Indexation of Wages and Salaries in Developed Economies
  • 1 0000000404811396 Monetary Fund

Indexation of wages and salaries occurs in response to a period of price inflation. It has tended to wither away during sustained periods of price stability.


Indexation of wages and salaries occurs in response to a period of price inflation. It has tended to wither away during sustained periods of price stability.

Indexation of wages and salaries occurs in response to a period of price inflation. It has tended to wither away during sustained periods of price stability.

To assess the advantages and disadvantages of indexation realistically, one must compare it with the likely alternative responses, such as frequent renegotiation of contracts, or a shift away from wage determination via formal bargaining between enterprises and unions under agreed procedures toward extemporary negotiations and industrial action on the shop floor.

It is hard to reach clear-cut conclusions concerning the advantages and disadvantages of indexation, since these must depend on the kind of situation that may be faced if indexation is not adopted, as well as on the widely different forms that indexation may take. Even the distinction between indexation and nonindexation is not clear. For example, should provision for the reopening of wage negotiations when the cost of living passes a certain level be considered as a form of indexation, or as a means of providing for wage bargaining at irregular, and probably shorter, intervals? Much discussion of indexation tends to be misleading because the assumptions made about wage behavior in the absence of indexation, or about indexation provisions themselves, are not generally applicable.

Section I of this paper discusses the theoretical case for indexation in the context of a different and probably more realistic model of wage determination than that implied by monetarist advocates of indexation. Section II describes wages indexation in practice, and the arguments for introducing or extending indexation in particular countries. It concludes that discussion of the economic case for indexation has tended to divert attention from the fact that wages indexation is primarily a means of mitigating the political consequences of rapid inflation, by avoiding the buildup of intergroup conflicts occasioned by changes in the distribution of income.

I. The Wage Bargaining Process and Indexation in Theory

price expectations in the wage bargaining process

The average rate of wage increase in an enterprise, an industry, or the economy as a whole is influenced by several factors, one of which is the actual or expected rate of increase in prices. Wage increases are also sought or conceded on the grounds of rising productivity, high profits, or scarcities of particular, or all, types of labor; and the level of wages in a particular sector may be raised as a result of strengthened organization of wage and salary earners there. These grounds for increases are not mutually exclusive but may be in part alternatives.

When wage bargaining relates to particular enterprises or industries, the rate of wage increase granted is likely to be influenced both by the actual or expected movement of the cost of living, which influences the size of wage claims, and by the actual or expected movement of prices for the specific output of the enterprise or industry, as this affects its profitability and the size of wage increases likely to be granted. These two influences do not necessarily correspond, for the economy as a whole. Not all domestic output enters directly into consumption, and consumption is not met exclusively out of domestic output. There are imports and exports, and outputs of capital goods (whose prices often fluctuate more widely than those of consumer goods).

The fact that in a “partitioned” labor market the influence of price expectations operates in part from specific price expectations to particular wage settlements has important consequences. With shifting patterns of demand and supply, it means that, at any moment of time, wage and salary earners will be in a position to secure above average gains in real income in those sectors where demand is strongest, or where productivity is increasing exceptionally rapidly.1 Over any short period, certain industries and professions will be able to increase their share in total value of output relative to their share in employment. Other sectors then have to accept a relative decline in the price of their outputs, if the average rate of price increase is not to accelerate, and the overall level of output and employment, associated with the given rate of monetary expansion, to decline. Thus, even when the authorities pursue a consistent monetary policy, demand conditions and price movements in particular sectors are liable to be influenced by the uncertain outcome of labor-management and supplier-buyer confrontations in other sectors; and possibly also by unintended changes in the monetary and fiscal situation brought about by changes in the rate of inflation and the level of output. Since in most economies there are marked differences in the degree of organization and bargaining power of wage earners as between sectors, the average rate of wage increase is likely to be higher or lower, depending on which sectors are experiencing relatively strong demands. These considerations suggest that even if a given rate of monetary expansion is maintained, demand conditions and price expectations in particular sectors cannot necessarily be foreseen with a high degree of certainty; and the average rate of increase in wages and prices and level of employment and output may remain continually liable to change.

the monetarist case for indexation of wages

Monetarist advocates of indexation focus on inflation as a phenomenon of a general movement of the price level induced by an excessive expansion of purchasing power. They assume that if a certain rate of monetary expansion is consistently maintained, the average rate of price increase will come to be correctly anticipated, with all prices and rates of pay increasing at about the same rate, regardless of whether indexation is explicitly adopted or not, and with employment stable at its “natural” level. When the rate of inflation is changing with changes in monetary policy, indexation of wages will be helpful as a means of securing a swifter adjustment of wage (and price) decisions to the new context, and so of avoiding “excessive” expansions or contractions of output, and distortions in the structure of production, arising from mistaken price expectations. They especially emphasize that when the authorities are seeking to reduce the rate of inflation, indexation will be helpful as a means of avoiding as much unemployment and dislocation of production as would occur if wage (and price) decisions were based on incorrect expectations of the continuance of the previous rate of inflation.2

Now the notion that the economy tends to “steady-state equilibrium” implies that behavior in the real economy adjusts to whatever rate of monetary expansion is maintained, so that the rate of monetary expansion ultimately affects only the absolute level of prices and does not influence the level of real output and employment. The idea that the real economy itself is repeatedly subject to powerful “autonomous shocks” is alien to this approach. Hence, it obscures the vital practical question whether it may be more feasible to maintain a higher level of real output and employment over time, in the face of autonomous shocks, with one rate of monetary expansion rather than another.

In the monetarist view, the economy may be subject to shocks, but since their nature and direction cannot be foreseen, policy should not be designed to deal with them. The potential impact of shocks is limited by the self-equilibrating properties ascribed to the real economy: the fact that the economy does not seem prone to explosive disequilibrium is taken to support the assumption that adjustment mechanisms are operating to restore equilibrium at some “natural” rate of unemployment. Now there are, of course, many constraints and frictions that mitigate and slow down the impact of shocks, and contribute to the apparent stability of the real economy (including the slow reaction of economic agents, imperfect knowledge, resistance to change and lack of mobility, and reactions to uncertainty). One of these checks and balances is the existence of a monetary system. Instead of presupposing self-equilibrating properties of the real economy, it is preferable to imagine it as a slow stream (moving with its own momentum under the force of changes in technique and political events) whose actual course is determined in part by the interplay between the changing “real” factors and the degree of monetary constraint that is maintained.3

In a changing real environment, where competitive markets and freedom of entry for new enterprises and wage earners are by no means general, even if a certain rate of inflation had been maintained and the movement of prices in general had come to be foreseen with some degree of confidence, the movement of particular prices and/or wage rates would still be apt to accelerate (or decelerate) as a result of “real” factors (the economy being continually subject to the largely unpredictable effects of technical change, changing scarcities of resources, changes in the degree of monopolistic organization of business and labor in different sectors, and to the impact of changing political forces both within the economy and within the world at large). It would in fact be difficult for confident price expectations and general agreement concerning the probable rate of inflation ever to become established. Changes in the strength of price and cost pressures in particular sectors, together with the degree of constraint implied by the constant growth of the money stock, would continuously be likely to produce an acceleration (or deceleration) in the average rate of price increase and a tendency for the level of employment and real output to fall (or rise) at the going rate of monetary expansion. Furthermore, the possibility of the changing situation provoking a change in monetary policy could not be ruled out.4

This view of the process of inflation poses a question concerning wages indexation that does not arise in the monetarist view of things, namely, whether wages indexation would render the movement of prices in general more sensitive to price and cost pressures developing in particular sectors of the economy. If so, it could make the task of the monetary authorities more difficult by increasing the risk that changes in the real economy could set off a faster rate of inflation and rising unemployment when the same rate of monetary expansion was maintained.

the rate of monetary expansion and relative price flexibility

We may find it easier to understand the causation and effects of indexation by first considering the influence of wage bargaining on the flexibility of relative prices in the economy, and how this is likely to be affected by the rate of inflation.

There are grounds for thinking that it may be easier to maintain a high level of employment with a moderate rate of monetary expansion that permits a gradually rising average price level than with either a lower rate of expansion, requiring a stable or falling price level if employment is to be maintained, or a faster rate of expansion that permits rapid price inflation. In other words, contrary to the monetarist view, there is likely to be a lasting trade-off between the rate of monetary expansion and the level of unemployment experienced over time, at least over a certain range of rates of expansion.

The reason for thinking so is that a moderate positive rate of price increase is likely to be associated with greater flexibility of relative wage rates at the going level of employment, rendering the average movement of prices, and hence the level of employment, less vulnerable to above average price and/or wage increases in particular sectors than under conditions of strict price stability or rapid inflation.

Consider an economy with a fixed labor force, and a technically determined trend growth of productivity of π˙ per cent per annum. The money stock is permitted to expand by per cent per annum, and the existing level of employment is maintained when prices rise on average by π˙ per annum. Suppose that the desired level of employment has been maintained with this rate of inflation for some time, but one sector then is able to secure a much larger increase in the price of its output than previously, as a result of some autonomous change in the structure of demand or supply, or as a result of more monopolistic organization. This will tend to cause some weakening of demand in other sectors at the going rate of monetary expansion. If the desired level of employment is to be maintained with the same rate of monetary expansion, prices in other sectors must rise less rapidly than before.

As long as moderate rates of inflation are experienced, wages and salaries are fixed, by contract or convention, in money terms and are adjusted for price changes by wage bargaining at relatively infrequent intervals (of a year or longer). Money wages are inflexible downward; rates of pay are not likely to be cut unless the level of employment is substantially reduced for some considerable period of time. But upward adjustments of wages can be postponed if an enterprise or industry faces a weakening of demand. Hence, the effective floor to the distribution of rates of wage increase without a marked rise in unemployment is the maintenance of agreed money wages. The fact that prices are known to be rising by, say, 3 or 4 per cent per annum has little if any effect in preventing the postponement of wage increases in sectors faced with a weakening of demand. So, if certain industries succeed in pushing up their prices at a higher rate and increasing their share in total expenditure, there is scope for reducing the rate of wage increase in sectors facing weakening demand, in spite of the downward inflexibility of money wages.

If, however, the monetary authority were to limit the growth of money strictly to that consistent with overall price stability at the expected rate of growth in productivity, price increases in some sectors would require price reductions in other sectors, which almost certainly could not be achieved without unemployment, given the downward inflexibility of money wages.5

Thus, over some range, a higher rate of monetary expansion associated with a higher (but still moderate) rate of inflation may increase the possibility that an acceleration of price or wage increases in one sector can be offset by postponement of wage and price increases in other sectors. The average rate of price increase and level of employment may therefore be less apt to change as a result of increased price or cost pressures in particular sectors of the economy. In other words, the authorities will be less likely to face the need to choose between validating a faster rate of price increase or tolerating a higher level of unemployment. If demands and supplies are price elastic, one may suppose that when there is greater short-term flexibility in relative wages and prices between industries, shifts in the structure of output and employment can be achieved with less unemployment and the maintenance of a higher level of real output over time.

Friedman and the monetarists assume that whenever a certain rate of inflation has persisted for some time, a situation resembling that under price stability will apply, as all prices and wages will be swiftly adjusted for the correctly anticipated rate of price increase. It would seem, however, that there is an important difference, in this respect, between a rapid and a moderate rate of inflation, because a moderate rate of price increase does not cause wages to be fixed in real terms by explicit or implicit indexation.

Experience suggests that a continued moderate rate of inflation does not lead to widespread adoption of indexation clauses, or to frequent renegotiation of wage rates in order to maintain the real value of wages. Neither the extra effort of frequent negotiations, nor the real cost involved in frequent changes in pay scales (as in any prices), is justified when prices are rising on average by 1 or 2 per cent a quarter. Furthermore, there is likely to be resistance to the idea of indexation, both on the part of employers, who fear that it may expose them to difficulties should the prices of their product tend to rise less fast than consumer prices in general or raise the size of wage increases that they are forced to concede, and on the part of union leaders, who see it as detracting from their power and influence.

By contrast, the onset of rapid inflation quickly leads to greater insistence on the maintenance of real, as opposed to nominal, wages. That is to say, wage fixing in real terms seems to depend more on the rate of price inflation experienced than on the length of time that inflation has persisted. Contrary to the view of some monetarists, notably Friedman, the weakening of money illusion in wage bargaining seems not to be a gradual process of adjusting expectations for rising prices but rather a swift response to the more obvious effects of large price and wage increases. Under rapid inflation, both the loss of purchasing power associated with fixed money wages under annual, or less frequent, wage settlements and the conspicuous relative gains of those wage earners who have just secured new wage contracts become sufficiently great to justify the extra effort and cost of wage bargaining at more frequent intervals or of adjusting wages automatically by indexation.

Under widespread and effective indexation (providing full compensation for changes in prices at short intervals) a sharp acceleration of price increases in one sector, such as was considered earlier, would be apt to provoke a faster rise in wage costs in other sectors, even while an increase in the degree of monetary constraint, owing to the faster rise in money incomes and prices, tended to cause a weakening of demand. Thus, indexation would appear to make it more difficult to maintain the existing level of employment with the existing rate of monetary expansion.

The situation need not be very different without explicit indexation, however, if the realization that real wages had declined significantly led to greater militancy and the refusal to permit postponement of wage settlements, even in the face of a weaker demand situation for the enterprise or industry and if some groups were insisting on swifter compensation for price increases outside or under the usual bargaining procedures.

In sum, rapid price inflation quickly leads to greater insistence on maintaining the real value of wages in the short run, by indexation or other means, and thereby lessens the relative flexibility of labor costs in different industries. The average rate of price increase becomes more likely to accelerate with any acceleration of the rate of price and wage increase in specific sectors of the economy, and it becomes increasingly difficult to maintain an existing, desired level of employment without validating an accelerating rate of price increase by faster monetary growth. Thus, it is probably easier to maintain a moderate rate of inflation for some considerable period of time than either a stable price level or a double-digit rate, of inflation. Rapid inflation would not seem likely to come about through a gradual acceleration of creeping inflation. It would usually be the result either of some quite unduly rapid expansion of monetary demand relative to real output potential or of some unforeseen catastrophic decline in real output potential owing to natural or political causes.

the effect of wages indexation

Indexation is merely one of the ways in which greater insistence on maintaining the real value of wages in the face of rapid inflation makes itself felt. It is very difficult to say, a priori, what effect, if any, indexation has, compared with the situation that would exist in its absence. There are, however, certain differences in its impact compared with the achievement of a similar degree of compensation for price increases by more frequent wage negotiations and more insistent wage bargaining: (i) indexation is more automatic, (ii) it continues until abrogated, (iii) it can be foreseen with greater certainty. Because of (i) and (ii), indexation may cause nominal wage increases to be less responsive to the existence of substantial unemployment and real wages to be better maintained when unemployment is heavy but prices are still increasing. If indexation provides for adjustment with a lag of six months or more, indexation may tend to prolong the impact of rapid price inflation and to delay the response of wage settlements to an easing of demand conditions. Because of (iii), indexation may increase the certainty with which producers in any particular sector can count on money wage rates, prices, and, probably, money income in other sectors rising in the immediate short run, in response to current price increases, even though the rise in money incomes might require to be validated by a faster rate of monetary expansion.

