Keynes versus the classicists on the effects of wage and price deflation

Keynes spent most of the 1920s arguing that Britain's economy was "stuck in a rut" with persistently high unemployment, but the situation worsened considerably in the 1930s.

To most economists who call themselves Keynesian, the claim that destructive disequilibrium processes are a crucial part of Keynes's eco­nomics would seem highly questionable, if not bizarre. This is not just because many never carefully read The General Theory from start to finish, but also because the book is almost universally understood to concern itself exclusively with short-term, stable equilibriums and to rely heavily on comparative static exercises that assume stability of equilibrium to derive economic policies.

This is partly Keynes's fault. His main policy goal was to convince economists and other influential people that a permanent increase in public investment would inevitably lead to a much larger (or multi­plied) increase in production and a substantial permanent decline in unemployment.

Recall that there are two essential processes in classical theory that quickly restore general equilibrium if the economy suffers a negative exogenous shock to AD while in general equilibrium. The first is that an excess labor supply will cause wage and price deflation that inevitably lead to real-wage deflation and thus to an assured rise in employment and output. Real-wage deflation will continue until the excess supply of labor is eliminated. The second disequilibrium process that restores gen­eral equilibrium is found in financial markets where any negative shock to AD is met by a quick and sharp drop in interest rates that restores AD to full-capacity level. We will examine Keynes's critique of the classical theory of efficient financial markets when we treat chapters 12, 13, 15 and 22 below.

The classical vision of an ideal capitalist economy requires that money­wage deflation always takes place without restriction in the face of high unemployment, that prices fall by less than wages, and that real-wage deflation always causes employment to increase until full employment is restored. In The General Theory, Keynes argued that none of these assertions are true - classical theory had the essential facts of disequilib­rium dynamics wrong.

We have already seen that Keynes persistently argued that serious deflation had catastrophic effects. He railed against repeated attempts by the government to deliberately create wage deflation in traditional export industries by increasing unemployment, calling the policy an economic disaster and a violation of basic principles of social justice. The fact that Keynes used the first substantive chapter of The General Theory to attack the classical theory of deflation indicates how important it was to him to destroy the idea that Britain could deflate its way to prosperity. Keynes also argued that the USA and global financial systems had become so fra­gile and over-leveraged that the serious deflation of the early 1930s had triggered a financial market collapse and global depression. Deflation under conditions of financial fragility is especially disastrous. This was a major theme of his 1923 Tract on Monetary Reform (CW 4) with its emphasis on a "regime of money contract."

In chapter 2, Keynes focused on the classical theory of the labor market, the market that directly determines the state of employment. He accepted what he referred to as the first postulate of classical theory - that the real wage is always equal to the marginal product of labor (MPL) or the increase in output caused by the addition of one more worker. In other words, we are always on the classical labor demand curve. Just as

Effects of wage and price deflation 211 in the neoclassical short-run theory of the labor market, classical theory assumed - in obvious conflict with reality - that the whole capital stock (K) is utilized no matter how many workers (L) are employed. Capital is a "jelly" that can be spread thick (if L is low) or thin (if L is high).¹ In the classical (and neoclassical) short run, it is assumed that the MPL always declines as output and employment rise and always falls when unemployment rises. This means that the real wage always falls as unemploy­ment rises.²

There were a number of important "facts" that Keynes believed the classicists got wrong about the labor demand curve.

The fact is that in the USA, with weak unions, nominal wages in manu­facturing fell by 21 percent from 1929 to 1933, but real wages rose by 4 percent as unemployment increased from 4 to 25 percent.³ Prices fell by slightly more than wages as the result of massive excess capacity and more intense competition, forces not considered in classical disequilibrium theory.⁴ In the UK, where unions were strong, nominal wages stayed rela­tively constant in manufacturing from 1929 to 1936, while the real wage rose by at least 10 percent because prices fell.⁵

Keynes concluded: "The change in real wages associated with a change in money wages... is almost always in the opposite direction" (CW 7, p. 10). This "fact" implies that the crucial classical assumption that wage deflation is a mechanism that automatically eliminates involuntary unemployment in a capitalist economy was not true. He observed that classical economists had never even attempted to empirically verify the central classical assumption that a fall in the money wage automatically leads to a fall in the real wage: "it is strange that so little attempt should have been made to prove or to refute it" (CW 7, p. 12).

