grow. The only way to escape from the disorder caused by inflation is to create an environment in which the distribution and use of real income are responsive to economic forces and the monetary system cannot be used to generate an excessive demand or to support a claim for an uneconomic level of wages. That is the task of the monetary authorities.


I. Consumer Prices and Wholesale Prices

Some form of consumer price index is always used when a decision has been made to link wage adjustments to the movements of a price index. Of course, no index can reflect fully the effect of changes in prices on the real income of all people, not even of all wage earners. An index of consumer prices is based on a collection of goods entering into the expenditure of a representative family. At best, this collection will only approximate the pattern of expenditure of most family groups of the same size and income. Larger or smaller families with larger or smaller incomes may find that the cost of living for them has risen more or less than is shown by the index of consumer prices. In a country with significant regional differences, the behavior of consumer prices may vary somewhat among different cities and between urban and rural areas. These difficulties are inherent in an index number which is an indication, rather than a precise measure, of changes in the cost of living.

Even under conditions of monetary stability, the index of consumer prices tends to rise gradually, relative to wholesale prices. The consumer price index contains a larger element of services and rents than the index of wholesale prices; and productivity in the service industries, and perhaps in construction, increases somewhat less, on the average, than productivity in agriculture, manufacturing, and mining. Thus, while the behavior of the goods constituent of the consumer price index ordinarily tends to approximate that of wholesale prices, the rent and service constituents ordinarily tend to rise relative to wholesale prices.

The secular relationship between the cost of living and wholesale prices may be destroyed for an entire generation after a period of large and prolonged inflation. Thus, in the United States, consumer prices rose by 85 per cent between 1937 and 1953, while wholesale prices rose by 96 per cent. In the United Kingdom, the cost of living rose by 127 per cent, while wholesale prices rose by 203 per cent, in the same period. Whatever imperfections there may be in these index numbers, it cannot be doubted that the rise in consumer prices since 1937 has lagged far behind the rise in wholesale prices.7 Practically all of this lag is accounted for by the failure of rents and services to rise as rapidly as the prices of goods.

In recent years, the cost of living in many countries has been rising relative to wholesale prices (Table 1). Where wholesale prices have resumed an upward course during periods of cyclical expansion, as occurred in recent years in the United States, they have soon been overtaken by the continued rise in consumer prices. Whatever special causes there may be, which reflect conditions peculiar to any country, the basic reason for this behavior of consumer prices is the gradual resumption of their appropriate relationship to wholesale prices. This can be seen from an examination of changes since 1953 in some of the constituent series in the index of consumer prices in the United States (Table 2). The consumer prices of food and apparel have behaved in much the same way as the wholesale prices for the same products. On the other hand, rent, medical care, and personal care, which previously lagged behind wholesale prices, rose considerably in the same period.

Table 1.

Indices of Wholesale Prices and Cost of Living, Selected Countries, 1953 and 1957

(1937 = 100)

article image
Source: International Financial Statistics (Washington).

1938 = 100.

Table 2.

Percentage Increase from December 1953 to March 1958 in Constituents of Index of Consumer Prices in United States

(1937 = 100)

article image
Source: Council of Economic Advisers, Economic Indicators (Washington), May 1958, p. 23.

This experience is of some importance in judging the future behavior of prices. Consumer prices may be expected to continue to rise until they have resumed their normal relation to wholesale prices. Such a rise in consumer prices during the next few years is entirely consistent with stability in the index of wholesale prices. As long as the wholesale price level is stable, the gradual rise in the index of consumer prices does not call for any far-reaching measures to restrain aggregate expenditure. At the same time, it would be unfortunate if wholesale prices were allowed to continue to rise on the assumption that there is something inevitable about higher prices in a world of full employment and rising wage rates.

The linking of wages to prices requires an adjustment of wages with the rise in consumer prices. As long as average wage rates do not rise more than the average increase in productivity in agriculture, manufacturing, and mining, the wholesale price level can remain stable, the effects of changes in the terms of trade being neglected. If the monetary authorities are successful in restoring monetary stability, average wage rates will rise with the average increase in productivity, wholesale prices will remain approximately stable, and consumer prices will rise slowly—much less than the rise in wage rates. The proper policy for the monetary authorities in a large industrial country is to aim at stability of some appropriate index of wholesale prices—restraining the expansion of credit when this index of wholesale prices rises, encouraging the expansion of credit when the index falls.

The question will be asked, Why should the monetary authorities aim at stability of wholesale prices rather than of consumer prices? The objectives of monetary policy, insofar as they transcend price stabilization, are related to employment and to the balance of payments. Long-period stability of consumer prices would require a slow but steady decline in wholesale prices. On the domestic side, such a fall in wholesale prices would tend to discourage investment and increase the difficulty of maintaining employment. On the international side, trade competition is on the basis of wholesale prices, and national price levels can be more readily kept in proper relationship if each country succeeds in keeping its index of wholesale prices approximately stable.

In determining the index of wholesale prices with whose stability monetary policy should be concerned, it is important to choose a collection of goods that does not generate instability in employment or trade through changes in the prices of constituent series. If import goods are included in the collection, a rise in import prices would require an offsetting fall in domestic prices to maintain stability in the index. Under certain circumstances, such a fall in prices could be depressing to the domestic economy. Similarly, a rise (or fall) in agricultural prices would require an offsetting change in the prices of industrial goods. Since agricultural prices are particularly subject to change because of variations in supply, there would always be the danger of inducing a rise or fall in industrial prices to offset the fortuitous behavior of agricultural prices. The price level to whose approximate stability monetary policy should be directed is the wholesale price level of domestically produced industrial goods.8

A wage-price link of the type discussed in this paper assumes that wages will be adjusted for any rise in consumer prices, subject to certain safeguards. This will protect wage earners against any significant fall in real wages arising from an investment inflation. In a progressive economy, in which more output is available for private consumption and investment, wage earners are entitled to a wage rise that is related to the average increase in productivity in industry, provided such a wage policy does not bring about a rise in wholesale prices and does not impair the export capacity of the country.9 If the wage base is appropriate for the economy and aggregate expenditure is not excessive, an increase in wage rates equal to the average increase in productivity in industry will generally be larger than the secular rise in consumer prices. Furthermore, such a wage policy will be consistent with maintaining approximate stability in the index of wholesale prices of domestically produced industrial goods.

II. Survey of Wage-Price Links in Various Countries

A. K. Sen Gupta 10


Wages and prices have not been formally linked in Argentina, but wage rates have been subject to government regulation. In the postwar period, wage adjustments were usually made biennially and often by government decree. The objective of these adjustments was not only to compensate workers for increases in the cost of living, but also to increase real wages, particularly in industrial employment. While per capita real income increased by only about 3.5 per cent in the decade 1945-54, the proportion of national income accounted for by wages and salaries rose from 47 per cent to 60 per cent. The rise in real wages was uneven as between different groups of workers and was achieved largely at the cost of rural producers. The rise occurred in spurts, increases being brought about by government action in different periods, followed in each case by a decline in real wages through a rapid rate of inflation.