It is often suggested that if a macroeconomic wage equation shows that the average level of wages responds to changes in the price level with a coefficient of unity, money illusion is absent, and indexation is unlikely to have an important effect in enhancing inflation. This need not be true, however.

The fact that wages on average keep up with prices does not, of course, imply that all wages do so continuously. It may result from a continuing combination of rising real wages in sectors of rapidly increasing productivity, or of high profitability associated with increasing demand, or in sectors where union power is increasing, combined with wages falling behind the average rise in prices in sectors of low growth, declining demand, and weak labor organization. The introduction, or extension, of indexation to these latter groups (for example, as consequence of the adoption of an indexed minimum wage) may then cause their wages to rise more rapidly than they otherwise would. It is not possible to say, a priori, what effect this will have on the average rate of wage increase. But it seems probable that the initial effect will be to promote a more equal rise of money wages throughout the economy around a higher average rate of increase (since the power of less favored groups to ensure that their wages rise more closely in line with other wages will probably not be offset by a reduction in the power of the most favored groups to secure above average wage increases).6

The rate of price and wage increase that is actually experienced must of course be influenced by the degree of financial constraint imposed by the authorities’ control over the budgetary situation and the supply of money and credit. But it is not obvious that it will be an easy, or politically acceptable, task for the authorities to prevent wages in general rising faster if indexation has the effect of increasing the rate at which wages rise in the weaker or less strongly organized sectors of the economy.

In an economy such as the United States, characterized by wide differences in the degree of competition prevailing in various sectors and a large nonunionized segment of the labor market, it would appear that if all wages and salaries 7 were automatically indexed on the previous rate of inflation, the effect of generalized and “full” indexation of wages would be to make inflation less painful and socially disruptive, but faster and more difficult to brake.8 A reduction in the rate of monetary expansion aimed at halting inflation might well produce a larger immediate rise in unemployment, loss of output, and economic dislocation (bankruptcies, companies going out of business, depressed areas, etc.) than if indexation did not apply. More unemployment would tend to be associated with a shift to tighter financial conditions precisely because there would be less effect in holding down wages and prices in the sectors most affected by weakening demand.

The argument that indexation may serve to lessen the strength of cost/push is most persuasive when labor is strongly organized both for wage bargaining at the industry or plant level throughout the economy and in national politics—with the result that the maintenance of full employment commands a high priority among policy objectives.9 In such a case, any well-publicized wage settlement, much in excess of the current rate of price increase plus the expected average rate of growth in productivity, will be likely to unsettle price expectations, even when it is believed that the authorities intend to maintain the existing degree of financial restraint and rate of monetary expansion. If wage earners in other sectors succeed in getting similar wage increases, the rate of price increase will be apt to accelerate, and the level of employment to decline if the rate of monetary expansion is maintained. It is not clear that the monetary authority will be able to avoid validating the higher rate of inflation, given the strength of political pressures for maintaining full employment. To be sure of protecting themselves against a decline in their relative real wages and a possible fall in their absolute real income in these circumstances, wage earners have to seek wage increases that match the highest rate of increase recently granted. That is to say, they are bound to bargain for matching increases in nominal wages, given the uncertainty over the rate of price increase that may eventuate.10

The fact that wage earners will try, as far as possible, to protect themselves in this manner against the unforeseeable outcome of particular groups’ efforts to push up their real wages then causes the average rate of wage increase to exceed the growth in real output. In this “scenario,” generalized cost/push is, in part at least, due to uncertainty; it does not necessarily reflect the situation that wage and salary earners have “nonnegotiable demands” for increases in real income much in excess of the growth in real output. Protection, via indexation, against the eventual rate of price increase, whatever it may turn out to be, may therefore offer a means of preventing an exceptionally large settlement from setting off a round of similar increases. Indexation on the movement of prices in general should be less inflationary than wage increases matching the highest rate of wage increase recently accorded, where that settlement may reflect such factors as exceptionally rapid growth in productivity in the sector, unusually favorable demand conditions there, etc. Thus, while it is usually correct to regard indexation as tending to reduce the flexibility of relative wages, in circumstances where great rigidity already exists, indexation may sometimes provide a means of securing increased flexibility. Where wage bargaining over much of the economy is centralized, as in the Scandinavian countries and in Ireland and Israel, indexation is likely to prevail, but this particular benefit will not apply.

In general, it may be helpful to consider developed market economies as approaching one or the other of two polar cases—the economy in which the authorities are politically able to maintain a restrictive monetary policy; full employment is not an overriding political goal, labor is not strongly organized in all sectors of the economy, and all wages do not tend to go up in line; versus the economy in which the authorities have political difficulty in maintaining a restrictive monetary policy; full employment commands a very high priority, labor is strongly organized in virtually all sectors, and wages tend to move upward closely in line (“relativities” being zealously guarded).

II. Indexation of Wages in Practice

present incidence and policy options

It is no accident that recent debate on indexation of wages and salaries has been strongest in countries where indexation was relatively unimportant, notably, the United States, the United Kingdom, the Federal Republic of Germany, and Australia. Policy options with respect to indexation of pay are usually limited. In a context of rapid price inflation, the authorities are not likely to be willing or able to resist the spread of some form of cost of living clauses. One may expect that a period of rapid price inflation will result in widespread indexation unless either wage and salary earners are sufficiently strongly organized for wage bargaining purposes to secure compensation for price increases by means of frequent renegotiation of contracts, wildcat strikes, etc., or labor is politically too weak to insist on some form of indexation. When indexation is widespread, it is only under exceptional circumstances (such as a balance of payments crisis under fixed exchange rates imposing a need to secure a lowering of relative costs by a large devaluation)11 that the authorities will be able to secure the abandonment of the system, as occurred in France and Finland. However, indexation may wither away naturally during a prolonged period of price stability.

The apparently haphazard division between countries having widespread or slight recourse to indexation, shown in Table 1, can be explained in these terms. In the majority of Western European countries, rapid price inflation during the late 1940s led to widespread indexation, which in most cases exists to this day, having usually been strengthened during the last few years.

Table 1.

Extent of Indexation of Pay in Developed Countries1

article image

Supplementary notes on the extent and form of wages indexation in most of the countries listed here are given in the Appendix.

The countries that had slight recourse to indexation fall into three groups: (i) the Netherlands, Sweden, and the United Kingdom, which experienced less than average open inflation in the immediate postwar period, and in which a politically powerful labor movement was able to protect real wages by other means than indexation, such as rationing, food subsidies, and price controls; (ii) the United States and Canada, where the postwar inflation was least severe and where wage earners were a less strongly organized force in national politics than in many European countries; and (iii) Austria, the Federal Republic of Germany, and Japan, where the labor movement was exceptionally disorganized and politically weak in the immediate postwar period.

characteristics of wages indexation in practice

In practice, indexation of pay diverges in significant ways from the concept implied in theoretical discussions of the case for generalized indexing, or “monetary correction.”

Indexation is rarely, if ever, complete in the sense that it ensures that all wages and salaries are automatically adjusted at short intervals for changes in the overall price level and that wage negotiations concern changes in real wages. Indexation provisions seldom apply to all workers or cover the whole pay packet; they often provide for adjustment only with substantial delay, usually on the basis of some less than comprehensive index of the movement of prices in general. Furthermore, there is usually provision for price declines to be treated unsymmetrically from price increases—or ignored. Indexation usually applies to negotiated rates of pay but not to additional earnings resulting from wage drift at the plant level, overtime pay, bonuses, special allowances for conditions of work, etc. Sometimes, as for example in Italy, indexation is applied to specific basic rates of pay, which have come to account for a progressively smaller share of total wages as time goes on.

The price index on which adjustment is based frequently excludes important items of expenditure, notably housing.12 Rental costs are sometimes based solely on rent-controlled housing. Highly processed items of increasing importance in household budgets (such as pharmaceuticals, consumer durable goods, and newly introduced products) and personal services are generally underrepresented in the basic cost of living indices.13

Both the partial adjustment of wages and the incomplete coverage of price changes for which compensation is provided are symptomatic of a different underlying objective from that of monetary correction. In many systems of indexation, the basic aim is not to protect the real value of wages in general but to protect the consumption levels of the lowest-paid workers, by ensuring that increases in the cost of basic essentials are reflected in wage rates.

Partial adjustment and the incomplete coverage of the basic price series mean that indexation generally has less powerful effects than might be expected in theory, so that in practice indexation does not tend to cause whatever rate of price increase may result from rising wage costs in the economy to be perpetuated.14 However, the biased coverage of price changes may have the effect of predisposing the economy to a faster rate of wage increase in certain circumstances.

The heavy weighting, in consumer price indices, of foodstuffs and staple items with a high commodity component in their total value meant that tight commodity supplies and sharply rising prices, such as occurred in 1973–74, would have a disproportionately sharp effect in inducing wage increases irrespective of the employment situation,15 the automatic adjustment of wages in turn compounding the rise in food prices and leading to further pressure on wage rates. Thus, the effect of indexation, contrary to the view advanced by advocates of monetary correction, has been to make the general level of prices especially susceptible to the influence of supply conditions in particular sectors.

Perhaps the most striking way in which indexation in practice diverges from the system propounded by its advocates is in respect to the time dimension. Indexation provisions frequently call for adjustment to be made annually. With such a lag, indexation may well tend to retard rather than to hasten the effect of deflationary policies, whereas if an expansion is under way, the lag is likely to cause the adjustments to be superimposed on wage increases that have already been granted rather than to be substituted for negotiated increases. Similar considerations apply if adjustment is called for after a large “threshold” price increase is exceeded. A balancing of considerations of the cost of frequent changes in pay packets versus the disadvantages of delayed adjustment suggests that quarterly changes may be the most favorable interval when the timing of adjustments is fixed.

Generally speaking, if indexation is not to be a disturbing factor, it is desirable that adjustments should be made frequently and smoothly, and preferably not on the same basis and simultaneously in all sectors (since lumpy and easily foreseeable changes in the mass of wages provide a favorable opportunity for price increases and tend to promote a relaxation of price discipline, producing greater subsequent wage adjustments than smoother changes). Differing provisions for adjustment under wage settlements for particular industries on the Belgian model may be preferable to widespread adjustments triggered by a single indicator, such as a change in the minimum wage, as was the practice in France during the 1950s.

the case for introducing or extending indexation in some countries

Recent interest in indexation has been fanned by the determined advocacy of Professors Milton Friedman and Herbert Giersch. It has not always been sufficiently emphasized that, while presenting a general case for indexation, those authorities were actually prescribing indexation as a remedy to the highly specific problems of a particular economy at a particular time.

Friedman’s case for increased indexation of wages in the United States

Indexation of wages and salaries is a comparatively minor element in Friedman’s proposals for monetary correction, but it is important as an illustration of certain pitfalls in the monetarist approach to the problem of inflation.

When Friedman propounded his ideas on indexation in 1974,16 a round of large wage increases appeared to be in prospect in the United States following the lifting of price/wage controls, which had had the effect of sharply reducing real wages. His immediate goal in advocating the wider use of escalator clauses was to enable enterprises to retain the benefits of long-term wage contracts, in obviating the considerable costs involved in frequent wage bargaining,17 but to avoid the risk of being stuck with excessively large wage increases should the rate of inflation moderate. His case for indexation essentially concerned the introduction of escalator clauses into long-term wage contracts coming up for settlement when more restrictive financial policies were about to take effect and a decline in the rate of inflation was in prospect.

Friedman considers it desirable that escalator clauses should be incorporated in a wider range of wage agreements, but he sees this merely as a temporary extension of existing practice, expecting the use of clauses to disappear once inflation is brought under control. “Private use of escalator clauses is an expedient that has no permanent role, if government manages money responsibly.” 18

His discussion of the advantages of indexation in the specific case of the General Motors Settlements of 1967 and 1970 is somewhat unsatisfactory, owing to its neglect of the fact that changes in the growth in total spending are associated with changes in the pattern of real demand. It is odd that, although he ascribes the inflationary expansion of the money supply to “the sovereign’s attempt to acquire real resources,” he then seems to forget the effects that this has in enhancing expenditure on certain real resources, with its consequent impact on the structure of relative prices and wages.

Thus, Friedman considers that the fact that General Motors paid lower real wages than expected, when the consumer price index and aggregate expenditure rose at a faster rate than anticipated after 1967, “was a stimulus to General Motors and no doubt led it to produce at a higher rate than otherwise.” 19 This does not necessarily follow, however. It is conceivable that sharply higher prices for items such as housing, services, and medical care, which contributed to the acceleration of the rise in consumer prices, were associated with shifts in the composition of aggregate expenditure tending to limit the rise in demand for automobiles. The steep rise in prices of steel and non-ferrous metals under the impact of the Viet-Nam war must also have had an influence in lessening GM’s incentive to produce more cars.

In the context of, possibly sharp, short-term changes in the real economy, it cannot simply be assumed that a fall in the real wages paid would necessarily cause General Motors to expand output more than it would otherwise, or vice versa. Suppose, for instance, that real wages had risen because of a fall in the price of food. Would this be likely to cause GM to produce at a lower rate? As wage earners buy more cars than do farmers, and as the demand for food is inelastic, this situation would seem more likely to induce an increase than a decrease in GM’s output.

In practice, not only unanticipated inflation but also unanticipated changes in government expenditure, financed out of taxation, may pose problems under long-term wage contracts. Suppose that the U.S. Government had financed the rise in the expenditure after 1967 out of progressive taxation, or a sales tax on consumer durable goods, or a gasoline tax. The price expectations of General Motors and the United Automobile Workers might have been fulfilled, but General Motors would probably not have been able to maintain the relative price of automobiles and the expected scale of output.

Friedman’s support for indexation apparently rests on the implicit assumption that the economy is in steady-state equilibrium with relative wages and prices at their equilibrium levels when an unanticipated inflation or deflation occurs. This is evident from his tendency to regard all changes in relative prices as distortions requiring to be corrected.20 If, however, one supposes that deflation follows after a period of inflation associated with changes in the structure of real demand, it is at least conceivable that the changes in relative wages and prices produced by the response to inflation in existing wage settlements for some sectors, and the adjustment to deflationary conditions in new wage settlements for other sectors, may tend to reverse the changes in relative wages and prices associated with the changes in structure of demand during the preceding period.21 More important, as shifts in relative prices are required, it is not clearly desirable to ensure, by widespread indexation, that wages in all sectors move more closely in line.

