《宏观经济学 Macroeconomics》课外读物：Money-Wage Dynamics and Labor-Market Equilibrium.pdf

Money-Wage Dynamics and Labor-Market Equilibrum* Edmund S. Phelps University of Pennsylvania If the economy were always in macroeconomic equilibrium then perhaps the full-employment money-and-growth models of recent vintage would suffice to explain the time paths of the money wage and the price level. But since any actual economy is almost continuously out of equilibrium we need also to study wage and price dynamics under arbitrary conditions. The numerous Phillips-curve studies of the past ten years have done this with a vengeance in offering countless independent variables in numerous combinations to explain wage movements. But it is difficult to choose among these econometric models, and rarely is there a clear rationale for the model used. This paper presents a modest start toward a unified and empirically applicable theory of money-wage dynamics. At the same time it tries to capture the role of expectations and thus to work into the theory the notion of labor-market equilibrium. I. Evolution of the Phillips Curve and its Opposition Keynes' General Theory (1936) and virtually all formal macroeconomic models of the postwar era postulated a minimum unemployment level-a full-employment level of unemployment-which could be maintained with either stable prices or rising prices. In this happy state, additional aggregate demand would produce rising prices and wages but no reduction of unemployment. The full-employment quantity of unemployment was identified as "frictional " and "voluntary "; and frictional unemployment was (mistakenly) assumed to be unresponsive to demand.l Hence there was no need to choose between low unemployment and price stability. * This study was supported by a grant from the National Science Foundation. A monetary economy can choose among different levels of frictional unemployment that correspond to different levels of aggregate demand and job vacancies. In fact, therefore, there is no unique full-employment quantity of frictional unemployment

MONEY-WAGE DYNAMICS AND LABOR-MARKET EQUILIBRIUM This doctrine depended on Keynes' notions of money-wage behavior At more than minimum unemployment, a rise(fall) of demand and em- ployment would produce a once-for-all rise(fall) of the money wage prices constant; a rise(fall) of the price level would cause a rise(fall)of th money wage in smaller proportion. Hence, in a stationary economy at least, his theory did not predict the possibility of a secular rise of money wage rates at normal unemployment rates-let alone wage rises exceeding productivity growth-only the one-time"semi-inflation"(Keynes, 1936, p. 301)of prices and wages during the transition to minimum unem- ployment. This doctrine was quickly disputed by robinson(1937, pp 30-31) wrote of a conflict between moderately high employment and stability. Dunlop(1938)suggested that the rate of change of the money wage depends more on the level of unemployment than upon the rate of change of unemployment, as Keynes had it. After the war, Singer(1947) Bronfenbrenner(1948), Haberler (1948), Brown( 1955), Lerner(1958), and many others wrote that at low albeit above-minimum unemployment levels there occurs a process of“ cost inflation,”"“wage- push infation,” ¨ income inflation,”“ creeping inflation,”“ sellers' inflation,”“ dilemma inflation,” or the‘ new infation”- a phenomenon which was attributed to the discretionary power of unions or oligopolies or both to raise wages or prices or both without"excess demand I believe this customary attribution of cost infation to the existence of such large economic units to be unnecessary and insufficient. Like the theory of unemployment, the theory of cost inflation requires a nor Walrasian model in which there is no auctioneer continuously clearin commodity and labor markets. Beyond that, it is not clear to me what monopoly power contributes. An increase of monopoly power-due, say, to increased concentration -will raise prices relative to wages at any given unemployment rate and productivity level; but employment rate, the real wage has fallen (relative to productivity) continuation of inflation will depend on other sources will stop and any enough to accommodate the higher markup this process ush theorists like Weintraub (1959) to treat innatic almost spontaneous, virtually independent of the unemployment rate over any rele vant range, and hence not induced by aggregate demand I once tested the hypothesis that the 1955-57 inflation was more of this character than were the two earlier post ar infla making the push"would be uneven in its sectoral incidence, so that the coefficient of correlation between sector price changes and sector output changes would (if the hypothesis were true) be algebraically smaller in the 1955-57 period than it was earlier(1961). It was stical significance of the decline was impossible to determine. Incidentally, Selden,s correlation test ( 1959) wrongly attributes significance to the positivity of the coef n 1955-57 instead of to the magnitude of tl 3 The answer of Ackley (1966)and Lerner(1967) that correspondin mployment rate and productivity level there is a natural real wage that is irreducible

