A paradox of thrift or Keynes's misrepresentation of saving in the classical theory of growth?

Ahiakpor, James C.W.

I. Introduction

Every year thousands of introductory economics students are made to accept as valid one of Keynes's lasting inversions of classical economics, namely the proposition that saving may be a private virtue, but is a public vice. According to Keynes, a community that seeks to increase its rate of saving would end up impoverishing itself and actually saving less, but the community that increases its consumption at the expense of saving would end up being richer and saving more. This proposition, frequently stated in macroeconomics textbooks as the "paradox of thrift," arises mainly from Keynes's definition of saving to include the hoarding of cash, contrary to the classical definition and language of the marketplace, but has received little recognition or criticism as such.(1) Rather than emphasizing Keynes's misrepresentation of the classical theory of income determination and growth in which hoarding constitutes a reduction in saving, most critics have pointed out only that saving may take other forms besides hoarding.(2) Thus, several reviewers of the General Theory [22], in which Keynes fully develops the argument, criticized it mainly for having assigned too much importance to the hoarding of cash, e.g., Pigou [37], Robertson [40], Viner [51], and Hawtrey [11].

Some modern textbook writers employ the loanable-funds view of savings to show the limited significance of Keynes's paradox of thrift proposition, but with little reference to the fact that it derives from a fundamental misrepresentation of saving in the classical theory of growth. See, for example, Baird [3], Boyes and Melvin [6], McConnell and Brue [31], and Case and Fair [7]. In the same tradition, Leijonhufvud [26, 196-97] also argues that saving is more than "an antisocial refusal to spend" and calls the paradox of thrift proposition "one of the most dangerous and harmful confusions ever taught as accepted economic doctrine." On the other hand, some other textbook writers insist that Keynes's proposition is logically sound (it can be demonstrated by the use of algebra or a saving-investment diagram), although also conceding that it fails to accord with experience in the real world, e.g., Henderson and Poole [15, 279-81] and Parkin [33, 224-25].

The principal proponents of the doctrine focus on elaborating Keynes's statement of the obvious (as if it were inconsistent with classical analysis): savings depend on the level of income, and the level of income is determined by the rate of investment, which in turn depends on anticipated consumption demand, e.g., Wonnacott and Wonnacott [52], Samuelson and Nordhaus [47],(3) Heilbroner and Galbraith [14], and Baumol and Blinder [4]. Some claim that the paradox of thrift is a good illustration of the "fallacy of composition" in economics: what is valid at the individual level may not be valid at the group level, e.g., Campagna [8], Bach [2], and Samuelson and Nordhaus [47].(4) Others also suggest that Keynes's argument may be valid only for advanced industrialized economies in certain periods but not for others. But in all these, little reference is made to the classical theory of growth against which Keynes was reacting. Thus, the teaching of Keynes's paradox of thrift proposition continues to thrive in spite of a long history of criticism. And of course, its confusion extends well beyond the level of sophomoric undergraduates, as I document below.

In this article, I restate the classical savings theory of growth to provide a basis for demonstrating that Keynes misrepresented the classical concept of saving to include the hoarding of cash, a mistake which led to his erroneous conclusions about the classical theory of growth. I use Keynes's arguments mainly in the General Theory to restate the paradox of thrift proposition,(5) and I rely on direct quotations from the classics and Marshall to establish Keynes's misrepresentation of their argument. I conclude that there is no such thing as a paradox of thrift other than Keynes's incorrect substitution of "hoarding" for "saving" in the classical theory of income determination. This recognition may help end the annual ritual of teaching students a fundamental confusion from which they later struggle to escape in courses such as development economics or finance. The discussion also shows that there is much of significance yet to be gained by paying careful attention to what the classics said, especially in assessing Keynes's claims against them. Even among the classics, as Samuelson [45, 1430] has observed, "their quarrels lasted because often they were quarrels over misunderstandings and definitions."

II. The Classical Theory of Growth

The classical theory of growth against which Keynes proposed his paradox of thrift argument simply states that economic growth is determined by the rate of saving or capital accumulation. The greater the amount saved out of income, the more capital goods, land, and labour services can be bought or hired for production. As Smith [50, 1:358-59] clearly states the argument,

Capitals are increased by parsimony,... Whatever a person saves from his revenue he adds to his capital, and either employs it himself in maintaining an additional number of productive hands, or enables some other person to do so, by lending it to him for an interest, that is, for a share of the profits. As the capital of an individual can be increased only by what he saves from his annual revenue or his annual gains, so the capital of society, which is the same with that of all the individuals who compose it, can be increased only in the same manner.(6)

Earlier, Smith argues that "Every increase or diminution of capital [i.e., saving] . . . naturally tends to increase or diminish the real quantity of industry, the number of productive hands, and consequently the exchangeable value of the annual produce of the land and labour of the country, the real wealth and revenue of all its inhabitants."