The case for wages indexation in the Federal Republic of Germany

In the Federal Republic of Germany, where the union leaders have been concerned to avoid provoking a marked inflation of costs, indexation is seen as the answer to the threat of wildcat strikes.22 It represents a possible solution to the union leaders’ problem of satisfying the interests of their membership and of avoiding the loss of influence over them to radical groups, while at the same time protecting the solidarity of the union movement and the Social Democratic Party now in power. One commentator has described the “political imperatives” as follows:

… the party cannot allow inflationary wage increases for blue-collar industrial workers and at the same time retain the allegiance of the new-middle-class voters so critical to its electoral victories; union leaders cannot supply too much flank protection for the government in the face of revolt from the rank and file; and both sides find the unemployment associated with economic slowdown (as in the 1966–67 recession) an intolerable way of fighting inflation—the party for political reasons … and the unions because unemployment is the most dreaded of all their specters.23

Professor Herbert Giersch has also pointed out that the inclusion of cost of living clauses in wage contracts promises the unions a means of avoiding the opprobrium of being responsible for large increases, while at the same time protecting them against “the outbreak of wildcat strikes following upon an unanticipated decline in consumer purchasing power.” 24 The union leaders reacted to the wave of wildcat strikes in 1969 with inflationary wage demands in 1970, and the same phenomenon occurred again in 1973.25

Thus, the advocates of indexation in the Federal Republic of Germany see it as a means of restoring the controlling role and moderating influence of the union leadership in wage bargaining, as well as a means of combining the advantages of long-term wage contracts with flexible increases in nominal wages.

Giersch argues that

… ever since flexible exchange rates and swift adjustments of parity values have become the general practice, we have seen that the central bank is fully able to apply the monetary brakes. … If escalator provisions are forbidden and not in use, there is a strong tendency to build the inflation rate of the past into the new collective agreements, especially in those cases where no loss in employment has occurred as yet.26

But to avoid major repercussions on physical output and the level of. employment, it is necessary that people keep their price and income demands flexible. These demands should be based not on inflation rates of the past, but rather on those lower future inflation rates which monetary policy will tolerate. This is exactly what will be achieved by concluding agreements containing cost-of-living escalator clauses.27

The case for threshold agreements in the United Kingdom

Arguments essentially similar to these were put forward by the Trades Union Congress when advocating the introduction of threshold agreements in the United Kingdom at a time of accelerating expectations of price inflation in 1971. The agreements were to stipulate that if, during the period covered by a wage settlement, the increase in retail prices exceeded a certain “threshold” figure, wages would be increased by a set amount related to the additional price increase. The threshold agreement would provide a guarantee of protection against a high rate of price increase, enabling the wage settlement to be negotiated on the basis of a lower rate of increase.28 The United Kingdom’s experience with threshold agreements, later in Section II, describes the situation since the eventual introduction of those agreements in late 1973.

The introduction of indexation in Australia

In Australia, indexation is seen as a means of strengthening the influence of the Australian Conciliation and Arbitration Commission over the general movement of wages and salaries and as a means of countering the growing tendency for collective bargaining and strikes relating to “over-award” wages.29

The Arbitration Commission rejected indexation proposals put forward by the unions and the Government in the 1974 National Wage Cases. In November, the Government announced that it would again press for indexation in the 1975 National Wage Case. In the meantime, in order to provide business with a respite from wage claims based on the rise in the cost of living, personal income taxes would be reduced on incomes below the level of average weekly earnings, as a form of compensation for price increases in the December quarter.30

The 1975 Award 31 provided for wage rates to be raised in line with the 3.6 per cent increase in the consumer price index (CPI) during the March quarter. But the Commission was not prepared to adopt an integrated wage fixing package, including indexation, for subsequent quarters without first seeing how the situation developed. Instead, it scheduled a further hearing, in July, to hear submissions on the following proposals.

Award wages should be adjusted quarterly for changes in the CPI. The Commission should hold short hearings four times a year to hear submissions for and against such adjustment. The form of indexation (and the extent of compensation against price increases) should be decided by the Commission on the basis of submissions to it, subject to the proviso that an increase of less than 2 per cent in the CPI would be fully compensated for; an increase of less than 1 per cent would be carried forward to the following quarter(s). In addition, each year the Commission should consider what increase in total wages should be awarded for increased productivity. The only other grounds on which pay increases would be justified would be changes in the nature of the work, skill, and responsibility required, or in the conditions under which work was performed (i.e., changes in “work value”).

The Commission considers that generally appropriate relativities between and within awards have now been established, on which indexation could be applied; but there may be isolated cases where “catching-up” may still be called for. The compression of relativities that has occurred in recent years should not be regarded as grounds for special increases, but could be considered in connection with decisions on the form of indexation or the distribution of gains in productivity. In the course of its conciliation activities, the Commission should guard against contrived work value agreements and other methods of circumventing the indexation scheme.32

Provided that these conditions for wage increases apart from indexation had been substantially met, the Commission would recommend the adjustment of award wages for price increases in the June quarter.

In explaining the ruling, the Commission pointed out that the case for indexation rested on three grounds: it would help to moderate inflation; it would provide an equitable basis for wage adjustment; and it would help to lessen the heat generated in industrial relations. The first proposition rested on the notion that the expectation of further large price increases led to large wage claims, which in turn produced the expected rate of price increase. The self-fulfilling element in the inflationary process could be eliminated by indexation. The opponents of indexation—the private employers and the four conservative State Governments—contended that an attempt to maintain the real value of wages, which had risen greatly in excess of prices and productivity in 1974,33 would add to inflation or increase unemployment or both. Professors Whitehead and Gates argued that elimination of price expectations would probably not greatly reduce wage demands, but would impart an additional automatic element to the rise in wages and would accelerate inflation.

For many years, the Commission had rejected claims for automatic adjustment on the ground that it preferred to keep wage movements under its direct control. However, with rapid price increases and large wage increases outside the National Wage cases, it had recently been difficult to avoid erosion of the real value of National Award rates. The unions and the Government did not rely on indexation as a means of reducing the current rate of inflation but regarded it as “an integral part of an equitable counter-inflationary policy.” 34 Indexation could be applied in National Wage cases only if pay increases apart from National Wage awards and indexation provisions were kept within strict limits. “Otherwise it is difficult to escape the conclusion that, even if it removed inflationary expectations from wage claims, indexation carries a high risk of accelerating inflationary tendencies if it adjusts excessive wage increases automatically for price increases every quarter.” 35

The Commission had been convinced that the final submission of the ACTU marked a new and positive approach to the interrelated questions of the method of wage fixing and indexation. The unions had argued that

The Commission can only gain control over wage fixation if it has something new, different and positive to attract wage and salary earners back in from the field. … the most positive action which the Commission could take at the present time would be to introduce wage indexation. … We submit that given the consensus which exists on the status of the Commission … there is a very real chance that wage indexation will work to the satisfaction of all if introduced by the Commission now, and the Commission using its status clearly states the terms on which its introduction is made.…

The introduction of wage indexation as envisaged by the A.C.T.U. would bring back immediately under the Commission’s umbrella, control over the majority of wage increases—[compensation for] price increases would be automatic for all and would not be sought on an industry basis—productivity hearings would be held annually and succinctly and would replace the larger national wage cases which now exist. Work value cases as traditionally defined would be arbitrated by individual members of the Commission. Other increases—which would be “unusual” or “rare and isolated” would be under the control of the Commission.36

The Commission emphasized that its decision to support indexation in principle had been materially influenced by the ACTU’s statement that work value and other increases would need to be small, and would “have a negligible effect on the total wages and salary bill.” 37 It had to be understood, of course, that such increases would include those sought by consent (i.e., settlements reached by collective bargaining and merely ratified by the arbitration authorities). The coming months would provide an opportunity for testing the viability of the suggested conditions under which indexation should operate.

In September 1975, the Commission ruled that wages should be increased by 3.5 per cent, in line with the increase in the CPI for the June quarter.38 In November, the Commission refused an application for an increase of 0.8 per cent, to cover the rise in the CPI in the September quarter, standing by its earlier decision that an increase of less than 1 per cent should be carried over to the next quarter.

In the event, the wages indexation policy proved to be rather successful in the context of heavy unemployment during 1975. The Commission’s guideline that wage increases over and above price compensation should be granted only in narrowly circumscribed “work value” or “relativities,” cases was, in the main, observed; and the percentage increases in award wages and average earnings in 1975 were less than half as large as in 1974 (when they reached about 38 per cent and 28 per cent, respectively).

When the turmoil which prevailed in wage fixation in 1974 is recalled, the development and implementation of this policy must be regarded as a considerable achievement, due to the willingness of all parties to industrial matters, and in particular the bulk of the trade union movement, to support the indexation system. It is also undoubtedly due in part to the depressed conditions in the labour market.…39

However, unless further policies are implemented to weaken the price/ wage spiral, the moderation of wage pressure in 1975 may prove to be temporary. Writing in November 1975, the Institute of Applied Economic and Social Research suggested 40 that if indexation was to succeed as a means of reducing inflation and supporting economic recovery in 1976, it would need to be supported by efforts to bring downward pressure on wage/price adjustments by reductions in indirect taxes and public sector charges; a commitment to an early introduction of income tax indexation;41 tax relief for enterprises; some form of price restraints; and a drive to promote the financing of increased public sector deficits by the sale of securities to the nonbank sector.

The future of indexation is uncertain at this time. The CPI for the fourth quarter 1975 was 5.6 per cent higher than the previous quarter—the largest increase since 1951. In the National Wage case now being heard, the Government has argued that the wage increase should compensate for only half the 6.4 per cent price increase between the second and fourth quarters of 1975. It is not clear whether the Commission, unions, and employers remain committed to indexation.

countries’ experience with wages indexation

Opinions on the usefulness or danger of wages indexation have differed enormously between countries, or in the same country at different times. In practice, views on the consequences of indexation tend to be colored by the success or failure of economic policies pursued in the country. In the United Kingdom, Denmark, Finland, Iceland, and Israel, indexation has been associated with severe economic problems and has been held to have had damaging results; but in Norway, and until recently in Belgium and Luxembourg, where the general economic situation was less difficult, wages indexation has been generally regarded as a necessary and useful device.

This concluding section concentrates on the recent experience of the United Kingdom, the United States, Norway, and Denmark, and comments briefly on the long period of automatic adjustment of the basic wage in Australia. Supplementary notes in the Appendix provide information on other countries and further details for the United Kingdom and the United States.

The United Kingdom’s experience with threshold agreements

The danger of wages indexation in the context of a sharp autonomous deterioration in the terms of trade is clearly illustrated by the recent experience of the United Kingdom, where indexation was introduced at an inopportune moment and in an unfavorable form.

Threshold agreements were initially introduced for one year in November 1973 with the immediate purpose of gaining union support for Stage III of the Government’s statutory price/wage policy, following a 90-day standstill on price and pay increases, and a period of strict control lasting about six months.42 When the agreements were adopted, it was reckoned that “there was a reasonable chance that retail prices would rise by only about 7 per cent during the coming year; import prices, it was thought, would level off, and national output would be rising at about 3 per cent a year.” 43 However, the National Institute had pointed out earlier that, given the uncertainty about import prices, threshold agreements would be something of a gamble. It had been suggested that the price index used for the threshold should exclude import prices, as had been tried in Denmark, but it was doubtful whether such an index would be accepted.44

If, by giving some insurance against the consequences of a rapid price rise, the basic wage increase figure can be brought down; and if,” as a consequence of that low basic figure, the threshold price rise is not reached; then obviously there is a gain. On the other hand, if the reduction in the basic figure is only marginal, and the rise in import prices exceeds the forecast, anti-inflationary policy would probably lose … any attempt to compensate for real income losses arising from adverse terms of trade only results in higher prices still.45

The riskiness of the gamble is revealed in a chart prepared by the National Institute (presented as Chart 1). This shows that before the start of Stage III, the retail price index was rising by nearly 7 per cent a year, although Stage II had had an obvious effect in keeping down the increase in retail prices, despite the extremely sharp rise in import prices occasioned by the world boom in commodity prices. Thus, even without the very steep further rise in import prices that occurred during Stage III, an accelerated rise in retail prices was to be expected that would trigger threshold payments. The effect of Stage III was to delay the impact of steeply rising import prices and the acceleration of retail price increases (which was actually enhanced by the raising of indirect taxes early in 1974), and to hold down the increase in average earnings for about six months. Thereafter, the increase in average earnings accelerated very steeply, exceeding the rise in retail prices and causing the rate of price increase to continue to accelerate despite the slowing down of import prices.

Chart 1.
United Kingdom: Retail Prices, Average Earnings, and Import Prices, 1966–74 1
Source: Economic Trends.1 Adapted from Chapter l, “The Economy in 1974,” National Institute Economic Review (February 1975), p. 31.

The form of indexation permitted in Stage III was likely to increase the severity of subsequent wage pressures. The combination of a pay-rise ceiling with provision for a liberal flat rate adjustment for each 1 per cent rise in prices above the 7 per cent threshold seems unfortunate. The fact that the ceiling was announced before the acceleration of price increases occurred led to a widespread feeling that the conditions upon which it had been based had been breached; and the sharp decline in real earnings during the first half of 1974, together with the express provision for automatic adjustment, contributed to strengthen wage claims later in the year.

The flat rate adjustment was dictated by the wish to afford the strongest protection against inflation to the lowest-paid workers. The increase of 40 pence a week was equivalent to about 1 per cent of average weekly wage and salary bill, and represented considerably more than 1 per cent of the lowest rates of pay. Consequently, an increase of 1 per cent in retail prices caused a rise of more than 1 per cent in below-average wages.46 The operation of threshold clauses was therefore apt to provoke a faster rise in wages than in prices by inducing wage claims for the maintenance of traditional relativities. In sum, in the British context of fragmented wage bargaining and jealous attention to wage relativities, the specific form of indexation that was adopted introduced a very dangerous element of instability into the system by making wages both more swiftly responsive to price increases and apt to rise more rapidly than prices. With a fluctuating exchange rate, this situation was likely to put pressure on the exchange rate, enhancing the rate of price increase and provoking a still faster rise in wages.