Keynes offered two reasons as to why this assumption should have been suspect even to classical economists. The first reason has to do with the nature of competition, assumed by classical theory to always be fierce or "perfect."

For it is far from being consistent with the general tenor of the clas­sical theory, which has taught us to believe that prices are governed by marginal prime cost in terms of money and that money wages govern marginal prime cost.

in almost the same proportion, leaving the real wage and the level of employment practically the same as before.

(CW 7, p. 12)

In the real world of the 1930s, fierce competition in an environment of high unemployment and crushing excess capacity should have been expected to drive prices down as fast or faster than marginal cost was falling - which is what actually happened. The core problem is that in the world of classical theory, there was no attempt to build an empirically and institutionally realistic theory of out-of-equilibrium processes, one that acknowledges the effects of excess capacity and destructive competition.⁶ Unrealistic stabil­izing out-of-equilibrium processes are assumed to exist in classical theory because they are necessary to defend laissez-faire policy. Recall that in "Am I a Liberal?" Keynes argued that classical "perfect competition" is a theory of a blissful state; it does not contain any theory of potentially destructive disequilibrium processes in real time. We have already seen Keynes's ana­lysis of the destructive character of intense competition in the British cotton and coal industries in the 1920s, as well as his argument that "perfect com­petition," which assumes that price equals marginal cost, would destroy the growing number of important industries with large economies of scale and, therefore, large fixed costs per unit. In his view, the classical theory of competition as a process was utterly bankrupt.

The second reason to reject the classical argument that real and money wages move in the same direction is that its assumptions are inconsistent. High unemployment may well force workers to accept lower money wages if they are not well-organized as in the USA in the 1920s and early 1930s. However, when employment falls, the classical demand for labor or the MPL function shows that the real wage rises because output per worker rises - which contradicts the central claim of the theory of efficient disequilibrium processes.

[W]hen money-wages are rising... it will be found that real wages are falling, and when money-wages are falling, real wages are rising. This is because, in the short period, falling money-wages and rising real wages are each, for independent reasons, likely to accompany decreasing employment; labour being readier to accept wage-cuts when employment is falling off, yet real wages inevitably rising in the same circumstances on account of the increasing marginal return to a given capital equipment when output is diminished.

(CW 7, p. 10)⁷

Second, involuntary unemployment is a "fact" that should not be theorized away for political or ideological convenience. In the depression,

Effects of wage and price deflation 213 there were millions of unemployed workers in the USA and Britain who wanted to work at the going wage but could not get jobs. Several statements in the chapter make Keynes's view on this clear: "Who would deny it?" he asked. (Well, Robert Lucas, Edward Prescott, and Thomas Sargent, among others.)

Third, unless workers have full cost-of-living adjustments in their wage contracts or wages are set by national bargaining among labor, capital, and the state, workers have no way to negotiate a bargain with employers for a real as opposed to a money wage. Labor as a whole cannot make a bargain with employers in which they trade lower real wages for more jobs, as the narrative associated with classical theory suggests.⁸ Classical theory assumes that if all workers would accept lower nominal wages, the real wage would fall, leading to lower unemployment. The "most fun­damental objection... flows from our disputing the assumption that the general level of real wages [and thus employment] is directly determined by the character of the wage bargain" (CW 7, p. 13). "There may exist no expedient by which labour as a whole can reduce its real wage to a given figure by making revised money wage bargains with the entrepreneurs" (CW 7, p. 13, emphasis in original).

