I. INTRODUCTION
The Phillips curve analysis is now employed to demonstrate the neutrality of money on real output and employment in the long run, but money's non-neutrality in the short run when changes in the quantity of money are unanticipated. The classical forced-saving doctrine – the proposition that rising prices cause wage, interest, and rental income earners to suffer a reduction in real consumption while profit rates increase in the short run because of the lagged adjustment of producers’ costs (wages, interest, and rentals) – which is commonly traced to Jeremy Bentham or Henry Thornton derives the same conclusion. Indeed, David Ricardo (4, pp. 36–37; italics added) traces the forced-saving doctrine to David Hume:
Mr. Malthus notices an observation, which was first made by Hume, that a rise of prices, has a magic effect on industry: he states the effects of a fall to be proportionally depressing…. It is said to be beneficial, because it betters the situation of the commercial classes at the expense of those enjoying fixed incomes; – and that it is chiefly in those classes, that the great accumulations are made, and productive industry encouraged.”Footnote 1
Phillips's (Reference Phillips1958) original analysis has little use for the classical quantity theory of money as a theory of the price level or the forced-saving doctrine derived from it. Fisher's (Reference Fisher1926) article explaining the phenomenon of rising prices along with a falling rate of unemployment in the short run, followed by rising nominal wages, interest, and rental rates because of increases in the quantity of money or credit (and vice versa), draws on the classical principle, but without reference to its name, other than “the principle underlying this relationship is of course, familiar” (1926, p. 787).Footnote 2 Some recent analysts also have made the connection between Phillips's work and Hume's monetary analysis and have cast Hume among believers in a downward sloping, long-run Phillips curve, such as Nelson (Reference Nelson1981a) and Humphrey (Reference Humphrey1982, Reference Humphrey1985), a claim Frenkel (Reference Frenkel1981) appropriately disputes.Footnote 3
In fact, several major classical economists frowned upon the injustice of the inflationary tax imposed on fixed-income earners by an excessive money creation intended to create the short-run positive effects on output and employment. That is why they did not recommend the forced-saving mechanism as a growth-promoting tool. Discussion of the forced-saving doctrine now tends to be consigned to texts in the history of economic thought, with hardly any mention in modern macroeconomics. Austrian macroeconomists who discuss the concept employ it rather to explain the changing structure of production, and thus do not relate it directly to the Phillips curve analysis; see for example Garrison (Reference Garrison2004). I seek also to bridge this gap in macroeconomics.
Indeed, before John Maynard Keynes's economics took hold in modern macroeconomics, employing the forced-saving mechanism was a commonplace; see Alfred and Mary Marshall (Reference Marshall and Mary Paley1879), Marshall (Reference Marshall1920), Fisher (Reference Fisher1912, Reference Fisher1913), Pigou (Reference Pigou1913, Reference Pigou1927, Reference Pigou1933), Hawtrey (Reference Hawtrey1913, Reference Hawtrey1919), Taussig (Reference Taussig1921), and Robertson (Reference Robertson1926), although without explicit reference to the name. Pigou, for example, used “forced levies” while Robertson used “imposed lacking”; see Hayek (Reference Hayek1933, pp. 218–19n) for some equivalent references to the classical forced-saving doctrine.
Keynes's influence in modern macroeconomics may be responsible for the absence of the doctrine's mention.Footnote 4 First, Keynes (Reference Keynes1936, p. 80) argues that the concept is relevant only to an economy operating at full employment, but which is mistaken. The lagged adjustment of wages, interest, and rental rates to a changing price level occurs because of the existence of contracts and/or lack of full anticipation of the behavior of prices, and this may take place whether or not an economy is at full employment. Ahiakpor (Reference Ahiakpor1997a) further explains Keynes's misattribution of the full employment assumption to the forced-saving doctrine. Secondly, Keynes thinks little of the concept because, he argues, “the new money is not ‘forced’ on anyone” (1936, p. 328), as if saving meant the hoarding of cash, but which it does not; Ahiakpor (Reference Ahiakpor1995) elaborates. Thirdly, Keynes (Reference Keynes1930, 1, p. 171n) attributes the doctrine mainly to “German writers” or Austrian economists, a tendency some analysts, including Steindl (Reference Steindl1995) and Laidler (Reference Laidler1991; Reference Laidler1999), to a considerable extent, still follow.Footnote 5 Ironically, Austrian economists, such as Mises (Reference Mises and Batson1934, Reference Mises1963), Hayek (Reference Hayek1931, Reference Hayek1932, Reference Hayek1933), and Haberler (Reference Haberler and Wright1932) trace the doctrine to the classical quantity theory of money.