Escalator clauses in the United States

A recent study 47 has emphasized the comparatively minor direct role of automatic adjustment for price increases in the United States, where less than 10 per cent of all nonagricultural employees were covered by escalator clauses at the beginning of 1975. Contrary to the common view, automatic adjustments under such clauses do not, on average, fully compensate for price increases but have covered no more than half the increase in the CPI in recent years. Past increases in living costs, if substantial, however, constitute an important element in union negotiated contracts; and unless there is a prospect of relative price stability, efforts are made to take the expected rate of price increase into account in the wage settlement, especially if the contract is to run for more than a year without provision for reopening. Substantial changes in living costs have also provided a powerful impetus to wage increases in the larger nonunion segment of the economy, owing to considerations of equity and the desire of employers, if the labor market is at all firm, to maintain morale, minimize turnover, and possibly forestall efforts at unionization. Consequently, even in 1974, when living costs rose at a rate unprecedented in peacetime, escalator clauses accounted for only about one fifth of the average increase for all workers under major collective agreements, for substantially less of all wage increases in the union sector, and for a small fraction of wage increases throughout the economy.

A more detailed empirical study of the effect of long-term wage agreements incorporating escalator clauses at the time of the Korean war48 concluded that during periods of rapid price increase, automatic wage increases “led” but did not exceed changes in closely related wages; that automatic adjustment tended to lessen the effects of short-term restraining influences in industries where wages tended to lag; and that they increased the sensitivity of the general wage level to price changes.

Soffer suggested that when agreements both with and without provision for automatic adjustment for price increases existed in bargaining units subject to similar demand conditions, escalator clauses tended to have a positive influence on the average rate of wage increase. This came about as follows. During a period of rising prices, it was almost impossible to negotiate wage increases as frequently and for similar amounts as those occurring under escalator clauses. Thus, “time inequities” or temporary divergence in rates of wage increase became widespread. “The more rapid the inflation, the greater are the immediate and expected lags of the relatively infrequently negotiated wages behind the automatically adjusted wages. These lags are rather quickly eliminated by supplementing the smaller increases,” 49 because smaller wage increases than those already granted are likely to be unacceptable to the union and its members. However, if the catching-up with automatically adjusted wages entails relatively larger current increases in other wages, this will be perceived by wage earners who are covered by escalator contracts as an adverse shift in their relative wages, requiring escalator increases to be supplemented by additional negotiated increases. With this kind of interrelation between automatic and negotiated wage increases, the influence of automatic adjustment on the movement of wages in general could not be assessed by comparing the movement of escalated and nonescalated wages, or the share of automatic increases in the average overall increase in wages.

Norwegian and Danish experience with indexation

Explicit compensation for increases in the cost of living has almost invariably constituted an element of central wage settlements in the Scandinavian countries, except in Sweden. In Sweden, price movements have influenced the national “frame” agreements more indirectly, the expected increase in foreign prices entering into the calculation of the “room” for wage increases in the open sector. Of the other four countries, Norway’s experience shows how the degree of compensation for price movements can provide a useful element of flexibility in nationwide bargaining over increases in nominal income in a favorable political and economic setting. Denmark and Finland illustrate the difficulty of controlling the rate of inflation, when indexation is pervasive, if there are strong unresolved conflicts over the distribution of income and a divisive political situation preventing the election of a government with a clear majority and mandate for action. Finland and Iceland also illustrate the problems posed by widespread indexation of wages in an economy subject to marked fluctuation of export prices and the terms of trade.50

Broadly uniform provisions for indexation apply throughout the private sector in both Norway and Denmark, owing to the dominant influence of the wage agreements negotiated, usually for a two-year period, between the central organizations of the trade unions and the employers’ associations. The system of wages indexation has been in many ways similar in the two countries, but in Norway the Government has had a much stronger influence on indexation decisions, and has been more closely involved in the process of income determination; indexation has been accompanied by frequent recourse to price freezes, price controls, and use of subsidies to keep the price index below the critical threshold that calls for a new round of negotiations.

Indexation is calculated to have resulted in about 70 per cent compensation for the rise in the cost of living in Norway.51 Up to 1966, indexation clauses generally provided for the so-called semiautomatic adjustment, that is to say, for reopening of negotiations if the cost of living rose by more than a stated number of points. More recently, fully automatic adjustment has provided for increases, of a certain absolute amount, to be granted for each point rise in the index over an agreed period, frequently one year.

Despite a year-long price freeze between late 1970 and late 1971, marked inflation of costs and prices from 1969 to 1972 stemmed from the excessively strong demand conditions that resulted from the relaxation of fiscal restraint during the upswing in 1969–70, and from resistance to the subsequent increases in taxation. Sharply higher indirect taxes and charges for public services,52 associated with a marked increase in old-age pensions and rising public expenditure, led to higher wage demands and increased wage drift. The phenomenon of egalitarian wage settlements followed by efforts by the higher-paid groups to restore their relative position in absolute earnings by local or individual bargaining also became apparent.

During 1973 the Government sought to reduce the rate of inflation by direct action in the field of wages, prices, and foreign exchange. The package deal between the Government and the labor market parties during the second quarter provided for a much lower degree of compensation for price increases than would otherwise have emerged. Wages were to go up by 0.45 for each 1 point rise in the CPI, compared with 0.70 in the previous settlement. As a quid pro quo, the Government undertook to subsidize food prices, to postpone increases in public service charges, to cut other public expenditure so as to offset higher subsidies, and to further increase subsidies if the CPI rose by more than 5 per cent between March and September, which it did. As a result of these measures, and of the effective appreciation of the kroner by about 9 per cent over the year, the rise in consumer prices was significantly slowed during 1973, declining from an annual rate of 9 per cent over the first half year to 6½ per cent over the second half. Firm demand management, perhaps coupled with a greater willingness to accept an “egalitarian” development of wages promoted by the measures taken to protect real wages, appears to have resulted in relatively slight wage drift during 1973.

All the main wage contracts and income arrangements for farmers and others were due for renegotiation in the second quarter of 1974. As a break with established practice in the preceding years, there was no centrally negotiated frame agreement, although intensive contacts between the central organizations of trade unions and employers continued. There was a general consensus, on the part of unions, in favor of reducing the wage contract period from two years to one year.

In connection with the wage negotiations, in March 1974, Parliament voted a reduction in wage earners’ contribution to social security, increased food subsidies, and a raising of social security payments and of children’s allowances for families with more than two children. Two-year wage settlements for the metal and engineering industries, which were broadly followed in other manufacturing industries, provided for an increase of 6.3 per cent from April 1974, a further 5 per cent increase from April 1975, and included an automatic hourly increase of NKr 0.10 for each percentage point increase in the CPI between March and November. This represented roughly 80 per cent compensation on average, but substantially less on higher wage earnings. The degree of compensation on the basis of the September 1975 index was to be negotiated later. The wage settlement for government employees incorporated similar indexation provisions.

Commenting that the weakening of price stability since 1971 had apparently been less pronounced in Norway than in most Western European countries, the Organization for Economic Cooperation and Development (OECD) suggested that government “policies must have played an important role in cushioning the repercussions of an inflationary development abroad on the domestic price and wage formation process.” 53 While the relatively favorable inflationary experience was assisted by the appreciation of the kroner in 1973 and 1974, a more important factor was the use of tax relief as a means of reducing nominal wage claims and of selected subsidies to keep down food prices.

In Denmark, government attempts to curb the inflationary impact of indexation in the face of balance of payments difficulties and declining export competitiveness have precipitated repeated political crises. A crucial factor predisposing the economy to inflation has been the unresolved conflict between the general union movement, which is strongly committed to a policy of wages “solidarity,” aiming at a more egalitarian wage structure, and the organizations of higher-paid wage and salary earners who are concerned to maintain their existing differentials. In practice, workers in the engineering and machinery industries have been prepared to participate in the central wage settlements, which provide fuller protection against price increases and proportionately larger contractual increases on lower wages, because they were confident of being able to restore their relative differentials via wage drift.

The two-year (or sometimes three-year) wage agreements negotiated between the Federation of Trade Unions (LO) and the central organization of employers (DA) cover less than half of all employees, but they set the pattern for agreements covering other workers. In the negotiation of the central wage agreement and other wage agreements, three components of wage increases are considered: the contractual increase in the basic wage rate; the cost of living adjustment; and the expected amount of wage drift, which includes wage increases resulting from the raising of piece rates and other increases in wages negotiated at the industry or plant level.54 The guarantee of automatic cost of living adjustment has generally aimed at a flat rate of compensation, providing for hourly wage rates to rise on average by about 0.7 per cent for each 1 per cent rise in the price index,55 but providing almost full compensation for the lowest-paid workers. However, since relative wage levels have been maintained by wage drift, the Danish system has in practice provided virtually full compensation against price increases for all wage and salary earners, as well as state pensioners and the recipients of unemployment and social security benefits. Civil servants have generally received full price compensation on the ground that they do not benefit from wage drift.

In 1967, the Government imposed a temporary freeze on profit margins on domestic sales, and secured the agreement of the trade unions and civil servants’ organizations to forgo the cost of living “step” increase expected to become due in January 1968, on condition that a one-time tax would be levied on unearned incomes. But the tax bill was defeated, and a general election followed.

Following a marked deterioration of the balance of payments in late 1969, the authorities again sought to influence the outcome of the pending wage negotiations. The Government proposed that employees should waive the next three “steps” of automatic adjustment in return for a lowering of income taxes, but this was rejected by the unions. When it became plain that at least one “step” increase would be required by the September level of the index, the Government raised the value-added tax from 12.5 per cent to 15.0 per cent and began to subsidize the employers on a temporary basis for the effects of automatic adjustment on their wage costs, by assuming part of the employer’s obligations for contributions to the state pensions fund.

The 1970–71 wage settlement, reached after mediation, again provided for 70 per cent average compensation for price movements by automatic adjustment. It was stipulated that automatic adjustment of pay in the public sector should be renegotiated if the terms of trade deteriorated markedly. Subsequently, it was agreed that one third of the cost of living adjustment on higher salaries would be postponed until 1977.

The “frame” agreement signed in March 1973 called for “standard” or “minimum” wage rates to be raised by 0.40 kroner per hour for each 3-point increase in the wage regulating index. This sum represented about 2 per cent of the average hourly wage rates but about 2.5 per cent of the lowest rates of pay. When it appeared that the automatic indexation formula would trigger a “three-step” increase in March/April 1974 and an even larger increase in September/October, the Government undertook to subsidize employers for 50 per cent of the adjustment becoming due in March, the cost being met out of a compulsory savings scheme levied on higher incomes. Price control was to be used to ensure that wage costs covered by the Government were reflected in lower price increases.

Faced with the prospect of further massive adjustments to wages in early 1975, the Government proposed the adoption of a general price/ wage freeze. The unions refused on the grounds that the freeze was likely to be more effective in preventing wage increases than price increases, and the Government was defeated on the issue.

Collective bargaining talks failed to produce a new frame agreement, and the state mediator’s draft proposal was rejected by both the unions and the employers. In order to avoid a major industrial conflict in an already difficult economic situation, the Government decided to impose the mediation proposal by law. The March 1975 legislation provided for continuance of automatic adjustment for about two thirds of price increases in the private sector, triggered by a larger price increase than previously, and for adjustments in the public sector and for salaried employees fixed in absolute amounts rather than in percentage terms.56

Danish experience illustrates how difficult it is to determine whether the existence of indexation per se has had an influence tending to enhance or to lessen cost pressure. Economists advising the Economic Council concluded in 1973 that wage increases under the two-year agreement would have been larger without indexation, because the union leadership had informed them that they would then have insisted on larger contractual increases. It is conceivable, however, that the employers might have resisted the larger wage claims in that case. In fact, indexation does not seem to have played a key role in the inflationary process. The essential problem has been lack of an effective constraint, to force the wage bargaining parties to reach agreement on a noninflationary average movement of wages, under a series of minority governments, which have been unable to enforce financial control or to resolve the competing claims of the general union movement and of other interest groups.57

Automatic adjustment of the basic wage in Australia

Seen in a longer perspective, the adoption of an effective system of indexation may in some circumstances help to make the economy less subject to cost pressure by retarding, or preventing, the development of a militant union movement, or of wage bargaining organizations at the factory level, or the growth of powerful unions in the public sector. This seems to have been true in Australia, where from 1922 to 1953 the pay of the majority of wage earners was adjusted upward (or downward) for changes in the purchasing power of the basic wage component.

Under this system, each worker in effect received a fixed minimum real wage plus additional purchasing power, not adjusted for price changes, if he qualified for a margin for skill, etc. Automatic adjustment served to keep the general level of wages and prices moving in line with those of other countries, but problems arose when import prices moved very differently from export prices, or when both were subject to extremely sharp short-term fluctuations, as in the early 1950s. The automatic adjustment of wages meant that an autonomous inflationary (or deflationary) stimulus produced a swift chain reaction on domestic costs and prices. The goal of the system was not price stability but near-stability of real wages and flexibility of money wages and prices.58

Movements of the price level in both directions were tolerable because almost all incomes were adjusted for changes in the price level, and, as Keynes remarked in 1923, “a change in the value of money, that is to say in the level of prices, is important to Society only insofar as its incidence is unequal.” 59

Australian experience can thus be seen as an illustration of Friedman’s thesis that while

widespread escalator clauses would make it easier for the public to recognize changes in the rate of inflation, would thereby reduce the time-lag in adapting to such changes, and thus make the nominal price level more sensitive and variable … But, if so, the real variables would be less sensitive and more stable—a highly beneficial trade off.60

It is, however, a revealing special case, since pressures by highly organized groups of workers, in a favorable bargaining position to secure increases over and above the increases in the basic wage component of pay, were restrained by the operation of the compulsory arbitration system. (One evidence of this is the fact that the wage structure that emerged was narrower than that in most other countries.) Thanks to this restraint, indexation did not tend to produce a faster rate of inflation than that experienced in other developed countries, and was not associated with chronic balance of payments difficulties as in some less successful cases.

Discussion of the economic case for indexation has tended to obscure the fact that indexation is primarily a means of mitigating the political consequences of rapid inflation, and of lessening intergroup conflicts occasioned by changes in the distribution of income. The principal economic benefits stemming from indexation are likely to be the avoidance of industrial unrest, frictions, loss of morale, and unproductive efforts in wage bargaining processes that would otherwise occur, and the establishment of conditions more conducive to long-range planning by enterprises and a political climate more favorable to business than might otherwise exist.

Even when there is evidence that wages indexation may have contributed to aggravate the rate of price inflation, indexation may also have been an important element in ensuring the survival of a democratic, free enterprise system. In Finland, France, and Italy, a divisive political situation in the 1950s was rendered less explosive by indexation of wages. It has been rightly emphasized that “it will not be possible to successively apply and abolish indexation in the light of economic policy. … It is too closely tied up with tradition and the system of industrial relations.” 61 However, the decision to introduce indexation will often be dictated by overriding “short-term imperatives” of the immediate political situation.

APPENDIX: Provisions for Indexation of Wages and Salaries in Selected Countries 62

belgium and luxembourg

Automatic adjustment of wages for changes in the cost of living was introduced in the mining sector in 1920 and had become widespread in Belgium by the late 1930s. It was discontinued in 1940 but was reintroduced in 1948, at a time of strong external demand pressure. About 95 per cent of all wage and salary earners are covered by automatic adjustments, based on the CPI.