Following the devaluation in October 1955 and the lifting of price controls over a long list of consumer goods, the cost of living index rose sharply. On February 17, 1956, the Government declared an emergency wage and salary increase of 10 per cent throughout industry and commerce, established a minimum monthly wage of M$N 1,120 for persons over 22 years of age, and took steps to eliminate the differential between wages of male and female workers. Similar wage increases and minimum monthly compensation were later extended to government employees. Wages of agricultural workers were increased by 20 per cent. Any further increase in wages in the private sector was left to be decided by collective bargaining between employers and employees. It was decided that revision of the labor agreements was to be initiated between March 1, 1956 and December 31, 1956 and that the new agreements would remain valid for 18 months. It is estimated that these various wage adjustments during 1956 accounted for an average increase in wages of about 14 per cent—a rate somewhat higher than in 1955—and that devaluation and the removal of some price controls and subsidies along with the wage adjustments increased the cost of living during 1956 by about 17 per cent, which was about double the increase recorded in 1955.

The situation deteriorated somewhat during the first half of 1957. The cost of living rose more rapidly, mainly under the stimulus of large wage increases in the last quarter of 1956; this high wage level was maintained in the first quarter of 1957. A part of the rise in prices was due to the abolition of some meat and bread subsidies in March 1957 and the partial relaxation of price controls. Wages remained largely frozen at the high level reached at the end of 1956. Consequently, pressures developed for wage readjustments. Relative stagnation in production during 1956 and the first half of 1957 further complicated the situation. There was, however, in the later months of 1957 a movement by some private business interests to cut prices of some essential commodities, to prevent a continuance of labor unrest.


A link between the so-called “basic wage” and the cost of living has been in use in Australia since 1913. The basic wage, together with secondary wage payments, including margins for skill, etc., for particular occupations, constitutes the minimum wage rate which is fixed and revised by federal and state authorities from time to time. The link is applied to the “basic wage” part of the minimum wage; the remaining secondary payments are not adjusted to the cost of living, but they are revised on other grounds from time to time.

Between 1913 and 1921, as cases came before the Commonwealth Court of Conciliation and Arbitration, adjustments based on movements in the relevant retail price index during the previous calendar year were made in the basic wage. However, during the period of rapidly rising prices toward the end of World War I, strong criticism developed that this system did not adequately maintain the basic wage in real terms. Consequently, in 1921, the practice was adopted of making automatic quarterly adjustments in the basic wage in proportion to variations in the retail price index during the preceding quarter. This system was continued until a judgment of September 12, 1953 of the Court led to its abolition.

Between 1946 and 1953, a substantial part of the rise in the basic wage level was in the form of cost of living adjustments. The Court, however, awarded additional increases of 7s. in December 1946 and of £1 in December 1950. In total, the average basic wage level for the six state capitals prescribed by the Court rose from £4 16s. in November 1945 to £11 16s. in August 1953.

In September 1953, the Court decided that the system of automatic adjustments conflicted with industry’s “capacity to pay,” and that it might have been an accelerating factor in the rapid increases in prices in Australia, particularly in 1951 and 1952. In fact, during 1952 and 1953, the price-wage spiral had continued despite a substantial increase in unemployment. The Court therefore decided to discontinue the principle of automatic adjustment of the basic wage in proportion to variations in the retail price index.

The Court nominally has jurisdiction over only those industries which extend beyond the limits of a single state. In terms of working population, its jurisdiction extends to about 50 per cent of Australian wage and salary earners; most of the other half come under the jurisdiction of state industrial tribunals. Up to 1953, the decisions of the Court in the matter of wage adjustments were followed by the various state authorities. After the 1953 decision, however, the system of quarterly adjustments of wages to prices was retained by the wage-fixing authorities of most states. As a result, anomalies developed in the wage structure in the course of 1954-55, leading to loss of labor and to strike threats in industries paying the lower federal basic wage. In May 1956, the Court rejected union claims for restoration of the system of automatic quarterly adjustments of wages, but favored annual review of the federal basic wage on the basis of the economy’s capacity to pay. Following discussions between federal and state authorities during the second half of 1956, which stressed the need to pursue uniform wage policies throughout the country, a few states discontinued the system of automatic quarterly adjustments of wages to prices. This, along with some wage awards by the federal authority (a wage increase of 10s. a week granted in May 1956 to male workers and another similar increase allowed in May 1957), brought federal basic wages closer to the basic wages paid to workers under the awards of the state tribunals.


In the earlier postwar years, wages in Austria were periodically adjusted to prices on the basis of nation-wide agreements which covered not only wages and prices but also broader aspects of fiscal and general economic policy. Between 1947 and 1951, organized labor, employers, and agricultural groups under government guidance and sanction negotiated wage limits, fixed agricultural prices, and participated in the determination of prices, taxes, subsidies, and pensions on a national scale. In the five agreements that were negotiated during this period, prices and wages were controlled and adjusted from time to time until the initial phase of reconstruction was over. Prices paid to farmers, retail prices of food, and other controlled prices were increased by each agreement; the resulting increases in the cost of living necessitated compensatory wage increases. This in turn raised industrial prices, and farmers demanded an adjustment of their prices. The agreements did not prevent a substantial price rise, but they limited somewhat the price effects of the inflation.

Between 1948 and 1951 there were annual average increases of 35 per cent in prices and wages. Yet because of the lags in wage adjustments and other price control and fiscal measures, the burden of a high rate of investment was borne by wages, distributed profits, and farm incomes alike, so that in a certain sense the incidence of the inflation of 1948-51 may be said to have been determined by public policy. The last two wage-price agreements demonstrated, however, the futility of continuing this kind of control, and the experiment was abandoned in 1952. Meanwhile, the economy had grown stronger through the recovery of production, and an effective stabilization program, along with a devaluation of the schilling in May 1953, was undertaken without major price and wage adjustments.

Although the system of general wage-price agreements under government auspices was abandoned, a number of wage contracts began to include sliding scale provisions. Such provisions usually require a rise in prices of the order of 5 to 10 per cent before a compensatory adjustment of wages can be applied. This is usually sufficient to ensure that seasonal or purely temporary price fluctuations do not lead to wage increases. Attempts are being made to change the cost of living index (which includes only 83 items) to include a much larger number of goods and to eliminate, as far as possible, seasonal influences from the price index.

While in industry a sharp increase in productivity has, in recent years, led to substantial increases in real wages without widespread reliance on wage-price links, problems have sometimes emerged from lagging farm income and the consequent pressure for higher agricultural prices. Such pressures grew again during 1956-57. However, the Government was careful to avoid price adjustments that would require wage adjustments and thus induce a new wage-price spiral.


The use of sliding scale provisions is widespread in Belgium. They appear in almost all wage agreements where they apply to both manual and white-collar workers. Some agricultural workers and government employees also receive automatic adjustments in wages related to increases in the cost of living. The price index used for such provisions is an index of retail prices compiled by the National Institute of Statistics. This is a new index based on prices in 1953 and replaces the older index based on 1936-38 prices, which was outdated. Usually, wages are adjusted every two or three months if the average price index of the preceding months exceeds a certain critical limit.

The decision to link wages with retail prices was reached in 1950 by national employer and employee organizations; by 1952 nearly all Belgian industries were covered by such agreements. Government policy has been directed toward keeping the cost of living stable, and between 1952 and 1955 this policy was markedly successful. With relative stability in the price level, the attention of organized labor shifted to gaining a larger share of increasing productivity. In May 1954, a joint declaration to this effect was signed by labor and employer organizations.