The classics did not presume that every saver was also the ultimate investor or purchaser of investment goods. The classics wrote about an economy in which there was money (cash) as well as financial intermediation by banks and brokers.(7) Also note that the classics used the term "capital" and "savings" interchangeably when discussing the actions of savers, e.g., Smith [50, 1:372-75], Ricardo [38, 1:363, 3:89-94],(8) and Mill [32, 3: ch. 23]. This explains why one typically does not see "saving" cited in the indexes of their works, and when cited, one finds "see capital," e.g., Mill [32, 3:1164].(9) Also see Marshall's Principles. On Keynes's misinterpretation of the classical usage of the term "capital" to mean "funds" or "savings", see Ahiakpor [1](10); also see Hartfield [10] on the widespread misuse of the term "capital" in modern economics.

Thus, saving or capital in the classical theory of growth refers to nonconsumed income from which revenue is derived, in terms of interest or profits, e.g., Smith [50, 1: Bk 2, ch. 1]; also see Marshall [29, 66]. Non-consumed income hoarded in cash does not constitute saving or capital since it does not earn interest or profits. Malthus, for example, clarifies the place of hoarding well when he argues: "No political economist of the present day can by saving mean mere hoarding." Quoted in Blaug [5, 166]. Rather, hoarding may be regarded as a form of 'spending' income, namely, the purchase of a non-income generating asset, cash.(11) As Smith [50, 1: 458] so well makes the point, "Money, like wine, must always be scarce with those who have neither wherewithal to buy it, nor credit to borrow it." [Emphasis added.] Thus, for those not engaged in the direct purchase of investment goods, there are three forms of spending income (Y), according to the classical argument: consumption (C), purchasing financial assets, i.e., saving ([Delta][FA.sup.d]), and buying or holding cash (pY).(12) Therefore, saving is equal to income less consumption and the proportion (p) of income households desire to hold in cash, i.e., [S.sub.] = [Y.sub.t] - [C.sub.t] - p[Y.sub.t] = [Delta][FA.sup.d].(13)

Note that pY is not the stock of accumulated cash but a flow within each period. Thus, if p = .1 and Y = $1,000, $100 of a period's income would be the amount desired to be held as cash while $900 would be divided between consumption spending and the purchase of financial assets or current saving. But should p instead be .4, $400 of current income would be the desired cash holding while $600 would be devoted to consumption and saving. Whether the desire to hold a greater proportion of current income in cash is realized or not depends on the existing stock of cash made available by the monetary authorities, less the quantity already being hoarded.

It is tempting to write the saving equation as S = Y - C - [Delta]H, as some commentators have suggested. But that formulation blurs recognition of the choice variable, p, in affecting the saving decision. Moreover, for the community as a whole, [Delta]H = 0 unless the monetary authorities change H. But the aggregation of [p.sub.i][Y.sub.i] over individuals does not amount to zero. Therefore, households' flow-demand for cash appears better represented by pY in the saving equation rather than [Delta]H, just as kY represents the total demand for money (cash) in the Cambridge equation, H = kY.(14)

Corresponding to the loaned out savings or accumulated capitals of households ([Delta]FA), is capital acquired by ultimate investors or producers: machinery, equipment, buildings, materials for further processing, produced goods yet to be sold (inventories), and, in a rather small proportion, cash needed to assist transactions of a business enterprise [50, 1:313; 30, esp. 46-47]. These constituents of capital are frequently classified in the classical literature as 'fixed' or 'circulating' depending on whether they yield "a revenue or profit without . . . changing masters" [50, 1:297; 29, 63]. But the existence of savings by households is logically prior to the accumulation of capital in the hands of producers.(15)

It is by abstaining from consuming all of their income and purchasing new financial assets ([Delta]FA), including bank deposits, bonds or stocks, that savers make it possible for investors (issuers of IOUs or financial assets) to borrow "capital" or the amounts saved, assuming no change in the economic or excess reserve ratios of banks.(16) That is, in value terms, [S.sub.t] = [Y.sub.t] - [C.sub.t] - p[Y.sub.t] = [Delta][FA.sub.t] = [I.sub.t] (ignoring the amount of cash held by banks as reserves). And the expenditures of consumers and investors add up to the total demand for goods and services currently produced in the economy. This is why it is impossible in a closed economy, and without an inflation of central bank credit or a reduction in the economic (or excess) reserves of commercial banks, for real investment in any period to exceed savings. And when such inflation of bank credit makes investment spending exceed private savings or the supply of "capital", this must be accompanied by a rise in the general price level, a phenomenon called 'forced saving' in the classical literature.(17)

The notion of savings being the source of funds for investment spending is what Mill [32, 2:70] states in the 'Principles' as follows:

The word saving does not imply that what is saved is not consumed, nor even necessarily that its consumption is deferred; but only that, if consumed immediately, it is not consumed by the person who saves it. If merely laid by for future use, it is said to be hoarded; and while hoarded, is not consumed at all. But if employed as capital, it is all consumed; though not by the capitalist.(18)

Perhaps "spent" would have been a clearer term than "consumed" in reference to savings "employed as capital." Ricardo [38, 2:449] makes the same point in criticism of Malthus, whom he accused of having failed to "remember that to save is to spend, as surely, as what [Malthus] calls spending." But Mill's use of the term "consume" interchangeably with "spend" is consistent with that of Smith's, of which Edwin Cannan also comments as lacking clarity.(19)

If income earners purchased a greater proportion of currently produced goods for immediate consumption, there must be a smaller amount left for producers (firms) to buy. Furthermore, increased consumption by income earners must also reduce the flow of private sector credit or loanable savings, assuming no injection of new credit by the banking system.(20) If investors maintain the same level of demand for savings (or credit) while there is an increase in consumption spending or a reduced flow of savings, the rate of interest must rise and the quantity of savings actually borrowed fall. The reverse is also true. A reduced purchase of goods for consumption along with an increase in saving lowers the rate of interest and increases the quantity of savings borrowed. The larger volume of borrowed savings matches the larger quantity of currently produced goods income earners have not purchased for immediate consumption, but which now could be purchased by firms for further transformation.

Note that there are numerous products that could be bought for household use or by firms, including fruits and vegetables, flour, fish, cement, automobiles, electricity, gasoline, building materials, and textiles. Thus, while some goods may fit a strict designation as consumption or investment goods, failure to recognize the possibility of substitution in alternative uses can hinder sound analysis.

We can derive the familiar (classical) relations between the rate of saving, the level of interest rates, investment spending and income growth from the above analysis. Given the demand for investment, a fall in the flow of savings causes a rise in interest rates, a fall in current investment, and a fall in future growth of output and income (Appendix Figure 1a). But a fall in interest rates due to increased saving, makes increased investment spending possible and promotes the growth of future output and income (Appendix Figure 1b). On the other hand, a fall in investment demand, given the rate of saving, would cause the rate of interest to fall, along with a slower growth of future output and income (Appendix Figure 2a).(21) However, a reduction in investment demand which coincides with a reduction in consumption demand would cause the rate of interest to fall by a greater amount, and prevent investment spending and the growth of output and income from falling by as much as in the previous case. An increased demand for investment, given the schedule of savings, would cause the rate of interest as well as output and income growth to rise (Appendix Figure 2b). But a rise in consumption demand (or a fall in the supply of savings), while the demand for investment increases, would raise the rate of interest by much more and slow the growth of output and income.

In the language of modern macroeconomics, "ex-ante savings" refers to the saving schedule and "ex-ante investment" refers to the investment-demand schedule, and the intersection of these two schedules depicts "ex-post saving and investment." Also note that the above analysis assumes an unchanged demand for money (cash). For example, an increase in the demand for money or increased hoarding would lower the price level and may also reduce the growth of output and employment, at least in the short run. I explain the consequences of variations in the demand for money on savings, investment, and interest rates in greater detail in section IV below. Also note that the above conclusions do not depend on the assumption of "full employment," however defined, and saving and investment spending refer to flows, not stocks.(22)

David Hume [17, 177] perhaps most eloquently states the classical analysis of equilibrium interest rate determination by the interplay of the supply and demand for savings (or capital) rather than money:

High interest arises from three circumstances: a great demand for borrowing; little riches to supply that demand; and great profits arising from commerce: and the circumstances are a clear proof of the small advance of commerce, and industry, not of the scarcity of gold and silver [i.e., money]. . . . Low interest, on the other hand, proceeds from the three opposite circumstances: a small demand for borrowing; great riches to supply that demand; and small profits arising from commerce." [Emphasis in original.]

Indeed, Smith [50, 1:376] finds Hume's explanation of interest rate determination, particularly the irrelevance of the quantity of money in explaining low interest rates, so clear as to be unnecessary to elaborate any further.(23)

The classical explanation of the role of saving in interest rate determination also underlies Marshall's [29, 581-82] argument:

Everyone is aware that the accumulation of wealth is held in check, and the rate of interest so far sustained, by the preference which the great mass of humanity have for present over deferred gratifications, or, in other words, by their unwillingness to "wait." And indeed the true work of economic analysis in this respect is, not to emphasize this familiar truth, but to point out how much more numerous are the exceptions to this general preference than would appear at first sight.(24)

But Keynes [22, 242] cites this statement as representing the classical view of the role of saving in the determination of interest rates with which he disagrees. Furthermore, he insists, it is the impatience of income earners to consume rather than save which is conducive to capital accumulation. I explain the basis of Keynes's disagreement below.