MONEY-WAGE DYNAMICS AND LABOR-MARKET EQUILIBRIUM 679 This doctrine depended on Keynes' notions of money-wage behavior. At more than minimum unemployment, a rise (fall) of demand and employment would produce a once-for-all rise (fall) of the money wage, prices constant; a rise (fall) of the price level would cause a rise (fall) of the money wage in smaller proportion. Hence, in a stationary economy at least, his theory did not predict the possibility of a secular rise of moneywage rates at normal unemployment rates-let alone wage rises exceeding productivity growth-only the one-time "semi-inflation" (Keynes, 1936, p. 301) of prices and wages during the transition to minimum unemployment. This doctrine was quickly disputed by Robinson (1937, pp. 30-31), who wrote of a conflict between moderately high employment and price stability. Dunlop (1938) suggested that the rate of change of the money wage depends more on the level of unemployment than upon the rate of change of unemployment, as Keynes had it. After the war, Singer (1947), Bronfenbrenner (1948), Haberler (1948), Brown (1955), Lerner (1958), and many others wrote that at low albeit above-minimum unemployment levels there occurs a process of "cost inflation," "wage-push inflation," "income inflation," "creeping inflation," "sellers' inflation," "dilemma inflation," or the "new inflationu-a phenomenon which was attributed to the discretionary power of unions or oligopolies or both to raise wages or prices or both without "excess demand."2 I believe this customary attribution of cost inflation to the existence of such large economic units to be unnecessary and insufficient. Like the theory of unemployment, the theory of cost inflation requires a nonWalrasian model in which there is no auctioneer continuously clearing commodity and labor markets. Beyond that, it is not clear to me what monopoly power contributes. An increase of monopoly power-due, say, to increased concentration-will raise prices relative to wages at any given unemployment rate and productivity level; but once, at the prevailing unemployment rate, the real wage has fallen (relative to productivity) enough to accommodate the higher markup, this process will stop and any continuation of inflation will depend on other ~ources.~ Some wage-push theorists like Weintraub (1959) appear to treat inflation as almost spontaneous, virtually independent of the unemployment rate over any relevant range, and hence not induced by aggregate demand. I once tested the hypothesis that the 1955-57 inflation was more of this character than were the two earlier postwar inflations, making the assumption that autonomous "wage push" or "profit push" would be uneven in its sectoral incidence, so that the coefficient of correlation between sector price changes and sector output changes would (if the hypothesis were true) be algebraically smaller in the 1955-57 period than it was earlier (1961). It was algebraically smaller, but the statistical significance of the decline was impossible to determine. Incidentally, Selden's correlation test (1959) wrongly attributes significance to the positivity of the coefficient in 1955-57 instead of to the magnitude of the decline. The answer of Ackley (1966) and Lerner (1967) that corresponding to every unemployment rate and productivity level there is a natural real wage that is irreducible

JOURNAL OF POLITICAL ECONOMY Similarly, I doubt that the existence of labor unions is remotely sufficient to explain the cost inflation phenomenon. whether the unions significantly exacerbate the problem-whether they increase that unemployment rate which is consistent with price stability-is, however, a difficult question The affirmative answer frequently starts from the theory, set forth by Dunlop(1950), that a union, to maximize its utility, seeks to "trade off the real wage rate against the unemployment of its members, raising the former (relative to productivity) until the gain from a further real wage increase is offset by the utility loss from the increase in unemployment expected to result from it. At an unemployment level below the unions the unions then push up wage rates faster than productivit But firms pass these higher costs on to consumers, so the real wage gains are frustrated, and as long as the government maintains the low unem- ployment level the rounds of inflation will continue I have trouble applying such a model to the American economy. Almost three-quarters of the civilian labor force do not belong to unions. This fact ts doubt on the quantitati ive importance of the model. And perhaps the fact goes much deeper. If the union members whom the unions make unemployed have no good prospect of future union employment, they will be inclined to seek employment elsewhere. If, at the other extreme, the union unemployment is shared in the form of a short workweek, this un- employment-while real enough to the extent that members do not moonlight"-does not add to the official unemployment rate as it is measured. Certainly the unions participate in the cost inflation process, and they may even increase a little the volume of unemployment consistent with price stability, But I should think that a union must offer its member ship a frequency of employment opportunities that is roughly comparable to that elsewhere in order to thrive and that appreciably reduced employ ment opportunities require a greater wage differential between union and other employment than is commonly observed. 4 Phillip: sful fitting of what to a scatter diagram of historical British data deprived the discussions of some of their institutional color, but epitomized the new concept of cost inflation---if by that term we mean(as i think most of the aforementioned writers intended) that kind of inflation which can be stopped only by a reduc- tion of the employment rate through lower aggregate demand and which structural changes, so that money wages will keep pace with prices until to increase, seems to me to be terribly implausible. In any IIs paper monopoly"argument 4 It is certainly likely, however, that an increase of union power, even if localized