Given the money wage, which is set through wage bargaining, the real wage will be determined by the aggregate price level.⁹ But the price level is directly determined in the goods market, not the labor market. This means that the real wage cannot be established until conditions in the output market are taken into account. AD - not the money wage - is the main determinant of the demand for labor. Keynes insisted that the MPL curve is not a demand curve for labor. Rather, given the demand for labor, which is primarily determined in the goods market, the MPL curve translates the level of labor demanded into an appropriate real wage.¹⁰ To simplify just a bit, Keynes said that AD determines labor demand through the production function, and then the amount of labor demanded determines real wage along the MPL curve - and not the other way around.

Finally, Keynes built his innovative, non-classical labor-supply function in chapter 2 based on the institutional fact that where organized labor is strong, workers will, if they can, resist money-wage cuts, causing the money wage to become downwardly rigid or at least downwardly sluggish. They will try to fight money-wage cuts for two main reasons.

First, in a depressed economy, the elasticity of employment with respect to a fall in the money wage will be very low. If steel workers take a pay cut in a depression, this will not lead to a significant expansion of jobs in the steel industry because the rate of growth of steel industry employment depends on the growth of production in steel-using indus­tries such as capital goods, autos, and consumer durables. Their growth of production in turn depends on the rate of growth of economy-wide AD, which steelworkers cannot significantly influence. Keynes said that

workers instinctively know this: "workers, though unconsciously, are instinctively more reasonable economists than the classical school" (CW 7, p. 14).

Second, workers also know that "reductions of money wages are seldom or never of an all-around character." Workers in stronger unions will be better able to strike or work to rule or otherwise obstruct produc­tion in order to resist wage cuts than workers in weak unions or non­unionized workers. Workers with less or no bargaining power therefore will be unable to resist wage cuts and will drop down the relative wage and status ladder, an outcome that is socially unjust and that workers know is unjust. All workers will thus try to resist money-wage cuts.

Since there is an imperfect mobility of labour, and wages do not tend to an exact equality of net advantage in different occupations, any individual or group of individuals, who consent to a reduction of money-wages relative to others, will suffer a relative reduction in real wages, which is a sufficient reason to resist it.. In other words, the struggle about money-wages primarily affects the distribution of wages between the different labour groups.

(CW 7, p. 14, emphasis in original)

Except in a socialized community where wage-setting is settled by decree, there is no means of securing uniform wage reductions for every class of labour. The result can only be brought about by a series of gradual, irregular changes, justifiable on no criterion of social justice or economic expedience, and probably completed only after wasteful and disastrous struggles where those in the weakest bargaining positions will suffer relative to the rest.

(CW 7, p. 267)

In interwar Britain, workers did indeed strike in the face of money-wage cuts, even when unemployment was high. The militant General Strike of 1926 was only the most prominent example of this.

Keynes made similar comments in other writings. In 1930: "A reduc­tion of wages [in Britain] can only be [achieved] as a result of a sort of civil war or guerrilla war carried on, industry by industry, all over the country, which would be a hideous and disastrous prospect" (CW 20, p. 419). In 1931: "In my country a really large cut in money wages. is simply an impossibility. To attempt it would be to shake the social order to its foun­dation" (CW 20, p. 546).

Keynes believed that though workers will struggle against and strike to prevent money-wage cuts if they can, they are likely to accept modest real-wage declines that arise from moderate price increases because this will not change their place in the relative wage distribution.

Effects of wage and price deflation 215

Every trade union will put up some resistance to a cut in money wages, however small. But since no trade union would dream of striking on every occasion of a rise in the cost of living, they do not raise the obs­tacle to any increase in aggregate employment which is attributed to them by the classical school.

(CW 7, p. 15)

Note that, contrary to textbook treatments of Keynes, downward money wage rigidity is not caused by "money illusion" - workers' inability to dis­tinguish between nominal wages and the purchasing power of their wage. There is no mention of money illusion in The General Theory.