Thus, Friedman (Reference Friedman1968, Reference Friedman1977) could have employed the classical forced-saving doctrine to make his point about the illusion of attempting to reduce unemployment below its natural rate by money creation and the need rather to regulate the quantity of money so that its variations do not become the source of economic fluctuations. The classical forced-saving phenomenon also could have assisted the research efforts of the new classicals in the 1970s and 80s; see Lucas (Reference Lucas1996). Indeed, one really does not need “the equipment of modern mathematical economics” to explain the short-run non-neutrality of money on output and employment from the phenomenon of lagging factor prices behind changes in the price level as Lucas (Reference Lucas1996, p. 669) suggests. All one needs is a lack of full anticipation of changes in the quantity of money and the price level, which is entailed in the classical explanation.Footnote 6
Furthermore, the fruitless efforts during the 1960s and 1970s to estimate reliable trade-off coefficients between rates of inflation and unemployment to guide macroeconomic policy (see Leeson Reference Leeson1997) might have been spared had analysts sooner made the connection between the Phillips curve analysis and the classical forced-saving doctrine. The doctrine explains the non-permanency of any short-run trade-off between inflation and unemployment. Besides, the period of short-run lags that produces the trade-off cannot be expected to be the same within a country over different inflationary (or deflationary) epochs, since people do learn over time and the institutional features that generate the factor price rigidities do change as well. Even now, the search for a “new Keynesian Phillips curve” continues; see for example Batini, Jackson, and Nickell (Reference Batini, Jackson and Nickell2005), in spite of Solow's (Reference Solow and Puttaswamaiah2002, p. 74) disavowal that he and Samuelson meant to encourage the search for “a menu of choice” with the Phillips curve. How much wasteful research effort could be saved by learning from monetary history.
II. THE ESSENCE OF THE PHILLIPS CURVE ANALYSIS
Phillips (Reference Phillips1958) reveals a fundamental pattern of behavior between the rate of unemployment and changes in money wage rates. For eight business cycle periods between 1861 and 1913 he finds that, on average, nominal wage rates rose faster in the United Kingdom during periods of low rates of unemployment and rose less fast or fell during periods of high rates of unemployment. Phillips explains the observed pattern by (1) the tendency of employers to bid up wages in times of business boom and to demand wage rate reductions during recessions, (2) the tendency of labor unions to demand cost of living adjustments during periods of rising product prices, and (3) a lag in the adjustment of wage rates behind changes in the rate of unemployment. A combination of these factors leads to the nonlinear negative curve he fits for the 1861-1913 period, while also noting a counterclockwise movement of pairs of wage rate changes and rates of unemployment over the business cycles. Subsequent period data, especially for 1913-1948, that include the two World Wars and Britain's return to the gold standard, are less “well-behaved.”
Indeed, one could draw a series of nonlinear, downward-sloping curves through the pairs of wage rate changes and unemployment rates above and below Phillips's fitted curve. Similarly, one could connect points of high and low (or negative) growth rates of nominal wages above and below the same unemployment rate to depict a vertical long-run Phillips curve. The same short-run curves could be drawn through the scatter diagram that Samuelson and Solow (Reference Samuelson and Solow1960, p. 188) produce using data on hourly earnings in manufacturing and rates of unemployment in the United States. Such depiction of the data has become the norm in subsequent analyses, with changes in the nominal wage rate being replaced by rates of inflation on the vertical axis, following Samuelson and Solow (Reference Samuelson and Solow1960), Phelps (Reference Phelps1967), and Friedman (Reference Friedman1977).
Phillips's explanations leave out the crucial factors determining changes in business activity that cause changes in the demand for labor. His approach was merely to infer changes in the demand for labor from changes in business conditions (1958, p. 283). Among the relevant determinants of changes in the demand for labor Phillips left out are (1) changes in the quantity of money (currency) or credit supplied by a central bank and (2) changes in the extension of credit by financial intermediaries, a reflection of their own demand for currency. Indeed, changes in the quantity of central bank money act directly on the supply of credit. Open market purchases or reductions in the discount rate increase the excess reserves of financial intermediaries, which then expand their loans or credit [S cr = S c + ΔH, where S cr = supply of credit, S c = supply of “capital” or savings = Y(1 − t) − C − ΔH h, Y(1 − t) = disposable income, C = consumption, H = currency or high-powered money, and H h = households’ hoarding of cash]. Interest rates fall below the level determined by the supply of savings relative to the demand for “capital” or investment fundsFootnote 7 – what the classicals called the “natural rate”; see Ricardo (1, pp. 91-92) and John Stuart Mill (3, p. 648). In fact, the “cumulative process” analysis, derived from deviations of the market rate of interest from the natural rate and commonly associated with Knut Wicksell, originates with the classics, such as Ricardo (1, pp. 363-64; 3, p. 91), which Wicksell (Reference Wicksell and Lindhal1935, pp. 79-80) himself acknowledges.