680 JOURNAL OF POLITICAL ECONOMY Similarly, I doubt that the existence of labor unions is remotely sufficient to explain the cost inflation phenomenon. Whether the unions significantly exacerbate the problem-whether they increase that unemployment rate which is consistent with price stability-is, however, a difficult question. The affirmative answer frequently starts from the theory, set forth by Dunlop (1950), that a union, to maximize its utility, seeks to "trade off" the real wage rate against the unemployment of its members, raising the former (relative to productivity) until the gain from a further real wage increase is offset by the utility loss from the increase in unemployment expected to result from it. At an unemployment level below the unions' optimum, the unions then push up wage rates faster than productivity. But firms pass these higher costs on to consumers, so the real wage gains are frustrated, and as long as the government maintains the low unemployment level the rounds of inflation will continue. I have trouble applying such a model to the American economy. Almost three-quarters of the civilian labor force do not belong to unions. This fact casts doubt on the quantitative importance of the model. And perhaps the fact goes much deeper. If the union members whom the unions make unemployed have no good prospect of future union employment, they will be inclined to seek employment elsewhere. If, at the other extreme, the union unemployment is shared in the form of a short workweek, this unemployment-while real enough to the extent that members do not "moonlightM-does not add to the official unemployment rate as it is measured. Certainly the unionsparticipate in the cost inflation process, and they may even increase a little the volume of unemployment consistent with price stability. But I should think that a union must offer its membership a frequency of employment opportunities that is roughly comparable to that elsewhere in order to thrive and that appreciably reduced employment opportunities require a greater wage differential between union and other employment than is commonly observed.* Phillips' successful fitting of what we now call the Phillips curve (1958) to a scatter diagram of historical British data deprived the discussions of some of their institutional color, but epitomized the new concept of cost inflation-if by that term we mean (as I think most of the aforementioned writers intended) that kind of inflation which can be stopped only by a reduction of the employment rate through lower aggregate demand and which despite structural changes, so that money wages will keep pace with prices until unemployment is allowed to increase, seems to me to be terribly implausible. In any case, if this paper is right, cost inflation theory does not require any such "double monopoly" argument. It is certainly likely, however, that an increase of union power, even if localized, will raise the average money-wage level at any constant unemployment rate (see Hines, 1964)

MONEY-WAGE DYNAMICS AND LABOR-MARKET EQUILIBRIUM thus raises a cruel dilemma for fiscal and monetary policy. The Phillips curve portrayed the rate of wage change as a continuous and decreasing function of the unemployment rate, with wage increases exceeding typical productivity growth at sufficiently low albeit above-minimum unemploy ment rates. Hence, if prices are tied to marginal or average costs, the smaller the level at which aggregate demand sets the unemployment rate the greater is the continuing rate of infiation Strikingly, Phillips found that the nineteenth-century data pointed to a trade-off between wage increases and unemployment in the same way as contemporary data. Lipsey's sequel(1960) showed a statistically significant Phillips-curve relation for the subperiod 1861-1913. In fact, this early Phillips curve was higher(by about one percentage point) than the Phillips curve he fitted to the period 1929-57. 6 Apparently the cost inflation ten dency, if real, is not""in history; in Britain anyway it may be no worse than it used to be But is the Phillips trade-off real, serious, and not misleading? I shall discuss briefly two challenges to the Phillips curve to which this paper is relevant. The first is the question of whether the slope of the wage increase- unemployment relation is great enough to pose a serious dilemma for ggregate demand policy. Though proponents of an American Phillips curve had tough sledding at first--numerous other variables were held to be important(Bowen, 1960; Bhatia, 1962; Eckstein and Wilson, 1962)- Perry's synthesis (1964)of much of this early work left a quantitatively important role for the unemployment rate(as well as for the profit rate and the rate of change of prices)in explaini nts in U.S. manufacturing. But in 1963 Bowen and Berry(1963)found that the decrease of the unemployment rate was far more important than the level of the unemployment rate in contributing to wage increases. The recent study of annual long-term wage data by Rees and Hamilton (1967)also showed a negligible (and statistically insignificant) relation between the eady-state unemployment rate and the rate of wage increase(though s By contrast, in the pure"demand inflation"of Keynes and the classics, a reduc ion of the price trend could be achieved without cost to output and employment since aggregate demand is necessarily superfluous to begin with "Demand inflatio ay be worth preserving, since a regime of mixed inflation"is conceivable My earlier paper(1961)contains a fairly complete taxonomy of inflations(see also Fellner, 1959). Incidentally, the occasional definition of cost inflation as an autono mous upward shift of the Phillips curve is very awkward and does not imply the policy dilemma"with which inflation analysts were concerned in the fifties 1862-1913 regression(his equation [10]) predicts a 2.58 per cent wage increase an nually, while the 1929-57 regression (his equation [13])predicts a 1.65 per cent 四 the same 3 per cent productivity growth in both periods, for bility would have permitted smaller unemployment in the latter s Table 2(p. 30)is evidence of the early Phillips curves under age increases after World War II