The derivation of Keynes's labor supply function is clearly based on an anti-classical methodology and reflects a methodological innovation. It is not derived by deduction from optimization processes over standard pre-given agent preferences given complete and correct information about the future, but rather inductively by a study of the facts. It is a historic­ally, behaviorally, and conventionally determined function. Workers resist changes in relative wages as much for "moral" as economic reasons. Wage cuts that lower their position in the vector of wages across jobs and indus­tries are considered unfair.¹¹

Of course, the real wage must also be in the labor supply function because workers are not irrational: they care about the purchasing power of their wage. For Keynes, labor supply is a function of both the current money wage and the current real wage. This means that every time the money wage changes in a disequilibrium process, the classical and neo­classical labor supply function shifts. The classical presumption that wage and price deflation will automatically cure unemployment cannot be derived from such a labor supply function.

The attack on the alleged employment-creating benefits of deflation as the main path to full employment was so important to Keynes that he returned to it in chapter 19, after the basic macro model in the book had been constructed. To discuss the arguments in chapter 19, we first have to briefly review the basic structure of the simple Keynesian AD/AS model.

To simplify somewhat for present purposes, we might say that in Keynes's theory, employment and income are primarily determined by AD - Say's Law does not hold. In his simple model with no government, AD is the sum of capital investment spending, consumption spending, and net exports. As we will discuss in detail below, Keynes assumed that future economic states are unknowable in the present. This means that corporations and households have to form expectations of the future based on mere guesswork or behavioral conventions to make decisions about how much to invest and how much of their current income to spend on consumption goods. The value of AD in any period thus partly depends on the state of expectations in that period.

As we shall see in our discussion of chapter 12, Keynes argued that expectations of the future are normally formed through extrapolation or projection from the relevant past. The longer this expectation formation process generates forecasts that are accurate enough to seem serviceable, the more "confidence" people will place in them. When people have confi­dence in optimistic expectations, they will be willing to spend more of their income on consumption goods and services and businesses will spend more on capital investment. When they have pessimistic expectations or have lost confidence in their ability to forecast with reasonable accuracy, they will spend less. Therefore, expectations must be incorporated into an exploration of the likely effects of falling money wages on employment in Keynes's model.

Chapter 19 reflects Keynes's insistence that analyses of out-of­equilibrium dynamics can only be adequately addressed if they allow key variables normally held constant in comparative-static exercises to become endogenous. This is especially important with respect to expectations and confidence. Classical theory asked a comparative-static question: would a once-and-for-all fall in the money wage raise employ­ment holding AS and, implicitly, expectations and confidence constant? Keynes argued that the appropriate question is: would a process of falling money wages over time with nothing arbitrarily held constant eliminate unemployment? How will a time-consuming process of falling wages affect the determinants of AD - the mec, the mpc, the rate of interest, and the trade balance? Keynes acknowledged that deflation would help the trade balance, but argued that its overall effect would be to lower AD and employment.

I want to focus on three issues discussed in chapter 19: the effects of deflation on expectations of future wages and prices, on the interest rate, and on financial fragility. Comparative statics are useless here because they incorporate the assumption that expectations never change in the disequilibrium process. Keynes argued that the most likely situation in a depressed economy is - in the absence of strong unions - a process of falling money wages over time caused by an excess supply of labor. He said: "The most unfavorable contingency is that in which money wages are sagging downwards and each reduction in wages serves to diminish confidence in the prospective maintenance of wages [in the near future]" (CW 7, p. 265). But this is the most likely outcome in periods of sustained high unemployment in the absence of strong unions.

Keynes said that in a depressed economy expectations of future prices are elastic with respect to falling current prices. He argued that when prices are expected to continue falling, the purchase of capital goods and consumer durables will be postponed in anticipation of lower prices in the future. This means that AD and employment will continue to fall, maintaining downward pressure on money wages. The takeaway here is that downwardly flexible wages in a period of high unemployment can

Effects of wage and price deflation 217 result in a deflationary wage and price spiral such as the one that took place in the USA in the early 1930s.