III. Keynes's Paradox of Thrift Proposition

In rejecting the classical savings theory of capital formation and growth, Keynes defines saving simply as "the excess of income over what is spent on consumption" [22, 74](25) without making any necessary link between saving and the purchase of interest- or profit (dividend)-earning assets, i.e., S = Y - C = pY + [Delta]FA. Thus, according to him, saving withholds demand from currently produced goods, resulting in the accumulation of inventories. This dampens the prospects of future profits (Keynes's "marginal efficiency of capital") for investors who then cut back on investment spending and reduce their demand for labor [22, 210-12]. On the other hand, increased consumption raises the demand for currently produced goods, increases the prospects of future profits, and encourages further investment spending and employment. Thus, he argues, "the level of output and employment depends, not on the capacity to produce or on the pre-existing level of incomes, but on the current decisions to produce which depend in turn on current decisions to invest and on present expectations of current and prospective consumption" [22, xxxiii].

Keynes also argues that "the growth of capital depends not at all on a low propensity to consume [i.e., a high propensity to save] but is, on the contrary, held back by it; and only in conditions of full employment is a low propensity to consume conducive to the growth of capital" [22, 372-73; 21, 1:173]. But note that by "capital" Keynes means capital goods, not loaned out savings, according to the classical definition. Thus, by focusing on consumption spending as the principal determinant of investment, hence aggregate demand, Keynes finds no positive role in his income determination process for savings.

Furthermore, according to Keynes, saving is a withdrawal from the expenditure stream without any clear prospects of increased spending in the future. To believe otherwise, he argues, is to suppose fallaciously the existence of a "nexus which unites decisions to abstain from present consumption with decisions to provide for future consumption" [22, 21]. Keynes [22, 210] elaborates this argument thus:

An act of individual saving means - so to speak - a decision not to have dinner to-day. But it does not necessitate a decision to have dinner or to buy a pair of boots a week hence or a year hence or to consume any specified thing at any specified date. Thus it depresses the business of preparing to-days's dinner without stimulating the business of making ready for some future act of consumption. It is not a substitution of future consumption-demand for present consumption-demand, it is a net diminution of such demand. [Emphasis in original.](26)

Only if savers gave producers notice of the specific things they plan to purchase as well as the dates they plan to spend - quite an impossible supposition - would their act of saving not significantly imperil production, Keynes [22, 210-11] also argues.(27) He does not recognize the classical definition of saving, such that it constitutes spending by someone else other than the saver, nor that increased saving lowers the rate of interest. Instead he argues that, because saving withdraws from consumption spending and discourages investment spending, the attempt by all individuals to save more must necessarily be frustrated by a fall in the level of total income [22, 84].

Keynes does recognize a connection between saving and investment when he argues that "no one can save without acquiring an asset, whether it be cash or a debt or capital-goods" [22, 81; emphasis added].(28) And it is the acquisition of capital goods that he usually defines as investment [22, 62, 75], a somewhat narrower concept than the classical definition of investors' capital explained above. This use of the term "investment", which is now part of the language of modern economics, also fails to recognize the fact that households consider their purchases of financial assets, including securities, as "investment."(29) But note that this comment does not invite "double-counting and theoretical confusion" as a previous anonymous reader fears, if we recognize that saving conceived as investment in new financial assets ([Delta]FA) by households, makes possible the purchase of newly produced investment or capital goods. And there is no increased saving or investment unless [Delta]FA [greater than] 0.

Keynes also acknowledges that the proportion of current output which is invested must be equal to the proportion saved. But he develops this link from an identity: "in the aggregate the excess of income over consumption, which we call saving, cannot differ from the addition to capital equipment which we call investment," because "the whole of . . . output must obviously have been sold either to a consumer or to another entrepreneur" [22, 64]. Thus, there is no necessary causal sequence of household savings transferring purchasing power to entrepreneurs who then undertake investment spending.

In developing the paradox of thrift argument, Keynes also firmly separates saving and investment demand from interest rate determination. Rather, he explains the determination of interest rates from the supply and demand for money (cash) or liquidity.(30) Thus, he argues that an increase in the rate of interest (which could have arisen from an increased investment demand, according to the classical theory),(31) has "the effect of reducing the amount actually saved" [22, 110], instead of increasing it. This would happen, according to him, because although the increase in the rate of interest may reduce consumption, it would also reduce the rate of investment upon which the growth of income and saving depends:

The rise in the rate of interest might induce us to save more, if our incomes were unchanged. But if the higher rate of interest retards investment, our income will not, and cannot, be changed. They must necessarily fall, until the declining capacity to save has sufficiently offset the stimulus to save given by the higher rate of interest. The more virtuous we are, the more determinedly thrifty, the more obstinately orthodox in our national and personal finance, the more our incomes will have to fall when interest rises relatively to the marginal efficiency of capital. Obstinacy can bring only a penalty and no reward. For the result is inevitable. [22, 111; emphasis in original.]