The sliding scale provisions in use in various industries involve the choice of a base date and a range in the retail price index between which the index may vary without requiring wage adjustment. Once the relevant price index moves beyond this range, wages are adjusted by a certain percentage, decided in advance, which may provide full or partial compensation for the price increase. The exact provisions regarding the base date, the range of the rise in prices that is permitted without requiring wage adjustment, and the extent of compensation for a given price rise differ from industry to industry. The most common provision is for a proportionate adjustment in wages when prices have risen by 2.5 per cent. The sliding scale provisions work the same way in both directions—upward and downward. However, under a special agreement, wages were not reduced early in 1955 although retail prices declined.


In Bolivia, massive wage increases have been repeatedly decreed by the Government during the past few years in an effort to compensate for increases in the cost of living and to raise the real wages of certain groups. Following the nationalization of the large tin mines in 1952, the general level of wage rates was raised by an average increase of 40-50 per cent in wages and salaries and in benefits to workers of two of the largest mining companies and by somewhat smaller increases for other workers.

In May 1953, the stabilization efforts included the devaluation of the boliviano from Bs 60 to Bs 190 per U.S. dollar; this, along with further increases in import surcharges, raised the cost of living substantially, through the corresponding marking up of controlled prices. To compensate for part of the increase in the cost of living, a flat increase in wages and salaries was granted to government and private employees. The adjustment was smaller for employees of the Mining Corporation than for most other groups, but prices of wage goods at the pulperias (company stores) were frozen at subsidized predevaluation levels. It was also decided to freeze wages for six months.

The wage-price spiral, which accelerated in 1954-56, was sustained by a number of wage adjustments secured by government decree. Effective January 1954, social security benefits were broadened to cover all classes of employees. By a decree of April 1954, workers’ wages were raised by 45-50 per cent and salaries of employees in commercial firms by 25-50 per cent, retroactive to January 1, 1954. With prices rising, a general increase of 40-75 per cent in wages was granted in the spring of 1955. This was followed by a 49 per cent increase in the cost of living by mid-1955. Pressures for wage adjustments that began again in late 1955 resulted in advance payments of Christmas and other year-end bonuses. On February 25, 1956, another round of wage and salary adjustments was decreed. Wages and salaries were increased by 52-89 per cent in private industry and commerce, by 46-76 per cent for civil servants and other salaried employees, and by 116 per cent for miners in the nationalized mines. Minimum wages and salaries were also increased. Still another round of wage increases began in June 1956 and was completed by November 1956; this increase averaged 40 per cent for all workers. Early in November 1956, the Government announced a wage freeze that was to continue until a stabilization program came into effect.

On December 15, 1956, a comprehensive stabilization program was put into effect involving major adjustments in the exchange system as well as in taxes, tariffs, government expenditures, wages, and social security. Most subsidies and price controls were removed, and efforts were made to eliminate the deficits of the principal government-owned enterprises. The system of subsidized sales of goods by pulperias was abolished in nationalized mines as well as in private mining, railway, and other enterprises. In compensation for the loss of subsidy, wages of workers formerly having access to the pulperias were raised by fixed amounts. In addition, as compensation for changes in the exchange rate and the resulting increase in the cost of living, a flat wage increase of Bs 1,300 per day was decreed for most workers. After this partial compensation for the devaluation and abolition of subsidies, wages were to be frozen for a year. In granting only partial compensation for the increase in the cost of living, the authorities were attempting to fix wages at an economic level, which was considered lower than the real wages that had nominally prevailed in certain earlier periods. The stabilization program has been largely successful, and the cost of living index declined in the first three quarters of 1957.


Wage and salary adjustments, usually in the form of massive increases for large classes of employees, have been made periodically in Chile. The primary purpose of the adjustments has been to compensate for increases in the cost of living. For example, at the time of the exchange reform of July 1953, legislation was adopted to compensate for the rise in the cost of living expected to result from the depreciation of the import rate for foodstuffs and other cost of living items. The legislation also extended the system of family allowances for workers. Massive adjustments of wage rates were made in 1954 and 1955 in an attempt to compensate for the price rises of 1953 and 1954.

Legislation existing in 1955 provided for the periodic adjustment of salaries of white-collar workers and government employees by amounts corresponding roughly to the increase in the cost of living during the previous year. Salary adjustments generally were made in January each year. The mechanism of the wage and salary revisions differed for (1) manual workers in industry, commerce, mining, and services, (2) white-collar workers in the private sector, and (3) government employees. Agricultural workers usually received annual adjustments in January; these adjustments were measured roughly by the revision each year of the minimum wages fixed by legislation. For employees in the private sector, yearly revisions were made in the minimum wage legislation. Nonagricultural laborers generally obtained, at the expiration of their collective contracts, a wage adjustment commensurate with the increase in the cost of living during the previous year. Some workers in the private sector (particularly employees of some of the larger mining companies) also had contracts providing for automatic wage revision when the cost of living index rose more than a certain percentage—generally 5 per cent—in any month. Government workers were entitled to salary increases of 25-90 per cent (the percentage depending on the level of their salaries) of the estimated rise in the cost of living index during the previous year. Each adjustment in wages and salaries resulted in a similar increase in the amounts of social security contributions paid by employers and the Government in certain fixed percentages of total wages.

Early in 1956, steps were taken, as part of a comprehensive exchange reform and stabilization program, to break the wage-price spiral. Under the previous law requiring automatic increases in salaries and wages, an increase of about 80 per cent was due in 1956, based on changes in the cost of living. This law was abrogated, however, and on January 23, 1956 the Congress approved a new law limiting automatic wage and salary adjustments. Salary readjustments for private employees carried out at the beginning of 1956 were generally limited to 50 per cent of the cost of living increase from December 1954 to January 15, 1956 (i.e., to 46.5 per cent, which is half the increase of 93 per cent in the cost of living); for bachelors with less than ten years of service in the enterprise, the adjustment was 33 per cent of the increase in the cost of living, i.e., 31 per cent. For government employees, the law provided increases of 44-50 per cent of the rise in the cost of living (in place of the earlier legislation permitting an increase of 25-90 per cent of the increase in the cost of living), depending on the level of salaries. This change not only limited the absolute increase in wages; it was also intended to check the narrowing of salary differentials among public servants.

The provision limiting wage increases to no more than 50 per cent of the rise in the cost of living between December 1954 and January 15, 1956 was made obligatory also for manual workers not under any system of automatic readjustment but whose wages were readjusted by collective bargaining or on the basis of changes made in the minimum wage. In some of the larger mines where escalator provisions with monthly readjustments were in operation, unions agreed to receive compensation for only 50 per cent of the rise in the cost of living beyond the “critical” point.

The new wage policy attempted to ensure that wages would be determined far more by the normal processes of bargaining than by law. For workers in private industry, any agreed increase in wages and salaries over the provisions of the law was valid, but for taxation purposes, the excess was to be considered as profits, rather than as costs of the enterprise. The new legislation also provided for the discontinuance of “automatic” wage adjustments in the future.

These wage stabilization measures were partly responsible for the slowing down of the rate of inflation in 1956. Although the cost of living rose by 38 per cent in 1956, this was less than half the increase in the previous year.

For 1957, the Government originally decided to limit wage increases to about 25 per cent. However, pressures within the Congress and outside resulted in wage increases of slightly over 30 per cent, amounting to about 80 per cent compensation for the increase in the cost of living during 1956. The wage-price spiral is likely to continue until effective steps can be taken to stabilize the cost of living index. During the first eight months of 1957, the cost of living index rose by about 29 per cent.