The increased supply of currency (H) increases the price level (P 1 = H 1/ky 0 > P 0 = H 0/ky 0, where k = proportion of income held as cash and y = real output or income), and while wage, rental, and interest rates are yet to rise, salaried workers, savers, and property owners suffer a diminution in their real consumption, hence the classics described the phenomenon as forced saving. (Some analysts fail to identify the lagged adjustment of wages, interest, and rentals behind the rise of product prices with the classical forced-saving doctrine, e.g., Humphrey (Reference Humphrey1991, pp. 6–9); perhaps because they identify the “capital accumulation” in the doctrine with only capital goods instead of with funds, e.g., Hansson (Reference Hansson, Eastwell, Milgate and Newman1987); see also note 5 above.) A decrease in banks’ excess or economic reserves in order to extend more credit amounts to a decrease in the economy's total demand for cash, H d = kPy, which also may raise the price level. In the absence of the central bank's credit expansion, the increased quantity of “capital” or investment funds would have been obtained only at a higher rate of interest.
On the other hand, a central bank's contraction of currency through an open market sale of securities reduces the reserves of financial intermediaries, causing them to contract their lending or supply of credit (S cr = S c − ΔH < S c). The resulting rise in interest rates (above the natural rate) and the excess demand for money thus created lead to a fall in the price level along with a slowdown in business activity. In the face of wage, interest, and rental contracts keeping factor costs from falling immediately, businesses incur reduced profits or losses. They lay off some workers and reduce production. A contraction of credit by financial institutions also has the same short-run deflationary effect on prices, output, and employment as a central bank's contraction of currency. Thus, it is the lagged adjustment of factor costs behind changes in product prices that generates the observed inverse relation between changes in nominal wage rates and rates of unemployment, not differences in perceptions of the price level between employers and workers, as Friedman (Reference Friedman, Eastwell, Milgate and Newman1987, p. 14) argues.
Fisher (Reference Fisher1926) relies on the above explanations to interpret his estimated relation between price-level changes and employment:
changes in the purchasing power of the dollar may very largely explain changes in employment. The principle underlying this relationship is of course familiar. It is that when the dollar is losing value, or in other words when the price level is rising, a business man finds his receipts rising as fast, on the average, as this general rise of prices, but not his expenses, because his expenses consist, to a large extent, of things which are contractually fixed, such as interest on bonds; or rent, which may be fixed for five, ten, or ninety-nine years; or salaries, which are often fixed for several years; or wages, which are fixed sometimes, either by contract or custom, for at least a number of months. For this and other reasons, the rise in expenses is slower than the rise in receipts when inflation is in progress and the price level is rising or the dollar falling…. Employment is then stimulated – for a time at least….
On the other hand, when prices are falling, expenses likewise lag behind and reduce profits, for exactly the same reason turned about. Consequently, during periods of falling prices profits are reduced, bankruptcies are increased, concerns shut down entirely or in part, and men are thrown out of work
(Fisher Reference Fisher1926, p. 787).Thus, Fisher concludes that
changes in the price level do definitely foreshadow or anticipate changes in employment … as the economic analysis already cited certainly indicates a causal relationship between inflation and employment or deflation and unemployment, it seems reasonable to conclude that what the charts show is largely, if not mostly, a genuine and straightforward causal relationship; that the ups and downs of employment are the effects, in large measure, of the rises and falls of prices, due in turn to the inflation and deflation of money and credit
(Fisher Reference Fisher1926, p. 792).Fisher (ibid., p. 788) also notes that the effects are temporary: “It is not a high price level that makes for full employment or a low price level that makes for unemployment. Whether a price level is high or low has, in the long run, nothing whatever to do with employment. … But if we sink from one [price] level to another, then during the time of falling, we do produce depression of trade and unemployment. Reversely, if we rise from one level to another, then during the period of rising we do for a time produce more employment.”
Fisher's Elementary Principles of Economics (1912), especially in chapter 10, as well as his Purchasing Power of Money (1913), especially in chapter 4, also employs the above analysis. In the latter work Fisher refers to “the salaried man, or the laborer, most of whom are silent but long-suffering, – paying higher prices, but not getting proportionally higher incomes” (1913, pp. 61–62) – to illustrate the forced saving imposed on wage earners. He also notes that “It is the lagging behind of the rate of interest which allows [business] oscillations to reach … great proportions” (1913, p. 71). Fisher further draws upon Alfred Marshall's (Reference Marshall1907) explanation that rising prices lower real interest rates and encourage increased business spending because “those working on borrowed capital pay back less real value than they borrowed, and enrich themselves at the expense of the community” (1913, p. 72; italics added); Marshall (Reference Marshall1920, pp. 493–94) restates the argument.