However, note that Keynes in the above argument does not indicate the source of the increase in the rate of interest, although he includes cash among the assets acquired in the act of saving [22, 81], and also argues that an increased demand for cash raises the rate of interest [22, 166-74]. But as explained above, an increase in the rate of interest due to an increase in investment demand, may increase the flow of savings out of current income and make increased investment possible (Appendix Figure 2b). The increased investment would then increase future income? Thus, Keynes in the above argument misinterprets what is, in effect, a comparative statics analysis which also can be illustrated by a model he himself [22, 178] summarizes in the General Theory, but could not attribute to the classics.

It is by the above misinterpretations of classical analysis coupled with his peculiar definition of savings to include hoarding that Keynes arrives at his proposition of the "paradox of poverty in the midst of plenty" [22, 30]. A poor community "prone to consume by far the greater part of its output," has a greater prospect of growth because "a very modest measure of investment will be sufficient to provide full employment" [22, 31] - but failing to note that savings provide the funds for investment. Keynes thus turns the classical savings theory of growth on its head by arguing that "capital is brought into existence not by the propensity to save but in response to the demand resulting from actual and prospective consumption" [22, 368]. His argument is in direct contradiction of Smith's: "Parsimony, and not industry, is the immediate cause of the increase of capital. Industry, indeed, provides the subject which parsimony accumulates. But whatever industry might acquire, if parsimony did not save and store up, the capital would never be the greater" [50, 1:359].

IV. Recognizing Keynes's Difficulties with Classical Growth Theory

If one defines saving as Keynes did, simply as the residual of income over consumption spending and thus include the hoarding of cash, i.e., S = Y - C = pY + [Delta]FA, then it is possible to attribute a depressing effect of increased saving on aggregate demand and income growth. Indeed, Blaug's [5, 163] inference that" 'intended saving' in the modern sense is equivalent to classical saving plus hoarding" helps in appreciating the point. One also has to assume that increased saving means increased hoarding (or increased demand for cash), just as Keynes did.(33) Thus, one could show consistency between Keynes's argument under these assumptions with short-run classical analysis using the Quantity Theory of money, summarized in the Cambridge equation: H = kY = kPy, e.g., Marshall [30, esp. 43-48], Keynes [20, esp. 81-88], and Pigou [37, 120-21]. Given the quantity of cash (H) supplied by the monetary authorities, an increase in k must cause a decrease in Y (from a decrease in P or y, or a combination of both) in the short run. Keynes [22, 84-85], in fact, reasons along these lines, although he does not directly invoke the quantity theory. The rate of interest rises, according to Keynes's reasoning, but need not rise in the classical analysis.

According to the classics, income may be paid in cash and used to purchase consumption goods and services, financial assets (saving), and partly held as cash, e.g., [50, 1:359, 458; 32, 2:70]; also see Marshall [30, 46]. If people desire to hold a larger fraction of their income in cash, they may do so by reducing either consumption spending or their demand for financial assets. A reduction in the demand for consumption goods would reduce the price level - a weighted average of the prices of consumption and producers' or capital goods - even if the demand for producers' goods remains unchanged. Total spending, output and employment may fall in the short run, but the rate of interest would remain unchanged, given an unchanged supply and demand for financial assets.

Only if the increased demand for money (cash) is at the expense of financial assets, would the rate of interest initially rise (an excess supply of financial assets). Again, output and employment would fall in the short run, even as consumption spending remains unchanged. The reduced investment spending would also create an excess supply of producers' goods, reduce their prices and therefore the general price level. However, a larger quantity of investment goods could later be purchased by entrepreneurs (when nominal wages fall) with the reduced flow of savings because of the fall in prices, and the economy then restored to its long-run growth path. The rate of interest also would fall to its initial level as the price level falls. This is the real balance effect mechanism Patinkin [34, 244-51] well explains (but substitute financial assets for bonds in his analysis). Also see, e.g., Hawtrey [11, 441]. Again, the short-run negative consequences of an increased demand for money are similar to those of Keynes's paradox of thrift argument. But an increased demand for money at the expense of financial assets constitutes a decrease in saving, not an increase as Keynes reasons.

The classics also were aware of the incidence of hoarding, which they correctly attributed to a disturbance in the state of confidence, e.g., Ricardo [38, 3:172, 6:289, 300-301]. They also noted that its consequence is to reduce the level of prices and economic activity; also see Marshall [30, 226]. It is because of these consequences that Ricardo [38, 9:167], for example, calls Cobbett "a mischievous scoundrel" for having recommended that people hoard money (gold), which Cobbett "knows will increase the value of money still more." Thus, Blaug's [5, 183] claim that the "saving-is-spending theorem encouraged the classical thinkers to ignore 'hoarding' " is an overstatement. But what is pertinent to the point of this article is that the classics did not confuse saving with the hoarding of cash as we find in Keynes's argument.(34)

Keynes's definition of saving to include hoarding also serves as a basis for his rejection of the classical proposition that the rate of interest is the reward for waiting or abstinence. He argues that "if a man hoards his savings in cash, he earns no interest, though he saves just as much as before" [22, 167] and insists that "the rate of interest cannot be a return to saving or waiting as such . . . On the contrary, . . . [it] is the reward for parting with liquidity [i.e., cash] for a specified period" [22, 166-67].(35) Thus, according to his argument, increased saving plays a double negative role: raising the rate of interest and reducing aggregate demand.