The use of the cost of living index in effecting massive wage increases has led to a critical re-examination in Chile of the adequacy of this index and to proposals for its reconstruction. The present index is largely outdated, and the weighting system is subject to criticism. For example, a study prepared by the statistical department of the Central Bank showed that the rise in the index in September and October 1956 was due largely to an increase in the price of onions resulting from a temporary winter shortage. While some items have excessive weights in the index, many others are believed to have weights that are too small to reflect their importance.


Adjustment of wages to changes in the cost of living in Denmark dates back to 1918. For wage earners, such adjustments have been regulated in postwar years on the lines of a voluntary agreement between nationally organized workers and employers which was concluded in March 1946. According to law, salaries of civil servants are also adjusted in accordance with changes in the cost of living. The provisions of this law have been used as a model for determining salaries of most white-collar workers, e.g., bank employees. The use of the sliding scale is widespread and covers most wage earners, salaried people, and even pensioners.

Wages are revised twice a year on the basis of the cost of living index in January and July. There is a flat-rate increase of wages of around 1 per cent for each six points of increase (equivalent to about 1.5 per cent) in the index (1941 = 100). Collective bargaining agreements are usually valid for two years, and it is believed that agreements for such a period can be secured only by guaranteeing workers’ real wages against serious impairment from inflation.

The weights in the cost of living index are based on an inquiry carried out in 1948 into the expenditure of a sample of families belonging to the working and the white-collar classes. The weight attached to food is low (about 27 per cent), and food, clothing, and rent together account for half of the total weight. Taxes on income are treated as expenditure in the sample family budget used for measuring changes in the cost of living; in periods of stress arising from a fall in productivity or a deterioration in the terms of trade, it is therefore difficult to take corrective fiscal measures, as higher taxes will result in increases in wages because of the resulting increase in the cost of living index.

In view of these rigid wage-price links, attempts have been made, particularly in periods of balance of payments crisis in 1955-57, to reduce domestic consumption in ways that do not increase the cost of living. Thus, the stabilization policy adopted on June 11, 1957 to reduce domestic absorption included (1) the exclusion of some increases in indirect taxes from the computation of the cost of living index used for wage adjustments and (2) a compulsory saving device. However, the cost of living continued to rise, and substantial automatic wage increases were again required during September-October 1957. Meanwhile, the possibility of omitting both direct and indirect taxes from the cost of living index used for wage adjustments is being seriously considered by the Government.

Though flat rate cost of living adjustments have been paid to virtually all workers throughout the postwar period, there has been no narrowing of differentials based on skill. Wages of skilled workers have gone up more both in absolute amounts and in percentages than those of unskilled workers. This is because a good part of the rise in hourly earnings was due to the “wage drift”11 (Table 3), employers paying above-award wage increases to more qualified workers in particular sectors. Because of considerable unemployment among unskilled workers throughout the postwar years, the “wage drift” was relatively small for them.

Table 3.

Components of Increase, 1947-55, in Average Hourly Earnings Granted to Male Industrial Workers in Denmark

(In öre per hour)

article image
Source: United Nations, Economic Survey of Europe in 1955 (Geneva, 1956).


Sliding scale arrangements have been widely used in Finland in the postwar years. The big wage increases immediately after the war exceeded the rise in the cost of living, and later attempts, through sliding scale arrangements, to maintain real wages at the rather high level then reached greatly handicapped the Government’s efforts toward price stabilization. The efforts of the farmers to increase their share of the national income were also an important factor in the price instability of the postwar period. Up to September 1951, the cost of living index used for wage regulation included direct taxes. In September 1951, however, all sliding scale arrangements were linked to a new cost of living index from which direct taxes were excluded. During the entire postwar period, the Government has found it necessary at times to limit the operation of the sliding scale arrangement by changing the official index or by direct price controls and subsidies. In 1951-55, prices were kept relatively stable by these means; also, stability was supported by a slight fall in import prices.

The stabilization agreement of 1951 provided for a 5 per cent wage rise whenever the new cost of living index (October 1951 = 100) increased by 5 per cent. If prices should fall, however, no adjustment was required unless the fall was at least 10 per cent.

The compulsory link between wages and prices, administered by the Government, came to an end when the Emergency Powers Act, in effect from 1951, expired at the end of 1955. Decisions on wages were then left to be settled by negotiation between the SAK (the Federation of Finnish Trade Unions) and the STK (the Federation of Finnish Employers). Price controls were abolished, milk subsidies were reduced, and rents of old houses were increased. Meanwhile, import prices of fuels, rubber, and metals were rising. All these factors led to a rise in the cost of living index from 101 in December 1955 to 108 in March 1956. The Federation of Trade Unions demanded either that the increase in the price of milk be canceled or that this increase be compensated by a wage increase; this led to a three weeks’ general strike. A compromise agreement between the unions and employees was finally reached, resulting in an increase by 11-12 per cent in average hourly wages. To prevent a wage-price spiral, the Government lowered the corporate income tax and the property tax, thus offsetting some of the increase in wage costs, and it also reassumed powers to control prices. New negotiations for wage adjustments were permitted only if the price index increased another 5 points (from 107 in February 1956). By a decree in May 1956 (covering the three crop years following September 1, 1956), farmers’ incomes were also linked to wages in such a way that a 1 per cent rise in wages would be followed by a rise of 0.85 per cent in farmers’ incomes in the form of a corresponding rise in the prices of agricultural produce.

Prices continued to rise during 1956. The wage negotiations which began in the autumn of 1956 between labor and employer organizations yielded no results, despite the mediation efforts of the government arbitrator appointed in November 1956. Late in 1956 a price freeze was announced. In March 1957, corporate taxes were again raised, and the Government recommended some rise in wages at a future date if the cost of living should continue to rise; it also decided to bring down some “critical” prices by reducing the producers’ price of milk and the consumers’ price of butter and cheese. Trade unions and farmers refused to accept this recommendation, which precipitated a political crisis.

Between February 1956 and July 1957, the cost of living rose by about 14 per cent. As production declined, unemployment grew and the deterioration of Finland’s international competitive position led to continuing balance of payments deficits. The trade unions, which had previously demanded full compensation for the rise in the cost of living, scaled down their wage demands in the third quarter of 1957 to modest increases of 1½-2 per cent for the year 1957. Most of the new collective wage agreements concluded during the third quarter of 1957 and expiring at the end of 1958 also contain a two-thirds compensation clause for any future wage increases, i.e., wages would be raised by 4 per cent when the cost of living index reaches 130 (October 1951 = 100), and by a further amount, making a total increase of 7 per cent, when the index reaches 136.

The restraint exercised in wage demands provided a suitable climate for the devaluation of the markka on September 15, 1957 to offset the adverse effects of an already inflated cost-price structure. It is not yet clear whether farmers will follow the trade unions in making concessions and agreeing to revise their index agreements and legally fixed income guarantees.


Sliding scale provisions are rarely used in collective wage agreements in France. A significant link between wages and prices is established, however, through the variation of the legal minimum wage rates in accordance with increases in the cost of living. During the early postwar years, minimum wage rates were so high relative to wages above the minimum that any significant change in them had substantial effects throughout the wage structure. This was particularly true in 1945-50, when the national wage structure was regulated in detail by the Government, which fixed occupational, geographical, and skill differentials as well as the minimum wage.

In 1950, wage differentials were formally returned to private control, leaving only the minimum wage rate under government control. Regional differences in minimum wage rates, however, remained. Wage differentials, based on skill, which had narrowed in 1945-50 owing to flat rate cost of living adjustments in the minimum wage rates, widened again until 1953. In 1954-55, there was a renewed tendency toward a narrowing of differentials as a result of three government decrees designed specifically to raise the earnings of the lowest paid groups.