Indexation has been accepted by employees as a means of guaranteeing the real level of negotiated wages, and by the employers as a means of avoiding industrial unrest. However, problems arose following the steep rise in world commodity prices and the effective devaluation of the Belgian franc vis-à-vis the deutsche mark and the Netherlands guilder in 1973. During 1974, the unions began to question whether indexation on the basis of the existing CPI guaranteed the maintenance of real purchasing power, while the employers defended the existing system. In several sectors, negotiations took place over provisions for special allowances for high living costs or for speedier adjustment. The possibility that indexation may aggravate the wage/price spiral has received increasing attention.

Guidelines for real wage increases are determined in national wage negotiations held every two years between representatives of the principal employers’ organizations and trade unions. The Government is barred by law from intervening, unless there is a threat to economic or social stability. The negotiations result in an accord interprofessionnel de programmation sociale, which serves as a model for sectoral labor contracts, with guidelines for increases in real wages and nonwage benefits. The adjustment of wages for price changes is established at this level. Enterprises then have some scope for modifying the sectoral agreement on an individual basis.

The exact method of adjustment varies from sector to sector, being determined by collective agreements. The reference periods for calculating changes in consumer prices vary from one to three months; adjustment is usually contingent on the change exceeding 1 or 2 per cent, or 1 or 2 percentage points in the index as calculated; sometimes there is provision for a one-month delay before wages are increased.63 In some cases, notably the banking sector, adjustment provides for a larger percentage increase in wages than in the CPI, presumably to cover a faster rise in other items, such as housing and progressive taxes. Indexation of pay in the public sector was introduced by law in 1960, and the same provisions apply to social security payments (including private pension schemes) and to the wage scales used in calculating social security contributions. Since August 1971, government employees and pensioners have been eligible for a 2 per cent increase in wages, salaries, and allowances whenever the two-month average level of consumer prices rises by 2 per cent, the increase becoming effective with a delay of one month. This level of consumer prices then forms the basis of calculating the next 2 per cent increase in prices, and so on.

The calculation of the official CPI is supervised by a joint commission representing employers and unions; any change in the index has to be approved by the Commission. Nevertheless, the index is not comprehensive: rent is excluded, heating oil has only recently been included, and drugs are not covered.

In Luxembourg, a uniform system of adjustment prevails throughout the economy. Provisions concerning the indexation of public servants’ pay, established by law in 1963 and modified in April 1972, also apply to all collective agreements for the private sector, and to the minimum wage and social security payments. Between the end of 1962 and April 1972, wages and salaries were adjusted when the quarterly level of the CPI had risen by 3.5 percentage points; subsequently, the “threshold” was reduced to 2.5. As in Belgium, the index does not cover rent.


There has been growing use of escalator clauses in long-term wage contracts in recent years. It is estimated that cost of living adjustments now apply to about 60–65 per cent of workers in manufacturing, and to about 20–25 per cent of all workers that are covered by major settlements—that is, to about 5 per cent of the total nonagricultural labor force. It was estimated in 1974 that automatic adjustment would result in raising wages much less than in line with the CPI, owing to the large proportion of contracts that provided for adjustment only if the rise in prices exceeded a certain threshold.


Indexation was gradually extended to all employees, as well as to the government-guaranteed incomes of farmers between 1947 and 1967, but was abandoned in 1968, when it became obvious following a large devaluation that the system was bound to lead to severe inflation and to undo the benefit of the devaluation. At the time of the devaluation, most sectors were covered by three-year wage agreements that had been concluded in 1965. Wages in the metal industry—a typical case—were to be increased by 3 per cent in January 1968, and by 3½ per cent in June; increases in the cost of living in excess of 3 per cent were to be fully compensated for at the beginning of 1968. Agricultural prices for the crop years 1967–68 and 1968–69 were governed by an act providing for biannual adjustments for changes in the price of agricultural imports and farmers’ costs. Almost all rental, construction, and insurance contracts, and many financial contracts, were also indexed.


Indexation of wages became important in France in the late 1930s when many collective wage agreements were negotiated. A March 1938 law on conciliation and arbitration provided for possible revision of the terms of settlements when the official cost of living index had risen more than 5 per cent. Arbitrators frequently made use of this provision to raise wage rates in line with the cost of living.

Indexation was widespread during the 1950s, when the minimum wage (salaire minimum interprofessionnel garanti—SMIG) was automatically adjusted for changes in the cost of living, and many collective agreements provided for the wage rate to be adjusted in line with SMIG. At the end of 1958, all indexation of wages, other than the SMIG, under statutes or collective agreements, on the level of prices, wages in general, or the SMIG, was forbidden by law; this remains so, apart from the institution in January 1970 of the new minimum wage allowing for growth (salaire minimum interprofessionnel de croissance—SMIC).

The SMIC is automatically increased when the cost of living index for urban households rises by 2 per cent, and is also revised at least once a year, on the advice of a special Joint Commission, based on the growth in average real per capita income as shown in the national accounts and current economic conditions. In August 1974, the SMIC was raised to a level that was 23.1 per cent higher than a year before, while the CPI has gone up by 14.6 per cent. The increase affected 800,000 workers directly, and an additional one fourth to one third indirectly; it is estimated to have raised the average wage rate index by about 1 percentage point.

An increasing proportion of collective agreements in the private sector provided for automatic renegotiation in the event of a marked rise in prices or a sharp change in the level of activity—this remains legal. In the public enterprise sector, notably, the electricity and gas industries, agreements concluded in 1969 and since renewed provide for automatic increases in wage rates in the event of a predetermined marked change in the value of the gross domestic product at current prices and in the level of production of the industry. Settlements for coal mining and railways have guaranteed an increase in real wages (generally of the order of 2 to 2½ per cent a year) by providing that if the increase in the price index is larger than foreseen, wage rates will be raised proportionately. A similar principle applies in the civil service. An increasing number of settlements in the private sector have included similar clauses, but up to now only about 10 per cent of wage contracts contain such clauses.

federal republic of germany

Under the Monetary Law of June 1948, as interpreted since 1961, any indexation of the nominal value of monetary assets is forbidden without express authorization by the Deutsche Bundesbank. The application of this law to wage settlements has been questioned as being contrary to the principle of free collective bargaining. In practice, the law has been taken to prohibit automatic adjustment clauses but not provisions for automatic renegotiation of contracts when the cost of living index exceeds a certain level. Beginning in 1963, in connection with the conclusion of long-term contracts covering more than one year, clauses of this type were introduced in a limited number of cases.


Wage increases are determined under a national agreement between employers and the Congress of Trade Unions, which applies to the private sector and to public employees whose pay is subject to collective bargaining. Agreements at the industry level are concluded under this master “frame” agreement.

Automatic adjustment was introduced in the national agreement of December 1970. Under this agreement, covering 18 months, there was to be an additional 4 per cent rise in wages during the last 6 months, to be supplemented by a further absolute increase if the rise in the cost of living over the first 12 months of the agreement had exceeded 4 per cent. This supplement was set at 15 pence per week for each 1 per cent additional increase in the cost of living. The same arrangement was maintained in the following national agreement, of July 1972.


Wages comprise four elements: the basic wage, cost of living allowance, fringe benefits, and wage drift. The basic wage is determined separately for each industry, largely reflecting productivity differences. The cost of living allowance is negotiated directly between the General Federation of Labour (Histadrut), a trade union covering 90 per cent of employees, and the Manufacturers’ Association. It represents a certain percentage of the basic wage up to a given maximum, depending on the rise in consumer prices. From 1965 to 1972, the cost of living adjustment was made once a year at the beginning of the year, but in 1973 increases of 14.3 per cent and 10.3 per cent took place in January and July; in 1974 a 14.8 per cent increase in January to compensate for the 26 per cent rise in the cost of living during 1973 was followed by a further 6 per cent when subsidies were cut in February. As in the previous cost of living increases, these were free of tax and were paid in full on basic wages of up to I £ 700 a month, and at lower rates on higher wages and salaries.


Cost of living adjustment was introduced in certain industries soon after World War II, and the system was gradually extended, in different forms, under the national wage agreements relating to agriculture, commerce, banking and insurance, and government service. The great majority of wage and salary earners in the private sector, and of government workers, are now covered by some form of sliding scale.

In the private sector, adjustment, usually at quarterly intervals, applies to the part of earnings that is distinguished as “indemnification for the cost of living.” This is a fixed sum that depends on the wage earner’s qualifications and age. Except in banking and insurance, where the value of each point of the sliding scale is established as a percentage of the total earnings of the worker, the compensation for price changes applies to less than half the present level of earnings. For each increase of 1 per cent in the cost of living index (1957 = 100), wage earners receive a sum equivalent to 1 per cent of the average wage for that type of work in 1957. (There are 10 different levels of adjustment in industry—5 for manual workers and 5 for clerical workers—and 8 levels in commerce.) 64 In the industrial sector, the absolute increase for skilled workers is about one third larger than for the unskilled; in commerce, it is more than twice as large. In the public sector, the value of each point increase is the same for all workers and is very low—about the same as for an unskilled worker in industry.

The special cost of living index on which wage adjustments are based is weighted according to the average expenditure of a family of four in 1957. Movements in the index are studied by a national commission representing the unions and employers’ organizations.

The system of adjustment has recently come in for criticism because, as the current level of the cost of living has risen greatly relative to the 1957 base, adjustments are required for much smaller percentage changes in the current cost of living. In order to limit the frequency of adjustments and the psychological repercussions caused by this effect, the Minister of Labor has asked the unions and employers to revise the system.


For a considerable period after World War II all increases in wages were subject to government sanction. No formal link existed between wages and prices, although changes in the cost of living were taken into account in wage decisions. In 1951, wages were prevented from rising in line with the cost of living in order to improve the balance of payments and to increase investment, and in 1957 there was only partial adjustment of family allowances.

Indexation of wages was not an issue until 1965, when the first medium-term wage settlements were drawn up. The question of indexation was referred by the Government to the Economic and Social Council, which judged it to be justifiable; a series of agreements of two to three years’ duration, providing for indexation, were signed. At the end of 1969, the Council recommended the insertion of indexation clauses even in one-year wage contracts. By 1971, practically all collective agreements provided for some automatic adjustment of widely varying form. In some cases, the initial wage increase took no account of the expected rate of price increase, and the adjustment for price increases was to take effect only after the period covered by the contract; in others, the initial increase took some account of the expected rise in prices and the remaining adjustment was made at the end of the contract period. In certain cases, the adjustment covered only a certain agreed percentage of the rise in prices. The adjustment sometimes called for equal percentage changes in all wages, sometimes for changes of the same absolute amount. In 1973, more than half the 13–14 per cent increase in wages was due to cost of living adjustments.

With the acceleration of inflation, the method of equal percentage adjustment tended to produce a rapid widening of wage differentials in nominal terms, which was unacceptable to the unions, which are concerned to secure a progressive equalization of wages. The general wage agreement for the metallurgical industry in April 1973 established that full compensation should be given for cost of living changes on wages of less than f. 28,000 a year; that compensation at half the rate of increase in prices should be given on the next f. 6,000; and that compensation at one fourth the rate of increase in prices should be given on wage earnings in excess of f. 34,000. The actual agreements signed under this accord resulted in 49 per cent of wage earners receiving full percentage compensation, and the others receiving varying absolute increases subject to a ceiling or floor.

The cost of living index used for adjustment measures the cost of expenditure of wage earners’ households. As in other countries where indexation is widespread, the index is studied by a joint committee representing the unions and employers’ organizations.

Wages and salaries in public service are indexed on the movement of wages in the private sector, as measured by the average of collectively negotiated wage rates. A provisional wage increase is granted at the beginning of each year, and the discrepancy between this and the movement in wage rates is measured at the end of the year, involving the payment of arrears.

The legal minimum wage is also indexed on the movement of collectively negotiated wage rates in the private sector; it is adjusted twice yearly, on January 1 and July 1, on the basis of the level of wage rates applying on October 31 and April 31. The Minister of Social Affairs may set a higher level for social reasons; and he has the power to raise the minimum, after consulting the Economic and Social Council, if the wage structure should change as a result of collective agreements providing for larger percentage increases to the lowest-paid workers.

In January 1974, the Government adopted a statutory incomes policy under which, inter alia, all wages, salaries, and fringe benefits were fixed at the level in effect at the end of November 1973, with the addition of flat rate increases of f. 15 a month (about 2 per cent of the average basic wage) in January and April. Negotiations between employers and the unions having again failed, the Government decreed a third increase of f. 15 a month, and a 3 per cent increase in compensation for price increases, to be paid three months after expiration of the old contract. Further compensation for price increases was to be made six months after expiration, subject to a deduction of 0.3 per cent for price increases resulting from increases in indirect tax and charges for medical care. As a result of these measures, negotiations between employers and employees could affect only fringe benefits.

No central wage agreement was concluded in 1975. Negotiations at the industry level resulted in an increase of about 13 per cent in average wage cost per employee, owing mainly to compensation for price increases carried over from 1974 or coming up in 1975. The Government intervened to reintroduce a statutory wages policy and imposed a wage freeze during the first half of 1976. An adjustment of 4.5 per cent for price increases in April-October 1975 was not affected. Automatic cost of living adjustment is being reconsidered, and the decision concerning adjustment in the second half of 1976 will depend on developments in the first half year.


Escalator clauses have been used since the early 1960s and now apply to almost all employees. In the private sector, collective agreements generally prescribe that new wage negotiations take place when the CPI has risen by 2.5–3.5 per cent. In the federal civil service, when the CPI has risen by a given number of points, wages are increased by determined amounts, fixed according to wage levels.

united kingdom

Escalator clauses have a long history in the United Kingdom. Sliding scales sometimes related to the cost of staple articles of consumption and sometimes relating wage rates to going prices for basic articles produced by the industry were a common feature of wage agreements in the nineteenth century; they served to bring about a reduction of wage rates in times of falling prices as well as to maintain real wages in times of inflation. Sliding scales were common during the early 1920s and declined in importance during the interwar years. They regained importance during World War II and applied to about 2½ million wage earners in 1945, falling to about 2 million by 1952. Sliding scales were virtually eliminated between 1967 and 1970 following the Incomes Policy White Paper of November 1966, which was strongly against automatic adjustment. After the collapse of incomes policy, the Trade Unions Congress began to advocate the adoption of threshold agreements, notably in its 1971 economic review, and they were eventually introduced, as part of Stage III measures, for one year in November 1973. (See The United Kingdom’s experience with threshold agreements, in the text.)

united states 65

During the Great Depression, the Roosevelt administration cut the salaries of federal employees by 15 per cent in 1933, in line with the decline in the cost of living. When the cost of living started to go up, several large concerns (including Standard Oil, General Electric, and U.S. Steel) introduced cost of living clauses. However, by 1941 only 3 per cent of companies had followed their example.