But had Keynes correctly recognized increased saving to mean increased demand for financial assets (as he did in the 1939 Times article cited in note 30 above), he could have recognized the proper connection between saving, borrowing (supply of financial assets) or investment demand, and interest rates, as the classics argued. He also could have recognized the validity of Marshall's restatement of the classical argument that high interest rates and limited growth are due to insufficient saving. Keynes [22, 242] might not, instead, have retorted:

The world after several millennia of steady individual saving, is so poor as it is in accumulated capital-assets, is to be explained . . . neither by the improvident propensities of mankind, not even by the destruction of war, but by the high liquidity-premiums formerly attaching to the ownership of land and now attaching to money [cash].

Keynes's definition of saving to include hoarding also may explain his rejection of Marshall's clarification that saving is just another form of spending, which he [22, 19] quoted:

The whole of a man's income is expended in the purchase of services and of commodities. It is indeed commonly said that a man spends some portion of his income and saves another. But it is a familiar economic axiom that a man purchases labour and commodities with that portion of his income which he saves just as much as he does with that he is said to spend. He is said to spend when he seeks to obtain present enjoyment from the services and commodities which he purchases. He it said to save when he causes the labour and the commodities which he purchases to be devoted to the production of wealth from which he expects to derive the means of enjoyment in the future.

In rejecting the above explanation, along with the Say-Ricardo 'Law of Markets', Keynes [22, 19-20] also argues that little role is assigned to money, and "Contemporary thought is still deeply steeped in the notion that if people do not spend their money in one way they will spend it in another." In a monetary economy, he argues, some income earners may save by hoarding cash instead of purchasing investment goods or hiring labor. But note Keynes's incorrect substitution of "money" for "income" in the quote he criticizes. The classics explained that income has to be spent one way or the other, on consumption goods, financial assets (savings) or investment goods directly, or purchasing the service flow of cash. Thus, Keynes's criticism appears to arise from his misinterpretation of the classical argument.

Keynes's inclusion of cash among assets acquired in the act of saving also explains his siding with Malthus against Ricardo in their debate over the possibility of a general glut, and extending the under-consumption theory of Malthus and his followers. Keynes finds quite puzzling Ricardo's success in arguing the invalidity of Malthus's concern that saving or parsimony may cause a reduction in the progress of wealth.(36) To him, "The completeness of the Ricardian victory is something of a curiosity and a mystery" [22, 32]. He makes little meaning of Ricardo's arguments, such as "Productions are always bought by productions or services; money is only the medium by which the exchange is effected. Hence the increased production being always accompanied by a correspondingly increased ability to get and consume, there is no possibility of Overproduction." Quoted in Keynes [22, 369; my emphasis]. He thinks Malthus, along with Karl Marx, Silvio Gesell, J. A. Hobson, and A. E Mummery, ought to have won the debate instead of Ricardo and his followers whose argument Keynes frequently misrepresented as having claimed that "supply creates its own demand" [22, 25; 23, 223], instead of "supply creates demand."(37)

However, in his admiration of the views of Malthus and his followers, Keynes takes their objections to Ricardo's statement of the impossibility of a general glut much further than they did. They recognized the need for saving to promote investment, but were worried about too much saving. Thus, Malthus in arguing with Ricardo noted, for example:

It is not, of course, meant to be stated that parsimony, or even a temporary diminution of consumption, is not often in the highest degree useful, and sometimes absolutely necessary to the progress of wealth . . . when the capital of a country is deficient, compared with the demand for its products, a temporary economy of consumption is required, in order to provide that supply of capital which can alone furnish the means of an increased consumption in future. [Quoted in Ricardo [38, 2:325-26]](38)

Keynes, on the other hand, considers the under-consumption theorists to have conceded too much, particularly by admitting that increased saving would reduce the rate of interest. Noting that Hobson and Mummery "were aware that interest was nothing whatever except payment for the use of money," Keynes considers their arguments to have "failed of completeness, essentially on account of their having no independent theory of interest" [22, 369-70]. He believes they should have constructed such a theory of interest from the supply and demand for money (cash), a mercantilist doctrine whose restatement Keynes [24, 250] claims to have "hit on" after failing to recognize a valid theory of interest in classical analysis. But the critics of Ricardo's argument appear to have learned enough corrections of the money (cash) theory of interest from Hume, Smith, Ricardo and Mill not to have repeated it.