In July 1952, legislation was passed by which it became obligatory (rather than discretionary as hitherto) to raise minimum wages in the same proportion as price increases whenever the retail price index showed an increase of at least 5 per cent and which was maintained for at least two months. The price base for minimum wages was 142 (1949 = 100) so that the sliding scale was to start operating when the cost of living index reached 149.10. The cost of living index was maintained relatively stable, so that no change in minimum wage rates to offset a price rise was necessary until August 1957. The relative stability in the cost of living index was maintained by control of various prices, by subsidies, by cutting indirect taxes, or by increasing imports and decreasing exports of particularly scarce goods included in the index.

Although, until 1957, minimum wage rates were not raised on cost of living grounds, they were sometimes revised (e.g., in April and October 1954 and in April 1955) for manual workers on grounds of an increase in productivity. Wages above the minimum were also revised through collective bargaining between employers and trade unions. After April 1955, the policy of periodic review of minimum wages on grounds of an increase in productivity was abandoned by the Government, and further changes in particular wages on the basis of increased productivity were left to be decided exclusively through collective bargaining. However, the Government adopted a series of other measures which increased wage costs. Thus, in September 1955 employers’ social security contributions were raised. In March 1956, the number of paid holidays was increased, and wages in the provinces were raised. A 10 per cent rise in the special security allowance to the lowest paid retired workers was granted, the cost to be financed through the budget.

During 1956 and the first half of 1957, there were continuous pressures on the cost of living which the Government sought to relieve, as before, by subsidies and other measures designed to keep the index below the “critical” point. Fiscal and commercial policies limited the rise in the cost of living index to about 1 per cent during 1956, but entailed considerable budget deficits and a substantial drain on foreign exchange reserves. The policy of stabilizing the cost of living index through such artificial devices was therefore abandoned in July 1957, when the index was allowed to rise beyond the critical level, to 150. Consequently, a 5.5 per cent increase in minimum wages was granted on August 1, 1957. This affected directly the wages of about 700,000-800,000 workers, or about 20 per cent of the working population. After the end of 1950, guaranteed minimum wages and the general wage level had drifted apart—hourly wage rates rising by almost 40 per cent and minimum wages by only about 26.5 per cent. However, following the 1957 rise in prices and the upward adjustment of minimum wages, new wage demands were made in other sectors as well.

The nature of the link between minimum wages and prices, as well as the adequacy of the cost of living index used in administering the link, has received critical re-examination in recent years. The cost of living index used for the wage-price link was calculated until mid-1957 on the basis of a standard budget of a family consisting of a man, his wife, and two children, and it included 213 items. There were 41 food items, 115 manufactured goods, 7 light and heating items, 47 services, and 3 sundry articles. Three main objections were raised against the use of this index as a basis for minimum wage adjustments: (1) it gave too great weight (58 per cent) to food; (2) it was based on family budgets of workers in the Paris area, although the resulting minimum wage adjustments applied to all French workers; and (3) it included a number of commodities and/or it attached weights to them which were not consistent with the typical budget of the lowest paid workers to whom the index guarantee applies. In 1957, a new cost of living index was constructed to serve as a basis for future minimum wage adjustments. The new index, unlike the old index, is based on the typical consumption of an unmarried, unskilled male worker; it covers only 179 consumer articles and uses different weights. It is, however, still based on the consumption habits of workers in the Paris area. Trade unions have, so far, refused to accept this new index as a basis for minimum wage adjustments, partly because its limited coverage of commodities ignores family needs.

A more radical change in the administration of the link affects the degree of price rise required for wage adjustments. Instead of 5 per cent, as earlier, a 2 per cent rise in the new cost of living index lasting for at least two months is to entail a corresponding 2 per cent increase in minimum wages. This would reduce the magnitude of the impact of the minimum wage adjustments at the points where they occur, though over a given period of time it may entail more frequent wage adjustments. Such smaller increases, it is believed, are also likely to create less pressure for simultaneous wage increases from workers whose wages are above the minimum wage level, and who constitute about 80 per cent of the total number of French wage earners.

The increase in minimum wages on August 1, 1957 was followed by export premiums and import taxes which were equivalent to a selective devaluation. To prevent resumption of a price-wage spiral, some prices were frozen and others cut by government decree.


Collective bargaining agreements in West Germany do not provide for automatic adjustments of wages to the cost of living. In the hyperinflation of the 1920’s, the period for wage revision under such agreements was reduced from two months in 1921, to one month in 1922, to a fortnight at the beginning of 1923, and to one week by the middle of that year. Agricultural wages were linked in 1923 to the prices of rye and potatoes, and industrial wages were often linked to some crude price index. The experience of the 1920’s has made businessmen and the Government extremely reluctant to permit any escalator provisions in collective bargaining agreements. Up to 1950, when the first phase of reconstruction was practically complete and real wages appeared to have reached their prewar level, wage demands were modest and, in spite of rising prices, they were not pushed to extremes. Unlike the experience in the period following World War I, strikes have been rare. The strike of metal workers at Hesse in the fall of 1951, which lasted several weeks, was the first major signal of wage pressures consequent on the sharp rise in prices in 1951.

Though no formal link exists between wages and the cost of living, an increasing number of collective bargaining agreements concluded during the price rise of 1951-52 were for such short periods that fairly frequent adjustments of wages to changes in the cost of living were in fact taking place. The effects of this practice are similar to those of the so-called wage-reopening clauses, which occur in certain collective bargaining agreements in West Germany. Thus, the agreements in February 1951 for the printing and allied industries stipulated that wage negotiations might be reopened if the cost of living index moved up by 7 per cent; in June 1951, this margin was reduced to 3 per cent. In the latest wage agreements for the printing and allied industries, however, such specific wage-reopening clauses linked to the cost of living have been dropped and replaced by a general clause which gives either side the right to give notice of withdrawal from the agreement if economic conditions should have changed.

The spate of strikes in 1954, when prices were stable, was an expression of labor’s demand for a larger share in increasing productivity. Pressures for wage increases grew as practically full employment was reached in 1955. The rise in the cost of living in 1955 and 1956, which was due largely to higher agricultural prices, further contributed to growing wage pressures in industry and resulted in higher wages. However, a considerable rise in productivity, the lag in wage adjustments, the maintenance by employers of agreed wage rates, and the refusal to accept wage-price links have made it possible for the German Government to deal with inflation through financial policy more effectively than has been possible in most countries.

The implications of wages-cost of living links have been widely discussed by the German press and studied by various government departments. Problems involved in seasonal fluctuations in the cost of living have also been discussed. The conclusions in all cases have been that such links would jeopardize the flexibility of the economy, handicap monetary and fiscal policy, and possibly lead to an acceleration of inflation.


Most of the gainfully employed persons in Israel have been paid wages subject to a cost of living allowance since early in the 1940’s. An automatic increase in wages is granted whenever the price index rises by 3 points (about 1.2 per cent) or more within a period of three months. This provision is not statutory, but it is included in the agreements between the General Federation of Labor (Histadrut) and the Manufacturers’ Association. The Government not only does not interfere while these two organizations are negotiating an agreement, but it undertakes to grant similar terms to government employees.