Pigou (Reference Pigou1927, p. 196) is an example of the familiarity of this type of macroeconomic analysis before Keynes's General Theory took hold. He interprets his own chart relating rates of inflation to changes in employment with some lag as being in “close accord” with Fisher's earlier work. And referring to the redistributive effects of unforeseen inflation, especially on wage earners and savers, Pigou (Reference Pigou1927, p. 157) observes that “price changes wrench the terms of contracts for loans and wages away from the terms that were intended when these contracts were made.” Similarly, Hawtrey (Reference Hawtrey1919, p. 262) considers the real wage reductions caused by the lagged adjustment of nominal wages as “one of the most serious evils of inflation.” Hawtrey (Reference Hawtrey1913, pp. 265–66) also explains fluctuations in output and unemployment by the lagged adjustment of nominal wages:
In a sense all the mischief of trade fluctuations arises from the tendency of changes in wages to lag behind changes in the value of money. The unemployment during bad trade is mainly the consequence of wages remaining on a level too high for the reduced stock of money. The fall of real wages during good trade is the consequence of delay in the increase of money wages corresponding to the increase in the stock of money. If the working classes would accept an early reduction of wages when a period of bad trade begins and if the employers would give an early increase of wages when a period of good trade begins, not only would the harmful consequences of a trade fluctuation be avoided, but the fluctuation itself might even be prevented.
Alfred and Mary Marshall (Reference Marshall and Mary Paley1879, pp. 156–57) earlier employed the lagged adjustment of factor costs to explain fluctuations in business profits and production:
when prices are rising, the rise in the price of the finished commodity is generally more rapid than that in the price of the raw material, always more rapid than in the price of labour; and when prices are falling, the fall in the price of the finished commodity is generally more rapid than in the price of the raw material, always more rapid than in the price of labour. And therefore when prices are falling the manufacturer's receipts are sometimes scarcely sufficient even to repay him for his outlay on raw materials, wages, and other forms of Circulating capital; they seldom give him in addition enough to pay interest on his Fixed capital and Earnings of Management for himself.
… a fall in prices lowers profits and impoverishes the manufacturer: while it increases the purchasing power of those who have fixed incomes … it impoverishes those who have to make, as most business men have, considerable fixed money payments for rents, salaries, and other matters.
Marshall (Reference Marshall1920, p. 515) repeats the explanation, noting that rising prices and profits enable employers to pay higher wages, but “experience shows that (whether they are governed by sliding scale or not) [wages] seldom rise as much in proportion as prices.”
Frank Taussig (Reference Taussig1921, pp. 298–99) gives the same explanations:
That wages go up more slowly than prices is one of the best-attested facts in economic history. It holds good of almost all sorts of hired persons…. It is due mainly to the force of custom, which is especially strong as to wages; … Of the fact there can be no question; when prices rise, the wages of hired workers do not rise as fast…. The business man who is nearest the ground, so to speak … profits from the relative stability of wages. Conversely, the business class as a whole commonly loses in periods of falling prices. Then, since the same forces tend to keep wages stable, a fall in prices brings loss. Probably wages feel the effect of falling prices less slowly than they do those of rising prices.
Fisher's (Reference Fisher1913) reference to the “long-suffering” salaried man or the laborer and Marshall's (Reference Marshall1920) to the delayed rise in wages also reflect the sentiments of several classical economists about forced saving as being unjust to fixed-income earners. Similarly, Taussig (Reference Taussig1921, p. 294) observes that a “different question of justice between debtor and creditor arises from the fact that money wages and other money incomes do not necessarily move in the same way as the prices of commodities.” Modern Phillips curve analysis is silent on the undesirability of the income or wealth redistribution effects of price level changes, a neglect worth reversing.
We next take up the major classicals’ discussion of the forced-saving doctrine. This will be followed by noting the equivalent analysis in modern macroeconomics.