Keynes also examined what is often called the "Keynes effect," in which falling prices combined with a constant money supply lead to a rising real money supply and a consequent reduction of the interest rate, ceteris paribus.¹² Keynes made two arguments against this proposition in chapter 19. First, the money supply is endogenous. As nominal income falls, the transactions demand for money falls. But both the demand for and supply of credit fall as well, which will lead to a decline in the money supply. Second, a steep drop in the nominal wage associated with high unemployment is likely to be accompanied by economic and political tur­bulence, as it was in Britain in the 1920s and in the USA in the 1930s. This causes investors to shun risk and seek capital safety. They will sell long­term securities that can suffer large capital losses and shift to liquid short­term assets that cannot suffer a serious capital loss. Stock prices will fall and interest rates, especially risk- and inflation-adjusted interest rates, will rise - as they did in the deflation of the 1930s. "If, moreover, the reduction in wages disturbs political confidence by causing popular discontent, the increase in liquidity preference due to this cause may more than offset the release of cash from active circulation" (CW 7, pp. 263-264). In an IS/ LM model, a rise in liquidity preference (or the desire to hold less of your wealth in the form of bonds) would be represented by an upward shift in the LM curve.

Moreover, when prices, incomes, and the value of collateral assets collapse in a financially fragile "regime of money contract," the nominal value of debts remains constant, but the real value of debts rises dramat­ically - as in the 1930s. Nominal incomes decline, but nominal debt values do not. This can lead to a wave of defaults and bankruptcies, the evapor­ation of new loans, and an unwillingness to roll over existing loans.

The depressing influence on entrepreneurs of their greater [real] burden of debt may partly offset any cheerful reactions from the reduc­tion of wages. Indeed, if the fall of... prices goes far, the embarrass­ment of those entrepreneurs who are heavily indebted may soon reach the point of insolvency, with severely adverse effects on investment.

(CW 7, p. 264)

We have seen that Keynes believed the combination of treacherously fra­gile balance sheets and rapidly collapsing asset values in the deflation of the early 1930s was a major cause of the depression.

Keynes went on to attack the widely held thesis that capitalism would be capable of generating sustained full employment if only there were no strong unions to limit or prevent money-wage deflation. The conventional wisdom of the time in Britain was that the main cause of sustained high unemployment in the era was not inherently ineffective disequilibrium

dynamics in free-market capitalism, but rather the existence of strong unions that prevented adequate money-wage deflation. This was the firm belief of classical economists and the presumption relied on in policy for­mulation. Support for the lockout in the coal industry in 1926 and for the brutal repression of the General Strike that followed was motivated not only by fear of revolution, but also by the semireligious belief in the neces­sity of driving down miners' wages to restore British trade dominance in coal as part of a general strategy of forcing large nominal wage cuts in all major export industries. Lower export costs would also facilitate the return to gold at prewar par. And, of course, price deflation was also attractive because it would enrich the politically powerful rentier class who owned long-term bonds and received interest payments denominated in nominal values.

Keynes insisted that free-market capitalism - not unions - was the cause of high unemployment. Where is that mentioned in your macro textbook? He attacked the claim that union-supported money-wage rigidity was the chief impediment to full employment in the strongest possible terms, arguing to the contrary that if a process of substantial wage deflation set in, it could completely destabilize the economy. British unions were performing a service to Britain by preventing a catastrophe, he argued.