In the first five months of 1956, the index used for the wage-price link rose by 4.2 per cent (from 239 to 249, September 1951 = 100), mainly because of higher prices of vegetables, the supply of which declined owing to adverse weather conditions, but also because of new government and municipal taxes. The Manufacturers’ Association, after agreeing to an increase in basic wages, refused to renew the cost of living allowance agreement. Early in 1957, through the mediation of the Government, the Histadrut and the Manufacturers’ Association reached a new agreement, valid for two years, by which wage adjustments will be made every six months (instead of three months as earlier) if the cost of living index rises by 3 per cent (not 3 points, or roughly 1.2 per cent, as earlier). However, if the index rises by as much as 5 per cent within a period of three months, an adjustment will be made before the six-month period has elapsed.

By lagging the link between prices and wages, the new agreement should make it easier to maintain price stability through fiscal and credit policy. The arrangement that collective bargaining agreements between employers and employees are to be for two years instead of one year will also contribute to greater stability in wage rates. A stricter control over premium payments and promotions has also been advocated in the 1957 agreement, so as to prevent “automatic” premium payments and fictitious upgrading of workers which in the past has been responsible for part of the “wage drift” beyond the nationally agreed basic wage rates. A good part of such “wage drift” has been in the form of incentive earnings, which did not exist before 1952 but have since become widespread.

Two other important changes have been made in the system linking wages and the cost of living. First, the minimum to which cost of living adjustments are applied has been increased from the first I£125 of the basic monthly salary to I£500 per month. The previous low limit had meant a narrowing of wage differentials, which provoked internal tensions in the Histadrut. Second, under the new wage system the pay which a worker received in January 1957 is considered to be his basic wage. This includes not only the basic wage paid in November 1956 but also the cost of living allowance and the frozen one third of the salary increases granted to workers under the wage agreements concluded in the beginning of 1956.

That this basic wage is too high relative to productivity is believed to be indicated by the pressure of inflation, the large import surplus, and the difficulty of providing employment for the growing number of immigrants. Partly because of such pressure, the cost of living index rose again by 3.2 per cent in the first six months of 1957, necessitating a 3.2 per cent rise in gross wages, under the index agreement, as from July 16, 1957. The pressure of wage inflation thus remains, although increasing productivity, restraint in wage increases, and strict monetary policy should eventually bring about a proper wage scale.


In Italy, a sliding scale of wages was applied in the northern provinces in December 1945. Various modifications have since been made in the cost of living index as well as in the provisions for adjusting wages. The use of the sliding scale has become widespread, covering most manual and white-collar workers and many government employees and agricultural laborers.

A national agreement for a wage-price link was made on March 21, 1951 between the Confederation of Italian Industries and the three large federations of trade unions. For the purpose of calculating wage adjustments, wages were divided into two parts—base pay and “temporary indemnity.” The “temporary indemnity” was to vary with changes in the cost of living after the base date, November-December 1950. Adjustments in the temporary indemnity were required every two months if the average of the cost of living index in the two preceding months had risen by over 0.5 point. For the purpose of the sliding scale, any variation of less than 0.5 point is ignored, and a variation between 0.51 and 0.99 point is treated as a full point. For each “point” of increase in the cost of living index, the agreement fixed a wage adjustment designated in lire, the number of lire differing according to geographic region and industry and the age and sex of the worker. Care was thus taken not to narrow wage differentials.

The sliding scale arrangements agreed to in 1951 did not work in the same way with regard to downward movements in the cost of living index. Adjustment was automatic for both upward and downward price movements, but there was a longer time-lag if wages had to be reduced. The sliding scale thus could operate downward and cause a reduction in the “temporary indemnity” only if the cost of living index showed a drop of more than 4 points from the highest peak previously reached. Another important provision of the 1951 agreement was that a variation of more than 25 per cent in the cost of living, compared with the base period, would entitle either party to the contract to request that the agreement be re-examined. With some differences in detail, sliding scale provisions, similar to those negotiated in the industrial sector in March 1951, were adopted in nonindustrial sectors, e.g., in trade, credit, and banking, and for farm workers.

The technical deficiencies of the index, the short time-lag between movements of prices and of wages, and the asymmetry between upward and downward adjustments were subjected to considerable criticism in early 1956, when bad weather conditions caused a shortage of fruits and vegetables and forced up the price index and, through the sliding scale, the wage level. On January 15, 1957 a new agreement was signed by the employers’ and workers’ organizations, which became operative from February 1, 1957. It differs in several respects from the agreement of March 1951.

Under the new agreement, the so-called “temporary indemnity” is to be changed at intervals of three months, instead of two months as hitherto. Price changes will be calculated by averaging the index numbers of the preceding three months. The new agreement also accepted the principle of equal variation for both upward and downward adjustments. In practice, however, the index has to drop by 1 whole point for the sliding scale to operate downward, whereas any rise in prices of more than 0.8 point requires an upward wage adjustment. Furthermore, if prices fall, the “temporary indemnity” will not be reduced in the quarter immediately following the fall, but in the quarter after that, and then only if the index remains at the lower level.

The base of the index number to be used for wage adjustment is the average of May-June 1956. The high wage level recorded at that time was responsible for the last rise in the “temporary indemnity” on August 1, 1956 under the agreement of March 1951. The use of any other base period, with a lower index, would obviously have meant that any rises in the cost of living that had already caused increases in the temporary indemnity would be taken into account a second time in calculating future wage increases. Some minor changes have also been made in the cost of living index used for sliding scale purposes, e.g., in the weights used for averaging local price indices and in the recording of fruit and vegetable prices so as to minimize seasonal fluctuations.


The Netherlands is the only European country where all increases in wages are subject to government sanction. No formal link exists between wages and prices. Changes in the cost of living have, of course, been taken into consideration in fixing wages. At times, with the cooperation of labor and employers’ organizations, real wages have been reduced to cut down consumption in order to increase investment or to improve the balance of payments situation. Thus, the wage and price agreement in March 1951 was designed to compensate workers for only half of the 10 per cent increase that had occurred in the cost of living. Again, in 1957, there was only a partial adjustment to a higher cost of living—this time through a restricted increase in family allowances instead of through wage increases.

In general, up to early 1954, wages followed prices with about a six-month lag, holding down real wages to close to prewar levels. There has been comparatively little “wage drift,’ payment of wages above the approved level being illegal in the Netherlands.12 Adjustments of wages to prices did not narrow (proportional) wage differentials for skill in the period 1947-55. Rigorous job evaluation and job classification have also prevented fictitious upgrading and similar devices to evade the standard wage levels. In 1954-56, wage pressures grew and substantial increases in real wages were permitted through higher wage rates and larger fringe benefits. Attempts were made to ensure that a fair part of the rise in wages in this period was met out of accumulated profits and current increases in productivity, without entailing a wage-price spiral. Greater flexibility in the wage structure was permitted in 1956, so that higher wages may be paid where labor is relatively scarce.


Wages and salaries in Norway are adjusted to changes in the cost of living by a system which seeks to avoid a rigid automatic link. Contracts renewed in the spring of 1956 included a reopening clause to the effect that, if in March or September 1957 the cost of living index should show an increase or a decrease of 6 points from the base figure of 150 (1949 = 100), either of the parties to the contract could request a revision of wages. No commitment was made regarding the amount of wage variation consequent on a 6-point change in the index, the exact amount being left to be fixed by negotiation between labor and employers’ organizations. If the contracting parties should fail to reach an agreement, the party that requested adjustment of wages (upward or downward) would be entitled within two weeks to dissolve the contract with one month’s notice. Such wage-reopening clauses are widely used because most wage contracts are made for a two-year period.