III. THE CLASSICAL FORCED-SAVING DOCTRINE
Henry Thornton (Reference Thornton and Hayek1802) takes his cue in discussing the phenomenon of forced saving imposed on fixed-income earners from Hume's (Reference Hume and Rotwein1752) essay, “Of Money.” He first agrees that national income may increase following an increase in the quantity of money, adding that “we may presume an encrease of paper to have exactly the same effect” (p. 238). He then argues that
It must also be admitted that provided we assume an excessive issue of paper to lift up, as it may for a time, the costs of goods tho not the price of labor, some augmentation of stock [i.e. “capital” or investment funds] will be the consequence; for the labourer according to this supposition, may be forced by his necessity to consume fewer articles, tho he may exercise the same industry. But this saving, as well as any additional one which may arise from a similar defalcation of the revenue of the unproductive members of the society will be attended with a proportional injustice and hardship
(Thornton Reference Thornton and Hayek1802, p. 239; italics added).Footnote 8Jeremy Bentham, also following Hume's monetary analysis, decries the injustice and hardships that are inflicted upon fixed-income earners through the forced-saving mechanism. He first considers the case of full employment: “If all were fully employed, [the increase of paper money] could not encrease industry” (1, p. 270). Next, he considers the case of less than full employment, where both employment and real output may increase from an increase in the quantity of money or bank credit: “If any were unemployed, or not fully employed, [the increased quantity of money] might encrease industry pro tanto” (ibid.). Noting that the “consequence would be the same if, instead of coin, paper money were added (if not in such excess as to destroy its credit),” Bentham lists the effects of an increase of money as “1. A tax upon capital, i.e. upon the capital of moneyed men whose money is placed out at interest, in so far as their debtors are able to pay them off. 2. In [an]other respect it is a tax on consumption, in so far as it raises the price of all commodities. Meantime it actually is productive of an addition to the mass of national wealth, in as far as it gets extra hands, or sets them to work at extra hours” (1, p. 271; italics added). But the increased output derives from “forced frugality, … at the expense of national comfort and national justice” (3, p. 349; italics added).
Bentham relies on the non-anticipation of increases in the quantity of money and the resulting inflation to argue the increase in output and employment: “If the accession was universally known and the principles of economy under this head universally known likewise, the money prices of things [instantaneously] would be doubled” (1, p. 270) from a doubling of the quantity of money – a new classicals’ rational expectations argument. Thus, claims that Bentham assumed only full employment, according to which “the supply of output as a whole ceases to be elastic, i.e. where a further increase in the effective demand will no longer be accompanied by an increase in output” (Keynes Reference Keynes1936, p. 26), as by Keynes (ibid., pp. 80–81), Hollander (Reference Hollander1987, p. 270), and Humphrey (Reference Humphrey1991, p. 9), are inaccurate.
David Ricardo's monetary analysis draws insights from Hume's “Of Interest” and “Of Money”; Ricardo (3, p. 90; 4, pp. 36–37; 5, p. 12). He also makes similar points as Thornton and Bentham regarding forced saving: “Depreciation of money … augment[s] riches by diminishing happiness … and it [is] advantageous only by occasioning a great pressure on the labouring classes and on those who live on fixed incomes” (6, p. 233). In his “Notes on Bentham,” Ricardo (3, pp. 318–19; italics added) explains that
There is but one way in which an increase of money no matter how it be introduced into the society, can augment riches, viz. at the expense of wages of labour; till the wages of labour have found their level with the increased prices which the commodities will have experienced, there will be so much additional revenue to the manufacturer and farmer they will obtain an increased price for their commodities, and can whilst wages do not increase employ an additional number of hands, so that the real riches of the country will be somewhat augmented. A productive labourer will produce something more than before relatively to his consumption, but this can be only of momentary duration.Footnote 9
Ricardo also considers the increase in the real riches of the country through the forced-saving mechanism as being brought about by “a violent and unjust transfer of property” (3, p. 93), and “It must be accompanied with a degree of injustice to individuals which requires only to be understood to excite the censure and indignation of all those who are not wholly insensible to every honourable feeling” (3, p. 123).Footnote 10 Thus he insists that “the augmentation of goods is the only legitimate cause of an increase of money” (3, p. 302, n. 3; see also 3, pp. 139–40).
Like Hume (Reference Hume and Rotwein1752, p. 40), who considered an inadequate supply of money as being “pernicious to industry” and employment in the short run, Ricardo was concerned about the short-run negative output and employment effects of an inadequate supply of money because of the forced-saving mechanism's operation in the reverse. Thus to avoid “the most disastrous consequences to the trade and commerce of the country” (3, p. 94), he prescribed a gradual contraction of the currency in an attempt to restore the value of the depreciated pound in terms of specie. A gradual contraction would be easier to adapt to or even anticipate than a large, sudden contraction.
Ricardo was well aware that nominal wages do not immediately adjust downwards when the price level falls, producing increased unemployment: “We know from repeated experience that the money price of labour never falls till many workmen have been for some time out of work” (9, p. 20).Footnote 11 Thus, he (1, pp. 359–60) declares “when I contemplate the evil consequences which might ensue from a sudden and great reduction of the circulation, as well as from a great addition to it, I cannot but deprecate the facility with which the State has armed the Bank [of England] with so formidable a prerogative.” Also, “I consider any Variation in the Value of the Currency as an Evil, from producing a Variation in the Prices of all Articles” (5, p. 442). Therefore, to secure price level stability and avoid the evil, Ricardo (1, p. 356) prescribed the regulation of paper money by its convertibility into gold.