Indexing was reintroduced in 1948 after a wage freeze during World War II and rapidly gained ground, applying to some 3½ million workers in the early 1950s. During the 1960s, the number fell to about 2 million, but it has risen rapidly in recent years, exceeding 4 million in 1973, 5 million in 1974, and 7 million in 1975.

Escalator clauses are confined largely to major wage contracts (covering at least 1,000 workers). In major contracts, coverage fell from more than 42 per cent in 1958 to less than 20 per cent in 1964 and had risen to 39 per cent by January 1974. In manufacturing, more than half (56 per cent) of the workers affected were then covered by some form of escalator provision, compared with one fourth of the workers outside manufacturing. However, coverage varies widely by industry. About 90 per cent or more of all workers under major settlements are covered in the primary metals, transportation equipment, trucking and communications industries, and a large majority (about three fourths or more) in the machinery and electrical equipment, tobacco, and retail distribution industries. At the other extreme, there is little provision for escalation in many industries, including textiles, apparel, lumber, construction, mining, and public utilities.

Some recent contracts have provided for automatic adjustment in pension benefits to compensate partially for a rise in the CPI. In addition to wage and private pension adjustments under wage contracts, about 29 million recipients of social security benefits, 2 million retired military and federal civil service employees, postal workers, and food stamp recipients are afforded some degree of compensation for price increases.

Until 1973, the most common wage adjustment formula called for a 1 per cent increase in hourly wages for each 0.4 point rise in the CPI (1957–59 = 100). In some more recent contracts, the formula has been switched to an increase of 1 per cent for each 0.3 point rise in the CPI based on 1967. The wages of the majority (75 per cent) of the workers covered by escalators are adjusted quarterly; 12 per cent receive semiannual adjustments; and 23 per cent, annual adjustments.

Since the rate of price increase began to accelerate, there has been a growing tendency for escalator clauses to include a maximum, or “cap.” By the end of 1973, somewhat less than 50 per cent of workers covered by escalator provisions were subject to a ceiling—the most prevalent limits being 11–14 cents an hour per year. (This represents compensation at the agreed rate for an increase of about 4 points on the CPI (1967 = 100), or an increase of about 2½ per cent on the 1974 level of the index.)

The existing CPI measures the change in the price of a fixed set of goods and services, representative of the average expenditure of the population of urban wage earners and clerical workers (accounting for some 55 per cent of total urban population and less than 45 per cent of the total population). In April 1974, the Bureau of Labor Statistics announced its intention to broaden the coverage of the index to include all urban households, and to discontinue the calculation of the existing index, for reasons of economy, in 1977. The proposal was hotly opposed by the unions. It has since been decided to begin publishing two CPIs in 1977: an improved index for urban wage earners to meet the requirements of collective bargaining; and an index for all urban households, to provide a comprehensive measure of price changes for the economy.


The Purchasing-Power-Parity Theory of Exchange Rates: A Review Articlelawrence h. officer (pages 1–60)

The purchasing-power-parity (PPP) theory involves the ratio of two countries’ price levels (absolute PPP) or price indices times a base period exchange rate (relative PPP) as the most important variable determining the exchange rate, but it allows both for other explanatory variables and for random influences. Such was the methodology of Cassel, who was the first economist to make PPP an operational theory. Other writers have suggested the use of factor costs in place of prices. The basic rationale for PPP theory is that the value of a currency is determined fundamentally by the amount of goods and services that a unit of the currency can buy in the country of issue.

Criticisms of PPP include the existence of tariffs and transport costs, the role of income in exchange rate determination, and the existence of non-current account items in the balance of payments. The direction of causality (prices to exchange rates) has also been called into question. The most damaging criticism is that the ratio of the price level of nontraded to that of traded commodities is not equal across countries, nor does it move uniformly in different countries over time. Most critics of PPP, however, do not reject the theory outright. They recognize the applicability of the theory under certain circumstances, such as moderate or large inflations. The principal empirical use of PPP is to measure the amount of disequilibrium of an exchange rate.

Several kinds of tests of the PPP theory have appeared in the empirical literature. One test demonstrates that the ratio of PPP to the exchange rate is systematically related to differences in per capita income across countries. Other tests correlate time series of PPP and the exchange rate or perform a comparative-static comparison of the two variables over time. More sophisticated tests involve the specification of a lagged relationship between PPP and the exchange rate, and the inclusion of explanatory variables in addition to PPP in exchange rate determination.

International Reserves and World-Wide Inflationh. robert heller (pages 61–87)

In this paper it is argued that there exists a systematic relationship between changes in the world-wide aggregate of international reserves and the rate of world-wide inflation. While the volume of international reserves in existence has no direct effect on price levels, the changes in national monetary aggregates that accompany changes in international reserves do have a significant lagged impact on national and international price developments.

The evidence examined shows that changes in global international reserves have a significant impact on the aggregate world money supply. There is an average lag of about one year in this relationship. It was also established that there exists a relationship between changes in the world money stock and changes in world prices. The evidence indicates a mean lag of approximately one and a half years in this relationship.

Estimates relating changes in global international reserves directly to changes in world consumer prices found a significant lagged relationship between these two crucial variables. The mean length of the lag was determined to be two and a half to four and a half years.

It is also argued that the sharp increase in international reserves that helped to trigger the world-wide inflation of the early 1970s had an effect on the structure of the recent inflation. In contrast to the post-World War II experience, the international sector represented an intensifying rather than a mitigating inflationary influence during the early 1970s.

While an increase in the U. S. dollar component of international reserves was the proximate cause of their increase in the early 1970s, there is little evidence that a more excessive monetary expansion in the United States than in other countries was responsible for this. Instead, a decrease in the demand for dollars as an asset by U. S. residents and private foreign entities alike resulted in a conversion of dollars into other currencies and a consequent expansion of the volume of foreign official dollar holdings in the form of international reserves.

Indices of Effective Exchange Ratesrudolf r. rhomberg (pages 88–112)

Seven indices of the effective exchange rate are calculated using the same set of exchange rates, the same base date, the same trade data for calculating weights, and the same sample of 15 countries. An index of the effective rate intended for use in connection with estimating or analyzing exchange rate effects on the trade balance must use as weights some estimates, however imperfect, of the relative influence of changes in the prices of various foreign currencies on the trade balance of the home country. The index that uses weights calculated from the Fund’s multilateral exchange rate model (MERM), in contrast to trade-weighted indices, is so constructed and is, therefore, preferable for analyzing effects of exchange rate movements on the trade balance. Indices that use as weights exchange rate effects on the current account balance, or some other balance of payments measure, would be useful for making a more comprehensive analysis of countries’ policies on external balance and domestic stability, but unfortunately the corresponding models have not yet been developed. In their absence, the MERM-weighted index recommends itself for these analytical purposes as well. Sometimes averages of various trade-weighted indices (such as an average of indices using bilateral imports and bilateral exports as weights, or an average of these and an index based on global export weights) are used as approximations of a general analytical measure of the effective exchange rate. Although this procedure is generally inadvisable, in some instances the discrepancies between the various indices may not be large.

An Indicator of Effective Exchange Rates for Primary Producing Countriesgérard bélanger (pages 113–36)

The need, for analytical and policy-making purposes, to represent the evolution of a given currency in terms of its total relationship to other currencies has led, in recent years, to a proliferation of methods aimed at incorporating in a single indicator of “effective” exchange rate the complex of direct and indirect influences on a currency’s value resulting from adjustments in its own and other currencies’ exchange rates. The purpose of this paper is to extend to primary producing countries the analysis developed in the Fund’s multilateral exchange rate model (MERM) for industrial countries, that is, to estimate the impact of a given set of changes in exchange rates on a country’s balance of trade after an adjustment period of two to three years. Among its applications, the model makes it possible to estimate the effective exchange rate of the currency concerned according to the following definition: the isolated change in the exchange rate of a currency that would be equivalent in its net effect on the trade balance to the whole complex of exchange rate changes that have taken place over a period of time. The proposed framework could be viewed as an attempt to identify this effect on the trade position and, hence, on the effective exchange rate of individual primary producers, which MERM currently includes as part of the “rest of the world.”

Changes in the exchange rates of all countries that produce or consume the commodities in the export trade of the country under study are analyzed to identify their effect on the price of these commodities on world markets. The effect on the country’s export receipts is identified on the basis of the response of its own supply to a change in price. The impact of changes in exchange rates on the country’s import demand is separated into two components—the induced effect of changes in export receipts on domestic expenditures (and, hence, on imports) and the effect of changes in the exchange rates on the price of the country’s imports. These components are specified as exogenous disturbances in a simple income determination system, resulting in an adjustment of the main economic aggregates, including imports. The second exogenous disturbance is quantified by using results generated by MERM.

Application of the model to two primary producers, Zaïre and Zambia, shows that the major influence on their effective exchange rates is the impact of exchange rate changes on the world price of their major export commodities, particularly copper. This result, in addition to arguments that highlight the shortcomings of the so-called export-weighted effective exchange rate index, suggests that none of the often-used “weighting” schemes (export, import, bilateral trade weights) are adequate alternatives for primary producing countries if the purpose of the analysis is to identify the effect of changes in exchange rates on the trade position of a given country.

The Advantages of Exclusive Forward Exchange Rate Supportwilliam h. l. day (pages 137–63)

The author has examined the official foreign exchange intervention mechanism of supporting only the forward exchange rate while maintaining control over the inflow or outflow of money by buying or selling reserves in a free spot market. The principal findings are as follows: (1) To achieve a desired total currency flow, the necessary and sufficient condition is to make official purchases of domestic currency in the spot market that are less than maturing official forward sales of domestic currency by the magnitude of the desired total currency flow. (2) The only constraint on the ability of a country to use the intervention mechanism to maintain zero net capital inflows is the availability of reserves to meet trade deficits and official losses. (3) There is no constraint on the ability of trade surplus countries to achieve offsetting capital outflows. (4) Trade deficit countries can secure offsetting capital inflows without disturbing interest parity for as long as the risk of official default is considered to be nil. (5) Interest arbitrage flows are inherently destabilizing when only the spot rate is supported, but they are inherently stabilizing when only the forward rate is supported. (6) Securing the required interest arbitrage flows will induce only minimal official losses for as long as interest parity holds. (7) Official losses owing to exchange rate depreciations or appreciations are likely to be less than they would have been if only the spot rate had been supported, provided that the intervention mechanism is not used to prolong the maintenance of an overvalued or undervalued exchange rate. (8) A lengthening and thickening of forward markets will encourage trade. (9) If appropriate policies in relation to the exchange rate are pursued, speculation can be expected to be stabilizing if the maturity period of the supported forward rate reflects the time period necessary for the policies to become effective. (10) If the maturity period of the supported forward rate of a trade deficit country was such that it created investment opportunities—covered against exchange risk—of similar maturity to the newly acquired liabilities of the world banking system to trade surplus countries, the strains on the world banking system would be reduced. (11) Conventional forward intervention policy under a supported spot rate has an imprecise effect on the total currency flow; the effectiveness of the policy is likely to be reduced by adverse speculation. The impact of the intervention mechanism on the total currency flow is precise; the effectiveness of the policy is likely to be enhanced by stabilizing speculation.

A Monetary Model of the Korean Economyichiro otani and yung chul park (pages 164–99)

The main purpose of this paper is threefold: (1) to remedy various aspects of the defects inherent in a single-equation approach to an explanation of a monetary phenomenon; (2) to construct a simultaneous-equation system of a quarterly monetary model of the Korean economy and to test the validity of the model against the data for the period 1962–74; and (3) to analyze influences on major macroeconomic variables, including real output, prices, and balance of payments, of alternative policy instruments, by way of simulation exercises.

The model is a nonlinear system with five behavioral equations (for inflation, imports of intermediate goods, imports of consumer goods, real output in the nonprimary sector, and currency/deposit ratio) and five identity equations. The model is monetary in that money plays a central role in determining key macroeconomic variables of the economy. It has its origin in long-run monetary models of balance of payments developed by Polak (1957), Johnson (1972), Dornbusch (1973), and others. But it differs fundamentally from these standard models in several respects. First, it is a short-run model in which prices and real output are determined endogenously. Second, the model can explain the effects of the monetary impulse on prices and real output. Third, the model contains an explicit supply function of real output based on profit-maximizing behavior in neoclassical theory and explicitly considers the importance of imported intermediate goods in the supply of real output.

Such a model is estimated by the two-stage least-squares method, using quarterly data for 1962–73. Then, the estimated model is subjected to a variety of fully dynamic simulation exercises to examine its properties with regard to stability, predictive ability, etc. These exercises indicate that: (1) the model is stable, has good predictive ability, particularly within the sample period, and is generally valid; (2) it possesses a remarkably symmetric response to the exogenous shocks given to the fiscal and monetary variables; and (3) it will, at least from a qualitative point of view, produce expected outcome of endogenous variables in response to changes in the policy variables.

Large Versus Small Price Changes and the Demand for Importsmorris goldstein and mohsin s. khan (pages 200–25)

This paper presents quarterly estimates of aggregate import demand functions for the period 1955–73 for 12 industrial countries: Belgium, Denmark, Finland, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, Norway, Sweden, the United Kingdom, and the United States. In addition, several empirical tests are undertaken, on both aggregate and disaggregate import data, to determine whether the (demand) elasticity of imports with respect to relative prices and the speed at which actual imports adjust to the desired level are both independent of the size of the relative price change.

In the estimated equations for total imports during the period 1955–73, it was observed that import demand was responsive to relative prices for 8 of the 12 countries. The sizes of these estimated price elasticities were found to be generally similar to those estimates obtained in earlier studies. Real income changes were also found to exert a significant influence on aggregate import demand; here, the quarterly estimates proved to be somewhat smaller than earlier annual or semiannual estimates.

Our results also suggest the presence of time lags in the response of aggregate imports to changes in relative prices and real income. Our estimates of these lags, however, tended to be shorter (generally between one and three quarters) than is sometimes assumed. Further, there was some tentative evidence that aggregate imports adjust more rapidly to real income changes than to relative price changes.

Finally, this study found no evidence that either the price elasticity of demand for imports or the speed of adjustment of actual to desired imports was significantly affected by the size of the relative price change. Further, this qualitative conclusion continued to hold when the tests were redone, in turn, with import data for manufactures, for a shorter time period, and with an alternative distributed-lag structure.