In early 1958, the labor and employers’ organizations agreed on a wage contract for three years instead of the customary two, and on raising to 160 (1949 = 100) the “ceiling” of the cost of living index on which wage adjustments are based.

The cost of living index used in such negotiations was based on family budget studies made between October 1951 and September 1952 covering 45 towns and 8 industrial centers situated in different parts of the country. It was a retail price index covering 370 items. The index was sufficiently broad not to be significantly affected by movements in only a few prices.


During World War II, Sweden had wage-price sliding scale arrangements on terms agreed between the Swedish Trade Union Federation (LO) and the Swedish Employers’ Confederation (SAF). Under wage-price agreements made during the war, the adjustment of wages to higher prices was seldom complete: it varied from nearly 75 per cent of the price rise in 1940 to about 50 per cent in the immediate postwar years. As a result of other types of wage adjustment, however, real wages were maintained in most industries at their prewar level. The LO-SAF agreement in January 1945 specified the wage adjustments to be made for each rise in the cost of living index. Wage adjustments for higher prices were usually made quarterly, but no downward wage adjustment could be made. Employers were entitled to cancel the agreement only if the cost of living had fallen to a level at which real wage rates would at least be equal to those prevailing in 1939.

At the end of the war, the Employers’ Confederation refused to accept any such sliding scale arrangements; they were accordingly discontinued after 1946 without significant opposition from the Trade Union Federation, which, in the interest of stability, agreed to exercise wage restraint in the immediate postwar years. Though wages were not linked to the cost of living index, attention was given to price changes in formulating wage demands. In some cases, “reopening clauses” were inserted in collective wage contracts. In general, wage increases were agreed annually by the employers’ and workers’ federations, with the Government acting, at times, as a regulating authority. It is estimated that over the entire postwar period the agreements have provided an annual average increase in hourly wage rates of about 5 per cent for male workers in manufacturing industries. In addition, overtime pay, increases in piecework, movements of labor from low paid to highly paid occupations, and the great demand for labor have resulted in a “wage drift” raising hourly earnings of male workers in manufacturing by a further 4 per cent annually.13 The period of relative price stability between 1952 and 1954 was accompanied by a more moderate rise in wages.

Partly because of higher excise taxes on some essential goods and some rise in prices of food, fuel and light, and rent, the cost of living rose in 1955 and again in 1956; this increased pressures for higher rates and revived interest in sliding scale arrangements. On February 21, 1957, an agreement was reached between the LO and the SAF providing for an increase in wage rates of 2 per cent or of 10 ore per hour (whichever is greater) for 1957, and a 2 per cent increase in piece rates and a 3½ per cent increase in time rates in 1958. The significant feature of this agreement was that it covered two years and was thus a reversion to the practice of longer agreements, which was usual before the war.

Because of the longer period covered by the agreement, the trade unions demanded a sliding scale provision to protect the workers against any sharp rise in prices. It was agreed, therefore, that if by November 1957 the cost of living index (1949 = 100) should rise to 150 (in March 1957 it was 143) the workers could reopen the contract to negotiate for a cost of living adjustment. The size of the adjustment was not specified, and if no agreement was reached either party might cancel the agreement on two weeks’ notice. Thus, any rigid linking of prices and wages was deliberately avoided. The agreement also provided that if the index should reach 150 as a result of fiscal changes, such as a new general sales tax or similar taxes, no wage adjustment could be asked for. Price increases resulting from such taxes would be deducted from the index to calculate the critical ceiling at which the contract could be reopened in November 1957.

United Kingdom

Sliding scales to adjust wages to the cost of living index were first introduced in the United Kingdom in certain industries toward the end of World War I. The system was gradually extended to other industries and services until in 1922 the total number of workers covered by such agreements rose to about 3 million, i.e., about 16 per cent of the labor force. In addition, wages in some branches of industry (e.g., the steel industry) were adjusted to the prices of the articles manufactured by the industries (for instance, the price of a ton of standard steel). The use of escalator clauses in wage agreements disappeared with the onset of the Great Depression. By 1939, however, the number of workers covered by such clauses was about 1½ million, i.e., about 8 per cent of the labor force.

During World War II, there was a further extension of sliding scale provisions. Though in some industries and services the sliding scale provisions in use at the beginning of the war were suspended or abandoned, the number of workers covered by sliding scales had risen by June 1944 to about 2½ million, or about 11 per cent of the labor force. Since the end of the war sliding scales have been abandoned in some important industries, and comparatively few new sliding scales have been introduced in other industries. Nevertheless, in April 1957, the total number of employees covered by sliding scale provisions was still about 2½ million (which at that time was only 10 per cent of the total labor force), as a result of increases in the number of workers employed in some of the industries which used sliding scales. The number of workers had increased in particular in the building industry, which employs about half of the workers covered by sliding scale provisions in the United Kingdom. Other important industries in which sliding scales are operative include mining and quarrying, glass, printing, iron and steel, and textiles.

While for particular categories of workers covered by sliding scale provisions a substantial percentage of the year-by-year increases in wage rates secured after the war was in the form of sliding scale adjustments, these adjustments were responsible for only a relatively small proportion of the wage rate increases received during 1945-56 (Table 4). Generally speaking, about three fourths of the increases in wage rates were agreed in voluntary negotiations between workers and employers. Of course, the rise in the cost of living has been an important consideration in wage negotiations, so that there is a strong informal link between the cost of living and wages.

Table 4.

Methods by Which Changes in U.K. Wage Rates Were Arranged, 1945-56

(In öre per hour)

article image
Source: Ministry of Labour Gazette (London, England), January issues, various years.

Of about 17 million wage earners, roughly 13 million were employed under collective agreements and another 4 million under statutory regulations.

Particulars relating to government employees were first introduced in 1948; therefore figures for 1945-47 are not strictly comparable with later years.

Wage increases relative to increases in the cost of living have been greater since 1953 than before, though the proportion of increases attributable to sliding scales has been somewhat less in the later period.

The collective agreements in which a sliding scale is usually included show considerable differences in the content of the provisions which link wages with price adjustments. Wage adjustments are usually made at intervals ranging from a quarter to a year. The movement required in the index of retail prices (the “cost of living” index used, as described below) before a wage adjustment is made ranges from 1 to 10 points. Most agreements provide for a wage adjustment of a specified amount per hour or per week for each change of 1 point in the price index. In the building industry, where sliding scales are most popular, wages have been revised annually and usually have varied |d. per hour for each variation of 3 points in the price index. In most industries where a sliding scale is used, wage earners are classified, usually as adult males, women, and juveniles, and different flat rate adjustments are applied to the cost of living bonus so that the differentials between the wages paid to these categories are not disturbed. The type of sliding scale actually used in any industry is decided, and may be changed from time to time, by collective bargaining within the industry.

From June 1947 until the end of 1956, sliding scale arrangements made use of the interim index of retail prices, which was based on an inquiry into household expenditure of about 11,000 working-class households in 1937 and 1938. This index was revised as from January 1952, on the basis of estimated consumption in 1950. A completely new index was introduced as from January 1956, broadening the categories of employees considered, reducing the weight given to food, and adding about 80 new items (including washing machines, secondhand cars, television sets, telephone rates, etc.). Most of the sliding scale agreements renewed during 1957 were based on this new index, with January 1956 as base.

United States

In the United States the sliding scale was used in some collective agreements in the early postwar years; the popularity of this instrument increased during the Korean boom and again during 1955-56.