Ricardo's arguments above appear to contradict Viner's (Reference Viner1937, p. 195) view that he had a “temperamental reluctance to explore the short-run and intermediate phases of economic process” and O'Brien's (Reference O'Brien1975, p. 164; Reference O'Brien2004, p. 196) claim that “Ricardo, focusing as usual on successive periods of long-run equilibriums, denied the damage of deflation, and the stimulating effect of rising prices.” Some also have interpreted Ricardo's views on the forced-saving mechanism as his having minimized its significance, such as Blaug (Reference Blaug1996, p. 159). Such an interpretation would be correct only if intended to convey Ricardo's cautioning against an attempt to promote long-term growth with increases in the quantity of money.
Unlike Ricardo, Thomas Malthus's discussion of the stimulative effect of inflation on output and employment, from increases in the quantity of money, focuses unduly on the short run. Production and employment are increased if more of the new money goes into the hands of producers than fixed-income earners: “on every fresh issue of notes, not only is the quantity of the circulating medium increased, but the distribution of the whole mass is altered. A larger proportion falls into the hands of those who consume and produce, and a smaller proportion into the hands of those who only consume … [working] to increase the national capital [and] to lower the rate of interest” (1811, pp. 96–97). Although he acknowledges the “manifest injustice” to fixed-income earners by inflation as “unjust transfers of property,” he argues that the “evil” would be “more than counterbalanced … by the advantage which the country derives from it” (p. 97).Footnote 12
Malthus believes that “a rise of prices is generally found conjoined with public prosperity; and a fall of prices with national decline” (1811, p. 97). He even suggests that the “increased command of the produce transferred to the industrious classes by the increase of prices” would so increase production as to restore “prices … to their former level” (ibid., pp. 97–98). His argument misses consideration of the subsequent increase in nominal wages, interest, and rental rates that would eliminate the initial advantage to producers. Thus, Ricardo (3, pp. 120–23) criticizes Malthus's argument, particularly the presumption that fixed-income earners do not save, and that inflation necessarily increases overall “capital” or savings. Malthus (Reference Malthus1836, p. 314) subsequently relies on voluntary savings to promote long-term growth.
J. S. Mill, who also draws insights from Hume's monetary analysis in “Of Money” (Mill 3, pp. 511, 564–65), expresses as much indignation at the income and wealth redistribution by unanticipated inflation as Thornton, Bentham, and Ricardo. He regards the “substitution of paper for metallic currency [as] a national gain: any further increase of paper beyond this is but a form of robbery” (3, p. 565). He considers the relief obtained by borrowers whose real interest costs are diminished by inflation an “unjust gain” for their having had a “part of the property of their creditors … gratuitously transferred to them” (ibid.). Mill argues that
the depreciation of the currency, when effected [through the excessive money or credit creation], operates to a certain extent as a forced accumulation. This, indeed, is no palliation of its iniquity. Though A might have spent his property unproductively, B ought not to be permitted to rob him of it because B will expend it on productive labour
(Mill Reference Mill1874, p. 118; italics added).Mill also discusses the short-run, expansionary effect of unanticipated increases in the quantity of money on output, a phenomenon he ascribes to a “delusion”:
The commonest cause of such delusion is some general, or very extensive, rise of prices (whether caused by speculation or by the currency) which persuades all dealers that they are growing rich. And hence, an increase of production really takes place during the progress of depreciation, as long as the existence of depreciation is not suspected; and it is this which gives to the fallacies of the currency school, principally represented by Mr. Atwood, all the little plausibility they possess. But when the delusion vanishes and the truth is disclosed, those whose commodities are relatively in excess must diminish their production or be ruined: and if during the high prices they have built mills and erected machinery, they will be likely to repent at leisure
(Mill Reference Mill1874, p. 67-68; italics added).Footnote 13Clearly employment and probably wage rates would increase during the period of rising prices, while employment would decrease when production decreases as the “delusion” vanishes.
Austrian business cycle theorists may find similarity between their prediction of idle productive capacity at the end of a boom with Mill's reference to builders of mills and erectors of machinery repenting at leisure when the expansionary phase comes to an end. However, for Mill, as for other classicals, the forced accumulation refers first to funds, and not necessarily to capital goods, as the Austrian theory argues, as presented by Hayek (Reference Hayek1931).