Indexation of Wages and Salaries in Developed Economiesanne romanis braun (pages 226–71)

Part I discusses the wage bargaining process and indexation in theory. The monetarist case for indexation supposes that if a certain rate of monetary expansion is consistently maintained, the rate of increase of prices in general will come to be correctly anticipated, regardless of whether indexation is explicitly adopted or not, with the level of employment stable at its “natural” level. Wages indexation will be helpful in securing a swifter adjustment of wage (and price) decisions to a change in monetary policy, and in avoiding an “excessive” expansion, or contraction, of employment and output and “distortions” in the structure of output, owing to mistaken price expectations. The monetarist view relies on the assumption that in the real economy powerful forces make for steady-state equilibrium, and that economic behavior quickly adjusts to whatever rate of monetary expansion is maintained. A less confident belief in the self-equilibrating properties of the real economy would suggest that, even when a consistent rate of monetary expansion was maintained, autonomous real factors resulting in stronger demand or cost pressures in particular sectors could provoke a persisting acceleration (or deceleration) of the general rate of inflation coupled with the reduction (or raising) of the level of employment. It would be difficult for confident price expectations for particular sectors and general agreement on the probable rate of inflation to become established in this kind of world. In this context, one would need to consider whether indexation of wages would not render the movement of prices generally more sensitive to price and cost pressures in some sectors, increasing the risk that autonomous changes in the real economy (or emanating from abroad) would set off an acceleration of inflation and increased unemployment under the existing monetary policy.

The argument that indexation may lessen cost pressure is most persuasive where labor is strongly organized and full employment commands a high political priority, and where any well-publicized wage settlement in excess of the current rate of price increase plus average growth in productivity is liable to unsettle price expectations, even when it is believed that the authorities intend to maintain the existing degree of monetary restraint. Protection via indexation against the eventual rate of price increase may offer a means of preventing one or two large settlements from generating a round of similar increases. While it is usually correct to regard indexation as tending to reduce the flexibility of relative wages, where great rigidity already exists, indexation may sometimes provide a means of securing increased flexibility.

Part II describes the characteristics of indexation in practice, and the case for introducing or extending indexation in particular countries. It is no accident that the recent debate on indexation of wages has been strongest in countries where indexation was relatively unimportant, notably the United States, the Federal Republic of Germany, the United Kingdom, and Australia. Policy options with respect to indexation of pay are usually limited. Under rapid inflation, the authorities are unlikely to be able to resist the spread of some form of cost of living clauses. Widespread indexation of wages, salaries, and pensions now applies in most Western European countries, generally dating from the period of rapid inflation following World War II.

Discussion of the economic case for indexation has tended to divert attention from the fact that wages indexation is primarily a means of mitigating the political consequences of rapid inflation, by avoiding the buildup of intergroup conflicts occasioned by changes in the distribution of income.



In statistical matter (except in the résumés and resúmenes) throughout this issue,

  • Dots (…) indicate that data are not available;

  • A dash (—) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist;

  • A singe dot (.) indicates decimals;

  • A comma (,) separates thousands and millions;

  • “Billion” means a thousand million;

  • A short dash (−) is used between years or months (e.g., 1971–74 or January-October) to indicate a total of the years or months inclusive of the beginning and ending years or months;

  • A stroke (/) is used between years (e.g., 1973/74) to indicate a fiscal year or a crop year;

Components of tables may not add to totals shown because of rounding.

International Monetary Fund, Washington, D.C. 20431 U.S.A.

Telephone number: 202 393 6362

Cable address: Interfund



Instruments of Monetary Policy in the United States: The Role of the Federal Reserve System. Ralph A. Young. 1973. US$1.25.

Membership and Nonmembership in the International Monetary Fund: A Study in International Law and Organization. Joseph Gold. 1974. US$10.00.

Voting and Decisions in the International Monetary Fund: An Essay on the Law and Practice of the Fund. Joseph Gold. 1972. US$6.50.

The Stand-By Arrangements of the International Monetary Fund: A Commentary on Their Formal, Legal, and Financial Aspects. Joseph Gold. 1970. US$4.00.

International Monetary Reform: Documents of the Committee of Twenty. English, French, and Spanish editions. Available on request.

International Reserves: Needs and Availability. Proceedings, IMF Seminar. 1970. US$6.00.

Surveys of African Economies

article image

Also published in French (Vol. 6, French in preparation). US$5.00 a volume (US$2.50 a volume to university libraries, faculty members, and students).

The International Monetary Fund, 1945–1965: Twenty Years of International Monetary Cooperation. (3 vols.) J. Keith Horsefield and others. 1969. US$12.50 a set (US$5.00 a volume).


Balance of Payments Yearbook. Monthly, loose-leaf, US$7.50 a volume; binder available for US$3.50. Annual, clothbound, US$6.00.

Direction of Trade. Monthly, with annual supplement. US$10.00 a year.

International Financial Statistics. Monthly, in English, French, and Spanish. US$20.00 a year. Specify language edition desired.

Staff Papers. Three times a year. US$6.00 a year.

University libraries, faculty members, and students may obtain the four subscription publications listed above at the reduced rates of US$12.00 for all four publications or US$5.00 for International Financial Statistics and US$3.00 each for the others.

IMF Survey. Twice monthly, with single issue in December, in English, French, and Spanish. Distributed by fastest mail. US$5.00 in Canada, Mexico, and the United States; US$10.00 in Central and South America; US$12.00 in all other areas. Specify language edition desired.

Finance and Development (jointly with International Bank for Reconstruction and Development). Quarterly in English, French, and Spanish. Available on application; specify language edition desired.


Series from International Financial Statistics, from Direction of Trade, and from the Balance of Payments Yearbook are available on tape for computer users, at a price of US$1,000.00 a year for each title, including a subscription to the book version (US$300.00 to universities).

Address inquiries to

The Secretary, International Monetary Fund, Washington, D.C. 20431



  • Australian Conciliation and Arbitration Commission, “National Wage Case, April 1975, Reasons for Decision” (unpublished, Melbourne, April 30, 1975).

    • Search Google Scholar
    • Export Citation
  • Australian Department of Labor, Wage Indexation for Australia? A Discussion Paper (Canberra, 1974).

  • [Australian] Institute of Applied Economic and Social Research, “The Economic Situation,” Australian Economic Review, No. 31 (3rd Quarter 1975), pp. 413.

    • Search Google Scholar
    • Export Citation
  • Banque Nationale de Belgique, “Recherche sur les causes des disparites des taux d’inflation en Belgique, en République fédérate d’Allemagne et aux Pays-Bas,” Bulletin de la Banque Nationale de Belgique (May 1975), pp. 362.

    • Search Google Scholar
    • Export Citation
  • Bernstein, Edward M. (1958), “Wage-Price Links in a Prolonged Inflation,” Staff Papers, Vol. 6 (November 1958), pp. 32368.

  • Bernstein, Edward M. (1974), “Indexing Money Payments in a Large and Prolonged Inflation,” in Essays on Inflation and Indexation, American Enterprise Institute for Public Policy Research—cited elsewhere in this Bibliography as Essays on Inflation and Indexation (Washington, 1974), pp. 7186.

    • Search Google Scholar
    • Export Citation
  • Blackaby, Frank, “Incomes Policies and Inflation,” National Institute Economic Review (November 1971), pp. 3453.

  • Braun, Anne Romanis, “Compulsory Arbitration as a Form of Incomes Policy: The Australian Case,” Staff Papers, Vol. 21 (March 1974), pp. 170216.

    • Search Google Scholar
    • Export Citation
  • Commission of the European Communities, “Machinery for Adjusting Wages and Salaries to the Cost of Living,” Working Paper of the Services of the Commission (unpublished, Brussels, November 14, 1974).

    • Search Google Scholar
    • Export Citation
  • Cunningham, Francis S., “Use of Price Indexes in Escalator Contracts,” Monthly Labor Review, Vol. 86 (August 1963), pp. 94852.

  • Dabscheck, Braham, “The 1975 National Wage Case: Now We Have an Incomes Policy,” Journal of Industrial Relations (Sydney), Vol. 17 (September 1975), pp. 298309.

    • Search Google Scholar
    • Export Citation
  • David, Lily Mary, “Cost-of-Living Escalation in Collective Bargaining,” Price and Wage Control: An Evaluation of Current Policies, Hearings Before the Joint Economic Committee, November 13–15, 1972 (92nd Congress, 2nd Session, Washington, 1973), Part 2, pp. 33241.

    • Search Google Scholar
    • Export Citation
  • Dernberg, Thomas F., “Inflation Indexing of the Personal Income Tax: The Case from the Perspective of Stabilization Policy” (unpublished, International Monetary Fund, February 14, 1975).

    • Search Google Scholar
    • Export Citation
  • Douty, H. M., Cost-of-Living Escalator Clauses and Inflation, Staff Report, Executive Office of the President, Council on Wage and Price Stability (Washington, August 1975).

    • Search Google Scholar
    • Export Citation
  • Fane, C. G., “Index-Linking and Inflation,” National Institute Economic Review (November 1974), pp. 4045.

  • Fédération des Entreprises de Belgique, Commission des Questions Sociales, “Liaison des salaires et des appointements à l’indice des prix a la consommation,” Document C.Q.S. 75/31 (unpublished, Brussels, February 27, 1975).

    • Search Google Scholar
    • Export Citation
  • First National City Bank of New York, “How Indexing Can Ease Inflation’s Pain,” Monthly Economic Letter (October 1974), pp. 79.

    • Search Google Scholar
    • Export Citation
  • Fishlow, Albert, “Indexing Brazilian Style: Inflation Without Tears?” Brookings Papers on Economic Activity: 1 (1974), pp. 26182.

    • Search Google Scholar
    • Export Citation
  • Friedman, Milton, “Monetary Correction,” in Essays on Inflation and Indexation (Washington, 1974), pp. 2561.

  • Garbarino, Joseph W. (1959), “The Economic Significance of Automatic Wage Adjustments,” in New Dimensions in Collective Bargaining, ed. by Harold W. Davey, Howard S. Kaltenborn, and Stanley H. Ruttenberg (New York, 1959), pp. 15475.

    • Search Google Scholar
    • Export Citation
  • Garbarino, Joseph W. (1960), Wage Escalation and Wage Inflation, The Brookings Institution (Washington, 1960).

  • Geluck, Jan, “Wage Indexation,” in Wage Determination: Papers Presented at an International Conference, Paris, 3–6 July 1973, Organization for Economic Cooperation and Development (Paris, 1974), pp. 177214.

    • Search Google Scholar
    • Export Citation
  • Gern, Jean-Pierre, “L’Indexation des salaires: ses répercussions économiques” (Neuchâtel, 1961).

  • Giersch, Herbert, “Index Clauses and the Fight Against Inflation,” in Essays on Inflation and Indexation (Washington, 1974), pp. 123.

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, “Wage Indexation, Inflation, and the Labor Market,” Staff Papers, Vol. 22 (November 1975), pp. 680713.

  • Goodman, J. F. B., and G. M. Thomson, “‘Cost of Living’ Indexation Agreements in Post-War British Collective Bargaining,” British Journal of Industrial Relations, Vol. 11 (July 1973), pp. 181210.

    • Search Google Scholar
    • Export Citation
  • Guenther, Jack D., “‘Indexing’ Versus Discretionary Action—Brazil’s Fight Against Inflation,” Finance and Development, Vol. 12 (September 1975), pp. 2429.

    • Search Google Scholar
    • Export Citation
  • Guglielmi, J.-L., “‘Echelle mobile’ ou ‘indexation’ des salaires en France?” Revue Economique (March 1955), pp. 21835.

  • Guitton, Henri, “Le problème économique de l’indexation,” Revue Economique (March 1955), pp. 187217.

  • Heller, Walter W., “Has the Time Come for Indexing?” Wall Street Journal, June 20, 1974.

  • Hughes, James J., “Cost of Living, Threshold Bargaining and Incomes Policy,” Industrial Relations Journal, Vol. 4 (Winter 1973/74), pp. 2333.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, Fiscal Affairs Department, “Adjustment of Taxation for Inflation” (unpublished, International Monetary Fund, May 23, 1975).

    • Search Google Scholar
    • Export Citation
  • Kafka, Alexandre, “Indexing for Inflation in Brazil,” in Essays on Inflation and Indexation (Washington, 1974), pp. 8798.

  • Keynes, John Maynard, A Tract on Monetary Reform (London, 1923).

  • Kredietbank, “Index-Linking of Wages,” Weekly Bulletin (April 14, 1972), pp. 15762.

  • Kriz, Miroslav A., “Inflation: the Case for Indexation,” Euromoney (June 1974), pp. 1923.

  • Laffer, Kingsley, “Cost-of-Living Adjustments in Australian Wage Determination,” Industrial and Labor Relations Review, Vol. 7 (January 1954), pp. 27887.

    • Search Google Scholar
    • Export Citation
  • Liefmann-Keil, Elizabeth, “Index-Based Adjustments for Social Security Benefits,” International Labour Review, Vol. 79 (May 1959), pp. 487510.

    • Search Google Scholar
    • Export Citation
  • Lowenstern, Henry, “Adjusting Wages to Living Costs: A Historical Note,” Monthly Labor Review, Vol. 97 (July 1974), pp. 2126.

  • Marshall, Alfred, “Remedies for Fluctuations of General Prices,” Contemporary Review (March 1887).

  • Morley, Samuel A., “Indexing and the Fight Against Inflation” (unpublished, University of Wisconsin, 1974).

  • Organization for Economic Cooperation and Development, Economic Surveys: Norway (Paris, January 1975).

  • Page, Mrs. S. A. B., and Sandra Trollope, “An International Survey of Indexing and Its Effects,” National Institute Economic Review (November 1974), pp. 4660.

    • Search Google Scholar
    • Export Citation
  • Peacock, A. T., and W. J. L. Ryan, “Wage Claims and the Pace of Inflation (1948-51),” Economic Journal, Vol. 63 (June 1953), pp. 38592.

    • Search Google Scholar
    • Export Citation
  • Perna, Nicholas, “Cost of Living Escalator Clauses in Collective Bargaining Contracts,” Federal Reserve Bank of New York, Research Paper No. 7334 (December 1973).

    • Search Google Scholar
    • Export Citation
  • Pontarollo, Enzo, John Hughes, and Roy Moore, “The Threshold in Europe: Inflation, Indexation and the Trade Union Response in E.E.C. Countries,” Trade Union Research Unit, Ruskin College, Oxford (unpublished, July 22, 1974).

    • Search Google Scholar
    • Export Citation
  • Ruben, George, “Major Collective Bargaining Developments—A Quarter-Century Review,” Current Wage Developments, Vol. 26 (February 1974), pp. 4254.

    • Search Google Scholar
    • Export Citation
  • Schultze, Charles L., Recent Inflation in the United States, Study Paper No. 1, prepared for consideration by the Joint Economic Committee (86th Congress, 1st Session, Washington, September 1959).

    • Search Google Scholar
    • Export Citation
  • Sen Gupta, A. K., “Appendix II: Survey of Wage-Price Links in Various Countries,” Staff Papers, Vol. 6 (November 1958), pp. 34368.