V. Some Implications and Conclusions

Keynes proposed his paradox of thrift argument in direct contradiction of the classics who argued that increased saving is the foundation of increased capital accumulation and economic growth. Thus, the validity of his argument must be judged with its target in mind. But in advancing his criticism, Keynes defined saving to include the hoarding of cash, contrary to the classical definition which limits saving to the purchase of interest- or profit-earning assets. The classical definition is also more consistent with language of the marketplace than Keynes's. People typically associate saving with the act of 'investing' in financial assets (bank accounts, mutual funds, stocks, bonds, or shares in a credit union), not the hoarding of cash. Thus, it may help to end the propagation of Keynes's paradox of thrift as a legitimate criticism of the classical savings theory of growth if this fundamental difference in definitions is explicitly noted.

Of course, several of Keynes's contemporaries complained about his misuse of terms. For example, Pigou [37, 119-20] refers to Keynes's "loose and inconsistent use of terms", while Robertson [41, 429] describes Keynes's definition of saving as "paradoxical," and together with Keynes's terms such as the supply of "finance" and "liquidity" he calls "verbal monstrosities" [42, 19]. Viner [51, 147-48] also notes that in Keynes's General Theory, "no old term for an old concept is used when a new one can be coined, and if old terms are used new meanings are generally assigned to them. . . . The old-fashioned economist must, therefore, struggle not only with new ideas and new methods of manipulating them, but also with a new language." Johnson [18] and Hazlitt [13, ch. 24] are among the modern equivalents of such criticism. Even Hicks, among the most effective propagators of Keynes's ideas, finds his "definitions of saving and investment" to be "new" [16, 239].

Keynes [24, 249-50] himself acknowledges the judgment of his critics that "the clue to the peculiarity of [his new] doctrine is to be found" in his definitions of "Income, Saving, Investment and other such terms." Yet, to his mind, he did not differ "substantially . . . from Marshall or any of the older economists" [ibid.]. Keynes may have believed this because his critics failed to point out exactly how inconsistent were his definitions with those of the classics. And since it is possible to derive almost the same conclusions from an increased demand for money, employing the classical Quantity Theory, as from an increase in saving, employing Keynes's definition of saving to include hoarding, it was quite an ineffective way to have attempted to correct him by pointing out that saving may take other forms besides hoarding. Moreover, such criticism concedes an unwarranted validity of Keynes's definition of saving, especially as he employs that definition to criticize the classical savings theory of growth.

The fact that several of Keynes's critics were anxious not to be too closely identified with what Keynes had labelled as "old fashioned orthodoxy," e.g., Hawtrey [11, 437] or "classicality," e.g., Robertson [42, 12] may explain their failure to restate definitions directly from the works of Smith, Ricardo, and Mill, or sometimes even Marshall. For example, Robertson [41, 435] considers it "most generous" of Keynes not to have regarded him as a classical economist! Some critics also probably took it for granted that Keynes, the economics professor, author of several books in economics, and editor of the Economic Journal, ought correctly to have understood the meaning of terms as employed in the classical literature, and were simply puzzled at his arguments, e.g., Robertson [41, 431]. In any case, it appears the critics felt certain that they could show the error of Keynes's arguments on the basis of his own definitions, just as several modern critics still do.

The failure of Keynes's contemporaries to correct his definition of saving also appears to have assured its survival in modern macroeconomics, along with its confusing implications. As Leijonhufvud [26, 175n] correctly notes, it is "one of the more common muddles of [modern] macroeconomics that we do not know what an 'increase in non-consumption' means for the state of excess demand in [other] markets." But the classical definition of saving is not prone to such a muddle, and facilitates easy recognition of the logic of the savings theory of growth.

It is now much harder to convince readers that Keynes could have arrived at his disagreements with the classics from a misunderstanding or misinterpretation. Rather, many tend to believe the view that Keynes studied economics at Cambridge "under the then-doyen of British economics, Alfred Marshall, as well as under . . . Arthur Pigou" [35, xix], and thus have difficulty accepting textual evidence suggesting his misinterpretation of the classics. As a commentator on an earlier draft wonders: "Keynes, a student of Marshall, fails to appreciate Marshall's restatement of the classical theory of interest. Possible, but plausible? I need more of a case . . ." Obviously, Skidelsky's [49, 166] account that Keynes's "total professional training [under Marshall] came to little more than eight weeks. All the rest was learnt on the job" is much less known. Thus, many would rather resort to seeking explanations of Keynes's difficulties with the classics in the complexities of models.(39)

An excellent example of such puzzling efforts at reconstructing Keynes's arguments is Leijonhufvud's [26, ch. 7] "Z-theory" in which he rejects Keynes's money or liquidity-preference theory of interest, declared by Keynes [24, 249, 250] himself to have been his novel contribution to economics. Leijonhufvud also rejects the paradox of thrift proposition but retains Keynes's definition of saving to include hoarding, e.g., [26, 156, n. 39], and appeals to a failure of interest rates properly to coordinate intertemporal information as being the essence of what Keynes was aiming to say [27, 84-85]. At the same time, Leijonhufvud [26, 171, n. 58] also declares: "As far as I can understand, Robertson was consistently right on every aspect of the interest rate controversy between himself and Keynes and Keynes' Cambridge followers."