The above views of Thornton, Bentham, Ricardo, Malthus, and Mill on the forced-saving mechanism appear to conflict with Laidler's (Reference Laidler1999, p. 89) claim that the forced-saving idea “never played more than a peripheral role in” the English classical tradition. Although most classicals did not advocate use of the mechanism to promote economic growth, they recognized its operation from unanticipated variations in the quantity of money or bank credit.Footnote 14
From the above classical and early neo-classical arguments we can envision a leftward movement along a short-run Phillips curve from unanticipated increases in the price level and a fall in the rate of unemployment, but a rightward (upward) shift of the whole curve as wage rates adjust to the rise of prices and the rate of unemployment returns to its original position. Similarly, their arguments depict a rightward movement along a short-run Phillips curve from an unanticipated decrease in the price level and an increase in the rate of unemployment, but a leftward (downward) shift of the whole curve when nominal wage rates decline in line with the fall of prices and the unemployment rate returns to its original position. Only an increased flow of savings or “capital,” consistent with the natural rate of interest, would increase production and permanently reduce the (natural) rate of unemployment.
IV. THE FORCED-SAVING DOCTRINE IN MODERN MACROECONOMICS
Modern macroeconomists, in effect, employ the classical forced-saving doctrine when they argue the non-neutrality of money and credit on output and employment in the short run because of the existence of contracts or lack of full expectations of inflation or deflation, but money's neutrality in the long run. Friedman (Reference Friedman1968, esp. pp. 7–11) pretty much restates the classical argument, partly in criticism of Phillips's original analysis, and concludes that “there is always a temporary trade-off between inflation and unemployment; there is no permanent trade-off. The temporary trade-off comes not from inflation per se, but from unanticipated inflation” (ibid., p. 11). He comes close to acknowledging the classical origin of his argument when he explains that an increase in monetary growth lowers interest rates and increases aggregate demand, to which producers respond by “increasing output, employees by working longer hours, and the unemployed, by taking jobs now offered at former nominal wages. This much is pretty standard doctrine” (ibid., p. 10). Indeed, Patinkin (Reference Patinkin1969) affirms that the lagged adjustment of factor prices behind product prices – the mechanism that generates the forced saving when prices are rising – was a standard principle taught at Chicago by the likes of Henry Simons and Frank Knight. See also Simons, et al., cited in Laidler (Reference Laidler1993, p. 1094).
Friedman (Reference Friedman1968, p. 7) thanks Wicksell for our being “acquainted with the concept of a ‘natural’ rate of interest and the possibility of a discrepancy between the ‘natural’ and the ‘market’ rate,” the basis upon which Friedman develops his argument of a “natural rate of unemployment” that could not permanently be altered by a central bank's money creation. But, as noted above (Section II), explanation of the “natural rate” of interest comes from Adam Smith and was picked up by Ricardo (see Mill 3, p. 648). Wicksell (Reference Wicksell and Lindhal1935, pp. 79-80) himself also acknowledges Ricardo's precedence in the “cumulative process” analysis of price-level changes generated by deviations of the market rate of interest from the natural rate. That analysis, being a part of the classical forced-saving argument, also was discussed by Alfred Marshall before Wicksell; see Patinkin (Reference Patinkin1965, pp. 631–32) and Ahiakpor (Reference Ahiakpor1999).
Hicks (Reference Hicks1967, pp. 162–63; italics original) also notes that
Inflation does give a stimulus [to production], but the stimulus is greatest when inflation starts – when it starts from a condition that has been non-inflationary. If the inflation continues, people get adjusted to it. But when people are adjusted to it, when they expect rising prices, the mere occurrence of what they had expected is no longer stimulating. Nor can the fade-out be prevented by accelerating the inflation; for acceleration of inflation can be expected too.
Hicks offers this argument as what he thinks the classics would say about the short-run effect of money on output: “some such short-period theory, somewhere at the back of their minds, though [the classicals] preferred not to emphasize it” (1967, p. 162). The classical views presented above argue what Hicks suggests.
Similarly, the new classicals argue that only unanticipated changes in the quantity of money affect real output and employment in the short run (Lucas Reference Lucas1996). Were changes in the quantity of money fully anticipated and acted upon by the general public, nominal wages and rental rates would rise with the expected rate of inflation. There would then be no tendency for real wages to fall and for more workers to be hired for increased production. That is, the short-run Phillips curve would immediately shift rightwards in anticipation of an increased rate of inflation. Also, the savings rate would immediately decline and/or the demand for credit increase to match an anticipated increase in the supply of central bank credit, and thus keep nominal interest rates from falling.