In terms of the number of workers affected, and of its importance in setting a pattern for other industries, the postwar escalator clause embodied in a contract between the General Motors Corporation and the United Automobile Workers in May 1948 was the most notable. This provision was retained when the contract was renewed in May 1950. According to this agreement, wage adjustments might be either upward or downward; and since the cost of living declined at times during 1948-50, downward adjustments then occurred. During the first two years (ended May 1950) of the agreement, there were two upward adjustments and three downward adjustments under the escalator clause.

By June 1950, sliding scale provisions applied to about 1.5 million workers. With the Korean boom, the use of sliding scale provisions grew. At the end of March 1951, about 500 collective agreements covering 2.5 million workers contained clauses providing for the adjustment of wages to the cost of living. The coverage grew to a peak of over 3.5 million workers in September 1952. Thereafter, with relative stability of the cost of living, the number of workers covered by sliding scale provisions declined to less than 2 million by January 1955. The increasing use of longer-term wage contracts in 1955 and the tendency for prices to rise in 1956 again stimulated the use of sliding scale provisions in collective agreements. It is estimated that in early December 1956 a total of at least 3.5 million workers employed under union contracts and another 300,000 unorganized workers had wages linked to the cost of living. Most of the workers covered by these clauses belong to the railroad, automobile, basic steel, meat-packing, and electrical industries.

During the Korean war, the Wage Stabilization Board was partly successful in checking the widespread growth of wage escalation provisions. Such clauses were carefully scrutinized to make sure that they conformed to certain conditions. Specifically, in January 1951 the Board limited upward escalation to a 10 per cent rise in wages (changed to 12 per cent in August 1951) and insisted on escalation working downward as well as upward. It also ruled that, in the absence of escalator clauses, wages could nonetheless be adjusted every six months (whatever the period of contract) to changes in the consumer price index. With this wage-reopening provision (which unlike escalation avoided the risks of a downward adjustment) and the expectation of price declines after the Korean fighting, escalation provisions were omitted in many contracts which came up for renewal in 1953 and 1954.

In most collective bargaining agreements where sliding scales were abandoned, all of the accumulated cost of living allowances were absorbed into the basic wage rates. In many contracts which retained sliding scales, similar modifications were made so that part or all of the current cost of living allowances was added to the basic wage rates. Such amendments generally insulated the higher base wage from any downward adjustment if prices fell, although in some cases some of the earnings were omitted from the basic wage rates and this “float” could fluctuate with future changes in the consumer price index.

The wage agreements negotiated in 1955 or 1956 were generally for longer periods. Because of this, they included not only provisions for adjusting wages to changes in prices, but also provisions for deferred (scheduled) wage increases. In 1956, workers covered by the United Automobile Workers’ contracts received deferred increases averaging 6 cents to 7 cents per hour and a rise of 6 cents per hour under the cost of living escalator clause. A new three-year agreement in the steel industry in the summer of 1956 similarly combined forward wage claims (apparently justified on grounds of expected improvement in productivity) with sliding scale provisions linked to consumer prices.

The price index used is usually the Bureau of Labor Statistics revised index of consumer prices, which measures price changes for the country as a whole, although some use is made of local price indices. The base date to which changes in wages and prices refer is fixed by bargaining, each party trying to fix the base date at which the ratio of wages to prices is deemed to be most favorable. Most adjustments are quarterly, but there is an increasing tendency for half-yearly adjustments, particularly in longer-term contracts. Some adjustments are annual. Although a few wage adjustments are proportional, most are on a flat rate basis. Typical sliding scale clauses in effect at the end of 1956 provided for a 1 cent change in hourly wage rates for each 0.5 point change in the consumer price index, or a 2 cent change in hourly wage rates for each 0.9 point change in the index.

A large number of collective bargaining agreements which do not provide for sliding scale adjustment have various wage-reopening clauses permitting a reconsideration of the contract whenever the consumer price index rises beyond a certain point.


Mr. E. M. Bernstein was Director of the Research and Statistics Department of the International Monetary Fund from the date of the Fund’s inception until January 1958. The author of Money and the Economic System and of numerous articles in economic journals, he was formerly Professor of Economics in the University of North Carolina and later Assistant to the Secretary in the U.S. Treasury. He is now an economic consultant in Washington, D.C.


A survey of the use of wage-price links in various countries is given in Appendix II.


The normal growth in credit money is equivalent to the increase in real balances of currency and deposits which the public wishes to hold, at stable prices, after allowing for the increase in money resulting from additions to monetary reserves. Bank loans out of savings deposits are treated as loans by the public.


In a country with a fluctuating exchange rate, where international payments are kept in continuous balance through exchange rate movements, the effect of the excessive demand will be manifested in a rise in prices and a depreciation of the exchange rate.


Between March 1956 and September 1956, monthly earnings of workers in Rio de Janeiro are reported to have risen by 35 per cent. In the same six-month period, the cost of living is reported to have risen 10 per cent. The change in real wages computed on the basis of these data is prima facie improbable. Such a computed change could certainly not have been matched by a corresponding change in the standard of living of wage earners.


“Labor Market Developments,” Federal Reserve Bulletin (Washington) September 1957, p. 1006.


While the wage-price link is usually based on an index of the cost of living or of consumer prices, the adjustment of wages for productivity, to avoid an uneconomic wage level, should be tested by an index of wholesale prices of domestically produced industrial goods. For a discussion of this point, see Appendix I.


There are a few, but very few, exceptions to this rule. In some instances, the larger rise in consumer prices, compared with wholesale prices, probably reflects an improvement in the terms of trade.


Alternative “stabilization standards” are discussed at some length in E. M. Bernstein, Money and the Economic System (Chapel Hill, North Carolina, 1935), Chap. XVIII; see pp. 435-36 for a discussion of the proposal for seeking approximate stability of the price level of domestically produced industrial goods. The concept of approximate stability is not intended to include a gradual rise or fall in prices over a long period. Rather, it is a recognition of the fact that prices will fluctuate in the business cycle and should respond to changes in the payments position. It is not possible, therefore, to maintain perfect stability of any price index. “For all practical purposes it would be sufficient if changes in the price level could be confined to small variations about an accepted norm.”


As this reference to export capacity indicates, a deterioration in the real international economic position of a country—that is, an adverse change in the reciprocal demand for exports and imports—would require a fall in export prices relative to import prices. To maintain a proper balance of payments with a fixed exchange rate under such conditions, it would be necessary for wage rates to rise less than the average increase in productivity in industry, so that there would be some small decline in the wholesale price level of domestically produced industrial goods.


Mr. Sen Gupta, formerly economist in the Special Studies Division of the Fund, is a graduate of Calcutta University, and he also studied at Harvard University. He is the author of “Growth Problems in International Trade,” Public Policy, Vol. VII (Yearbook of the Graduate School of Public Administration, Harvard University).


The excess of actual wages paid over wage rates fixed by collective bargaining.


The increase in average hourly earnings beyond average hourly wage rates in industry was significantly lower in the Netherlands in the postwar years than in many other countries, e.g., Denmark, Sweden, and the United Kingdom. There was, however, a significant “wage drift” in particular sectors, resulting in part from the increasing use of piecework. The “wage drift” was particularly substantial in the building industry.


This is a rough average. Year-by-year classifications of wage increases by negotiation and by “wage drift” are given in Gosta Rehn’s article, “Swedish Wages and Wage Policies/’ The Annals of the American Academy of Political Science, March 1957.