MONEY-WAGE DYNAMICS AND LABOR-MARKET EQUILIBRIUM 681 thus raises a cruel dilemma for fiscal and monetary p01icy.~ The Phillips curve portrayed the rate of wage change as a continuous and decreasing function of the unemployment rate, with wage increases exceeding typical productivity growth at sufficiently low albeit above-minimum unemployment rates. Hence, if prices are tied to marginal or average costs, the smaller the level at which aggregate demand sets the unemployment rate the greater is the continuing rate of inflation. Strikingly, Phillips found that the nineteenth-century data pointed to a trade-off between wage increases and unemployment in the same way as contemporary data. Lipsey's sequel (1960) showed a statistically significant Phillips-curve relation for the subperiod 1861-1913. In fact, this early Phillips curve was higher (by about one percentage point) than the Phillips curve he fitted to the period 1929-57.6 Apparently the cost inflation tendency, if real, is not "new" in history; in Britain anyway it may be no worse than it used to be. But is the Phillips trade-off real, serious, and not misleading? I shall discuss briefly two challenges to the Phillips curve to which this paper is relevant. The first is the question of whether the slope of the wage increaseunemployment relation is great enough to pose a serious dilemma for aggregate demand policy. Though proponents of an American Phillips curve had tough sledding at first-numerous other variables were held to be important (Bowen, 1960; Bhatia, 1962; Eckstein and Wilson, 1962)- Perry's synthesis (1964) of much of this early work left a quantitatively important role for the unemployment rate (as well as for the profit rate and the rate of change of prices) in explaining money-wage movements in U.S. manufacturing. But in 1963 Bowen and Berry (1963) found that the decrease of the unemployment rate was far more important than the level of the unemployment rate in contributing to wage increases. The recent study of annual long-term wage data by Rees and Hamilton (1967) also showed a negligible (and statistically insignificant) relation between the steady-state unemployment rate and the rate of wage increase (though By contrast, in the pure "demand inflation" of Keynes and the classics, a reduction of the price trend could be achieved without cost to output and employment, since aggregate demand is necessarily superfluous to begin with. "Demand inflation" may be worth preserving, since a regime of "mixed inflation" is conceivable. My earlier paper (1961) contains a fairly complete taxonomy of inflations (see also Fellner, 1959). Incidentally, the occasional definition of cost inflation as an autonomous upward shift of the Phillips curve is very awkward and does not imply the "policy dilemma" with which inflation analysts were concerned in the fifties. At a constant price level and an unemployment rate of 2 per cent, Lipsey's (1960) 1862-1913 regression (his equation [lo]) predicts a 2.58 per cent wage increase annually, while the 1929-57 regression (his equation [13]) predicts a 1.65 per cent annual increase. At the same 3 per cent productivity growth in both periods, for example, price stability would have permitted smaller unemployment in the latter period. But Lipsey's Table 2 (p. 30) is evidence of the early Phillips curve's underestimation of the wage increases after World War 11