Conversely, an anticipated decrease in central bank money (credit) creation would have no effect on production and employment because nominal wage and rental rates would immediately fall and thus keep their real rates unchanged. This is the equivalent of a leftward shift of the short-run Phillips curve in anticipation of a fall in the rate of inflation or a deflation. There also would be increased savings and/or decreased borrowing to match a central bank's credit contraction, which would keep nominal interest rates from rising. Such emphasis on the non-anticipation of a central bank's monetary actions as the key to the short-run non-neutrality of money reflects the mechanism underlying the classical forced-saving doctrine. Instead, Lucas (Reference Lucas1996, p. 679) thinks that “The discovery of the central role of the distinction between anticipated and unanticipated money shocks resulted from the attempts, on the part of many researchers, to formulate mathematically explicit models that were capable of addressing the issues raised by Hume.” There may only have been a re-discovery with the mathematical models. An interpretation of Hume's (Reference Hume and Rotwein1752, p. 38) argument that an increase in the quantity of money enables “a set of manufacturers or merchants” to hire “more workmen than formerly, who never dream of demanding higher wages, but are glad of employment from such good paymasters” as indicating the non-anticipation of rising prices on the part of the workmen hitherto unemployed would fit that explanation.Footnote 15
However, the version of forced saving argued by Austrian economists does not lend itself readily to a representation by the Phillips curve analysis. They disagree on the process by which forced saving occurs. Rather than a rise in the price level reducing the purchasing power of fixed-income earners, hence their real consumption, the Austrians argue that it is principally the fall in the nominal rate of interest induced by an increased supply of bank credit that enables producers to bid resources away from consumption goods into capital goods production.Footnote 16 The reduced availability of consumption goods and the subsequent rise of their prices is what constitutes the forced saving in the Austrians’ account. Thus, all categories of income earners may bear some forced saving in that version.Footnote 17
The Austrians also do not necessarily argue a subsequent rise in the real wage, interest, and rental rates following the revision of contracts to reflect the increased rate of inflation or the price level, as the process that terminates the initial positive effects of credit or monetary inflation. Their strict designation of producer's goods from consumption goods also assures them a longer impasse in the economy's adjustment process than the classical analysis suggests. Besides, their version appears unable to deal with the case of falling prices from a decreased money or credit creation or from an increased demand for money (currency), as occurred during the Great Depression. Thus, instead of endorsing calls for increased money creation to deal with the ongoing depression, the Austrians advocated against doing anything of the sort.
V. SUMMARY AND CONCLUSION
The forced-saving doctrine was an integral part of classical monetary analysis. It explains the mechanism by which unanticipated increases in the quantity of money or credit increase the rate of production and employment in the short run, but not in the long run. It also explains why unanticipated decreases in the quantity of money or credit decrease the rate of production and employment in the short run, but not in the long run. Significant in that explanation is the lagged adjustment of nominal wage, interest, and rental rates behind changes in product prices. Modern Phillips curve analysis is thus an illustration of the doctrine with respect to wage rates.
Phillips's original diagram illustrates this lagged relationship between wages and the price level through changes in the rate of unemployment, but without relating the phenomenon to changes in the quantity of money or credit. He had little use for the classical quantity theory of money as a theory of the price level, following Keynes (Reference Keynes1936). Fisher's (Reference Fisher1926) analysis correctly has been recognized as explaining the same phenomenon Phillips sought to illustrate. Since Phelps (Reference Phelps1967) and Friedman (Reference Friedman1968), modern macroeconomists essentially have been restating the classical doctrine, but without its explicit acknowledgment. Instead, some analysts increasingly have been associating the doctrine mainly with Austrian business cycle theory, while minimizing its analytical significance in classical macroeconomics or sometimes even denying that the classics employed that doctrine.
A greater appreciation of the classical heritage of the Phillips curve analysis as outlined above may lead to a more general acceptance of the classical view with respect to monetary policy: changes in the quantity of money (currency) should be directed at meeting changes in its demand so as to keep the price level constant,Footnote 18 a view also argued ardently by Fisher (Reference Fisher1913). He considers “the evils of a variable monetary standard [to be] among the most serious economic evils with which civilization has to deal” (1913, p. ix). Following the classical logic, interest rates should be left alone to be determined in the various “capital” or credit markets by the supply and demand for voluntary savings. Policies encouraging increased voluntary savings would then promote increased rates of investment, production, and employment at lower “natural” rates of interest. Monetary inflation does not promote long-term growth because there are forces constantly working to reverse its initial positive effects, the classics taught. There is also room for re-emphasizing in modern macroeconomics the classical moral outrage over the redistribution of real incomes from wage, interest, and rental earners to the benefit profit earners that the short-run inflation-unemployment trade-off entails.