    • Search Google Scholar
    • Export Citation
  • Shiskin, Julius, “Updating the Consumer Price Index—An Overview,” Monthly Labor Review, Vol. 97 (July 1974), pp. 320.

  • Slichter, Sumner H., “Do the Wage-Fixing Arrangements in the American Labor Market Have an Inflationary Bias?” American Economic Review, Papers and Proceeding of the Sixty-sixth Annual Meeting of the American Economic Association, Vol. 64 (May 1954), pp. 32246.

    • Search Google Scholar
    • Export Citation
  • Soffer, Benson (1954), “Cost-of-Living Wage Policy,” Industrial and Labor Relations Review, Vol. 7 (January 1954), pp. 19299.

  • Soffer, Benson (1959 a), “The Effects of Recent Long-Term Wage Agreements on General Wage Level Movements: 1950–1956,” Quarterly Journal of Economics, Vol. 73 (February 1959), pp. 3660.

    • Search Google Scholar
    • Export Citation
  • Soffer, Benson (1959 b), “On Union Rivalries and the Minimum Differentiation of Wage Patterns,” Review of Economics and Statistics, Vol. 41 (February 1959), pp. 5360.

    • Search Google Scholar
    • Export Citation
  • Staller, Jerome M., and Loren M. Solnick, “Effect of Escalators on Wages in Major Contracts Expiring in 1974,” Monthly Labor Review, Vol. 97 (July 1974), pp. 2732.

    • Search Google Scholar
    • Export Citation
  • Trades Union Congress, Cost of Living Threshold Agreements (London, 1972).

  • Ulman, Lloyd, and Robert J. Flanagan, Wage Restraint: A Study of Incomes Policies in Western Europe (University of California Press, 1971).

    • Search Google Scholar
    • Export Citation
  • White, William H., “Inflation and the Linking of Financial Contracts to the Price Index” (unpublished, International Monetary Fund, 1974).

    • Search Google Scholar
    • Export Citation
  • Willey, Richard J., “Trade Unions and Political Parties in the Federal Republic of Germany,” Industrial and Labor Relations Review, Vol. 28 (October 1974), pp. 3859.

    • Search Google Scholar
    • Export Citation
  • Woodward, A. C., and R. Senkiw, “Indexation in Finland and Its Impact on Inflation” (unpublished, September 10, 1974).

  • Vega, Ramón, “Wage Escalator Clauses and Inflation: The Experience of Selected European Countries” (unpublished, U. S. Treasury Department, October 18, 1973).

    • Search Google Scholar
    • Export Citation
  • Zoeteweij, Bert, “The Adjustment of Wages to Changes in the Cost of Living,” International Labour Review, Vol. 66 (August 1952), pp. 89112.

    • Search Google Scholar
    • Export Citation

Mrs. Braun, Advisor in the Research Department, is a graduate of Cambridge University. She formerly worked as an economist at the Oxford University Institute of Economics and Statistics, in the U.K. Ministry of Production and Board of Trade, the Organization for European Economic Cooperation, and the United Nations. She has published a number of articles in the fields of incomes policy and international trade.

In addition to colleagues in the Fund, the author is indebted to Sir E. H. Phelps Brown and Manfred Willms for helpful comments. Views expressed are those of the author.


Charles L. Schultze is the leading exponent of the concept of inflation originating neither in an overall excess of money demand nor in autonomous upward wage push but in shifting excess demands on particular sectors in an economy characterized by price and wage rigidities. See Schultze (1959).


See Friedman (1974), p. 31; and Giersch (1974), especially p. 10. Friedman’s and Giersch’s views are discussed in Section II.


See Joan Robinson, “What Has Become of the Keynesian Revolution?” Challenge, Vol. 16 (January/February 1974), pp. 6–11.


The monetarist approach begs the question whether it is legitimate to treat the economy as tending to steady-state equilibrium, under a constant rate of monetary expansion, if some paths to the steady state are politically impassable.


Note that rapid, widespread growth in productivity would have a similar effect to a moderately rising price level in permitting flexibility of relative price movements to be achieved via postponement of wage increases.


A possible exception to this situation may occur in an open economy, which maintains a fixed exchange rate in relation to its principal trading partners, as foreign competition may set a ceiling on the rate of price and wage increase in the more highly organized sectors of rapid growth in productivity.


Friedman’s advocacy of wages indexation in the United States, and his commendation of the Brazilian system, applied only to wage settlements in the unionized sector.


“Because of its uneven impacts, then, our existing system throws sand into the gears of inflation. Indexing would oil the gears and speed the process of inflation,” Heller (1974).

“The ‘escalator’ features of the GM-type formulas make the general level of wages highly sensitive to even small fluctuations in the Consumers’ Price Index,” thereby limiting the effectiveness of mild monetary or fiscal policies to check inflation. “Reducing the lags, frictions, delays and uncertainties which slow the average rise in wages probably lubricates the inflationary process,” Soffer (1959 a), pp. 54–55.


The United Kingdom is the most obvious case where the conditions described apply, but they are now more or less fulfilled in the Federal Republic of Germany and Australia (as well as in Belgium, Denmark, and the Netherlands, where indexation is already widespread).


As a study by the Australian Department of Labor (1974) observed (p. 12), indexation “avoids a situation where workers base wage claims on the highest expected rate of inflation. If a union is successful in achieving such a claim, this tends to become the norm for which other unions strive and the expected rate of inflation becomes a self-fulfilling prophecy.”


Or perhaps the threat of runaway inflation under flexible exchange rates.


Where home mortgage costs are included in the cost of living index with a substantial weight, as in the United States, the effect will be to make the level of wages unduly sensitive to the effect of restrictive monetary policy upon interest rates, and hence to weaken the effectiveness of such a policy in curbing inflation. See William H. White, “House Depreciation and Interest Costs and the Reliability of Recent Measured Increases in the Cost of Living” (unpublished, International Monetary Fund, April 28, 1967).


It is conceivable that the retention of the disproportionate weighting given to food reflects the fact that the unions are better placed to secure compensation for increases in the price of industrial products in the course of wage bargaining.


The same factor enhanced the disinflationary impact, on wage movements, of increasing supplies of agricultural products and falling prices on world markets during much of the 1950s and the early 1960s.


For a comment on Friedman’s earlier views on indexation, see Soffer (1959a), p. 54, footnote 3.


Writing in 1960, Garbarino explained why the length of the labor contract had become a major bargaining issue in many wage negotiations in the United States, with management usually pressing for multiyear agreements and indexation clauses representing a technique that made possible the negotiation of long-term contracts.

“… each contract negotiation has the potential of upsetting the external balance of power in the relationships with union leaders, political leaders, government agencies, customers and competitors. This consideration is reinforced by the intraorganizational turmoil that may result from the negotiations. For a period of months before the expiration of the current contract the union and the company may be engaged in a public debate over the anticipated issues in bargaining. Actual negotiations typically occupy the two months before the contract deadline and may run beyond that date in an atmosphere of increasing tension, sporadic stoppages, and decreasing worker efficiency. Even if a strike is avoided, the changes in the new contract have to be integrated into the operating practices of several levels of supervision and adjudication of differences of interpretation through the grievance procedure may have to be undertaken,” Garbarino (1960), pp. 1–2.


Ibid., p. 34.


“Different activities have different time-speeds of adjustment. Some prices, wages and production schedules are fixed a long time in advance; others can be adjusted promptly. As a result, a slowdown of total spending produces substantial shifts in relative prices, which will sooner or later have to be corrected; the correction in turn will cause economic disturbances.” Ibid., pp. 31–32.


Suppose, for instance, that sectors subject to smaller increases in demand during the inflationary period are the last to conclude wage settlements allowing for the inflationary average rate of price increase, and that sectors that experienced larger increases in demand associated with the inflation are the first to bargain settlements in the expectation of deflationary conditions. The resulting changes in relative wages and prices would then seem to be conducive to securing the relative wages and prices appropriate to the structure of demand in the absence of inflation.


Ibid., p. 59.


See Giersch (1974), p. 10.


In 1970, collective bargaining settlements raised wages by 13.6 per cent, excluding improvements in fringe benefits. “In January 1973, two thirds of those steel workers voting in North Rhine-Westphalia turned down a union-negotiated package of 8.5 percent … By July, the metal workers’ union estimated that over one hundred concerns had been affected by wildcat strikes, and most managements were giving in to the strikers with ‘inflation bonuses’ of one kind or another,” Willey (1974), p. 58.


Ibid., p. 4.


See Blackaby (1971), p. 49.


The Discussion Paper prepared by the Department of Labor in 1974 concluded that “… it seems likely that the over-award element of earnings drift has increased partly because of the inadequate compensation for the impact of inflation on award wages in the 1950s, and the less frequent cost of living adjustment of award wages in the sixties. That is, not only are over-award payments a reflection of the demand conditions prevailing in the labour market, but also of attempts by trade unions to preserve the real wages of members in times of inflation. … It can be argued therefore that the absence of automatic cost of living adjustment has been one factor in the increase in over-award payments in the post-war period, especially in times of inflation. While there was positive earnings drift even at the height of the Korean War inflation, automatic adjustments operated to reduce the dimensions of earnings drift.” If wage indexation were to be introduced, “that element of over-award payments catching up with, and anticipating, inflation could be reduced,” Australian Department of Labor (1974), pp. 25–26.

For a description of these problems and an explanation of the wages terminology, see Braun (1974), especially pp. 172–76 and 186–92.


The Australian Council of Trade Unions (ACTU) had pressed for “indexation of taxation” to provide protection against the erosion of real aftertax income in a resolution of September 1974. An official Committee of Inquiry studying the impact of progressive taxation under inflation produced its report in May 1975. Witnesses in the 1975 Wage Case stressed the part played by growing recognition of the impact of progressive taxes under inflation in increasing the scale of wage claims in the last few years. The Commission stated that it had been impressed by this point, and considered that the viability of its proposals for wage fixing would depend on the Government’s sensitivity to this problem in its tax decisions. (See footnote 41.)


Ibid., pp. 33–36.


Average weekly earnings rose by 7.2 per cent in real terms in 1974.


Ibid., p. 14.


Ibid., p. 16.


Ibid., p. 20.


Ibid., p. 22.


The President of the Commission also recommended indexed increases for metal trades workers based on actual pay rates, including over-award payments, enabling existing relativities to be maintained. Although not a binding ruling, the decision is likely to be widely followed in other industries.


Ibid., p. 9.


In May 1975, the Committee of Inquiry on Inflation and Taxation recommended full indexation of personal income tax in the 1975–76 budget, changes in the system of valuing stocks and calculating depreciation allowances, and the exclusion of indirect taxes in calculating the CPI. These proposals were not, however, accepted by the Parliamentary Taxation Review Committee. The new income taxes introduced in 1975–76, although generally lower, are no less progressive than before. Consequently, the Institute has pointed out that if prices rise by 20 per cent, a man earning $8,000 with three dependents would experience a decline of 5½ per cent in aftertax real income if his pay was adjusted only for price increases. This would generate pressure for wage increases outside the guidelines. On the other hand, it does not necessarily imply that employees would not accept tax rate increases for specific purposes.


Payments under threshold agreements could be made outside the limits for wage and salary increases of £2.25 a week (or of 7 per cent of the total wage bill of the enterprise). The agreements allowed a pay increase of up to 40 pence a week to be given when the retail price index rose 7 per cent above the October 1973 level, and a further increase of 40 pence for each subsequent 1 per cent rise in the index. Threshold payments were treated as special supplements, and were not included in base rates for calculating overtime, etc. The cost of living in April 1974 was 9.8 per cent higher than the October level, resulting in payment of three threshold allowances, and setting off a round of claims for threshold agreements. Despite the attention that they attracted, by June 1974 threshold agreements applied to only about 20 per cent of all wage and salary earners, and civil service pay was not covered. There was, however, a growing tendency for threshold agreements to spread to other industries.


“Appraisal,” National Institute Economic Review (February 1974), p. 5.


With the wisdom of hindsight it seems that, given the general expectation that the sharp rise in import prices would not continue, the Government might have been able to secure agreement to hedge its bet by providing that the threshold for adjustment would be raised if a very steep rise in import prices was to occur.


“Summary and Appraisal,” National Institute Economic Review (August 1973), p. 5.


In November 1974, when the average increase in wage rates owing to threshold payments, in settlements coming up for renewal, was 12.8 per cent, the range was from more than 20 per cent for Woolworth shopgirls to less than 10 per cent for dockworkers. See Frank Wilkinson, “The Outlook for Wages in the Short Run,” Economic Policy Review, Cambridge University, Department of Applied Economics, No. 1 (February 1975), pp. 26–27.


Douty (1975), pp. 45–46.


Ibid., p. 53.


Finland is the subject of an unpublished study by Woodward and Senkiw (1974).


Agricultural incomes are also indexed on the CPI. Increases in old-age pensions are decided by the Storting (Parliament), on the principle that they should go up in line with the general level of incomes. Social security contributions and benefits, including sick pay, are increased automatically in line with the general level of incomes.


Of the increase of 26 per cent in consumer prices between 1969 and 1972, at least 7 per cent was due to increases in indirect taxes, and another 4–5 per cent to successive increases in employers’ social security contributions.


OECD (1975), pp. 15–16.


In some small sectors, indexation provides for wages to increase in line with the index of average earnings published by the employers’ confederation.


The wage regulating price index differs from the regular consumer price index as indirect taxes are excluded, and in the weighting of items covered. The special index is published quarterly. The January and July figures are especially important, as most agreements call for adjustments twice yearly, in March and September, on the basis of the previous quarter’s level of the index. Unlike the consumer price indices in many other countries, the Danish index does not exclude or underrepresent the cost of housing.


The wage regulating index, which was 147 in January 1975 on the base January 1971 = 100, was shifted to a January 1975 base, and the adjustment step was raised from DKr 0.40 to DKr 0.60 (roughly 2 per cent of average hourly earnings).


“There is no technical problem about how to end inflation. The real obstacles are political, not technical,” Friedman (1974), p. 25, opening sentences.


Friedman (1974), pp. 44–45.


Geluck (1974), pp. 189–90.


This Appendix draws on information prepared by the European and Western Hemisphere Departments, and in country surveys published by the OECD, in addition to the sources listed in the Bibliography.


Until the late 1960s, the scale of the adjustment also depended on the region of the country, but this differentiation was abolished by agreement with the unions in 1968/69.


For further information, see Douty (1975).


The author is grateful to Mr. Karl Casserini, of the International Metalworkers’ Federation, for providing unpublished material, including the study by Pontarollo, Hughes, and Moore (1974).

IMF Staff papers: Volume 23 No. 1
Author: International Monetary Fund. Research Dept.