It follows therefore that if labour were to respond to conditions of gradually [rising unemployment] by offering its services at a gradually diminishing money wage, this would not, as a rule, have the effect of reducing real wages and might even have the effect of increasing them, through its adverse influence on the volume of output.¹³ The chief result of this policy would be to cause a great instability of prices, so violent perhaps as to make business calculations futile in an economic society functioning after the manner of that in which we live. To suppose that a flexible wage policy is a right and proper adjunct of a system which on the whole is one of laissez-faire, is the opposite of the truth. It is only in a highly authoritarian society, where sudden, substantial, all- around changes could be decreed that a flexible wage-policy could function with success.

(CW 7, p. 269, emphasis added)

This is not the only place in The General Theory where Keynes argued that the unrestricted downward flexibility of wages - a cornerstone of clas­sical stability analysis - would be disastrous. "For if competition between unemployed workers always led to a very great reduction in money wages, there would be violent instability in the price level" (CW 7, p. 253). As Keynes argued in the Tract on Monetary Reform, violent instability of prices would create intolerable uncertainty and could easily trigger both a collapse of capital investment and a financial crisis in a state of systemic

Effects of wage and price deflation 219 financial fragility. I have never seen this view - that a substantial degree of downward wage and price rigidity is a condition of existence of a cap­italist free-market economy - attributed to Keynes.

Notes

1 This assumption makes excess capacity zero at all levels of production.

2 There was no compelling economic reason for Keynes to accept the classical MPL function. Indeed, he published an article in 1937 (Keynes 1937) that rejected it. The article cited empirical work by Dunlop and by Kalecki that showed that over a wide range of output, the MPL is constant. This implies that the real wage is constant - deflation cannot lower the real wage. My guess is that Keynes accepted the classical first postulate to sooth the feelings of economists who were open-minded but not yet converted to his views, the people who were his main target audience. It seems that in The General Theory Keynes was willing to accept as much of received doctrine as he could while still being able to develop a theory that supported his radical interventionist policies.

3 Data from Carlstrom and Fuerst (2001). See also United States Bureau of the Census (1975).

4 The classical theory of the labor market assumed perfect competition or com­petition of maximum intensity; variations in the degree of competition and excess capacity were assumed away.

5 Total labor income fell because of increased unemployment and fewer hours worked per week.

6 The theory of destructive competition, associated both with Schumpeter, who explored the process of "creative destruction," and with Marx is explored in Crotty (2017, chapters 10 and 11). Schumpeter assumed the forces of creation were always stronger than those of destruction because he assumed Say's Law always holds.

7 The last fragment of the quote is puzzling because when L falls, K/L rises and L/K falls, which should lower the marginal product of capital. What Keynes presumably means here is that since it is assumed in the "short run" that cap­ital is fully employed no matter how many workers there are, the K/L ratio rises as employment declines, causing the MPL to rise. Since real wages are positively related to the MPL, they will rise as well.

8 This is not an explicit assumption: classical theory understands that the wage bargain is only over a money wage. But since it assumes that money-wage cuts automatically translate into real-wage cuts, it is as if the bargain is over a real wage. "The traditional theory maintains, in short, that the wage bargains between the entrepreneurs and the workers determine the real wage; so that, assuming free competition amongst workers, the latter can, if they wish, bring their real wages into conformity with the marginal disutility of the amount of employ­ment offered by the employers at that wage" (CW 7, p. 11, emphasis in ori­ginal). In reality, workers can agree to take a money-wage cut, but they have no control over the determination of the price level.

9 In the absence of unions and/or tight labor markets, the wage "bargain" is simply a take-it-or-leave-it dictate of the employer.

10 I use the word "primarily" because the real wage has a feedback effect on AD.

11 The mec (defined in chapter 11 of The General Theory) and the impact of liquidity preference on the interest rate are two other examples of convention­ally or behaviorally determined variables.

12 As the amount of money needed to facilitate transactions declines due to deflation, it creates an excess supply of money that will be used to buy bonds, driving bond prices up and interest rates down.

13 Recall that as output falls, the MPL rises. If we assume that the economy is always on the MPL curve, unemployment and real wages move in the same direction.

16