Indeed, Robertson was the one who probably debated Keynes the most on the latter's use of terms (definitions), including saving and investment as against investment-demand in the theory of interest, suggesting that Keynes failed to recognize the fact that "English words in -ing sometimes denote a process (requiring translation into Latin by an infinitive or gerund) and sometimes denote the object to which the process has been applied (requiring translation by a neuter past participle passive)" [42, 15]. Robertson, e.g., [42, 28-30] also scorns Keynes's use of that effective debating device, the "full-employment" assumption attributed to classical analysis, as being irrelevant to their dispute. Yet Leijonhufvud employs the "full-employment" assumption to characterize classical analysis and does not fault Keynes for his definition of saving. Thus, the Keynesian confusion persists.

In a similar vein, a previous anonymous reader invokes the full-employment assumption to divert attention from recognition of the crucial role of Keynes's definition of saving in having led to his erroneous argument, and sees little merit in my clarification:

. . . the classical model starts with the assumption of full employment. For growth to occur, saving is an absolute necessity . . . [And] if one starts with full employment with pure competition, etc., everyone is a classical economist . . . So in the end, I do not see much to be gained from resurrecting old arguments, and revisiting old territory.

Such attitude towards reexamination of the literature sustains the persistence of what the reader yet calls "basically a puzzle"!

Sometimes the paradox of thrift proposition has been used to suggest that governments should promote increased consumption rather than saving, if they wish to promote economic growth.(40) For example, Samuelson and Nordhaus [46, 171-74] draw implications from the paradox of thrift proposition to warn against "President Reagan's tax cuts put forth as a means of promoting saving" in the U.S. However, recognizing the basis of Keynes's paradox of thrift argument, hence its error, we may develop a greater consensus on the virtues of saving at the community level, just as Keynes [22, 361 n] quotes Adam Smith out of context to have argued: "What is prudence in the conduct of every private family can scarce be folly in that of a great Kingdom."(41) As several critics already have pointed out, recommending increased consumption rather than saving as a means of promoting increased economic growth is fundamentally an erroneous idea, whether the economy is at "less than full employment" or not.

Thus, employing a consistent definition of saving, namely the purchase of interest- or profit (or dividend)-earning assets, not the hoarding of cash, we find that there is no such thing as a "paradox of thrift." Increased saving, does not reduce the level of aggregate demand nor the growth of output and employment. (This fact also has implications for IS-LM analysis, taken up elsewhere.) Indeed, few would dispute the fact that countries with high savings-income ratios grow faster than those with low saving ratios. But an increase in the desire to hoard cash does reduce aggregate demand and the growth of output, employment, and income, at least in the short run. And that result is no paradox.

It is well known that definitions drive the conclusions reached from models or arguments. Thus, we need to pay careful attention to the meaning attached to definitions in economic analysis, especially in evaluating Keynes's disputes with the classics. Manipulating algebraic equations or graphs is often no good substitute. It also appears that far too many analysts have attempted to clarify Keynes's arguments for easy understanding rather than evaluating their validity against classical propositions. Thus, for example, Paul Samuelson could have helped to end propagation of the paradox thrift doctrine had he explained it as having arisen from Keynes's misinterpretation or misrepresentation of the classical definition of saving instead of illustrating its seeming correctness with the famous savings-investment diagram.

It is one thing for Keynes to have misinterpreted or misrepresented the classical definition of saving, and to have developed his critique of their savings theory of growth as a result. And we may yet find some weak points in the classical theory, for example, not being clear how quickly prices and interest rates would adjust. But it seems quite inexcusable for modern analysts to continue to present the paradox of thrift proposition as if it were a valid criticism of the classical theory of income determination and growth.

1. Leijonhufvud [25, esp. 29-30, 264] comes close to recognizing Keynes's misinterpretation of the classical definition of saving, but rather defends him as restating "the most ancient proposition of monetary theory" [25, 29n]. He also incorrectly argues that Keynes was merely misapplying the classical theories of interest and income determination which dealt with a barter economy [25, 28-29].

2. Among the early critics are Pigou, who argues that "Saving may take the form of hoarding" [37, 125; emphasis in original), but "a man decides to save . . . not in order to accumulate money in a stocking or in a bank balance, but in order presently to buy a house or a motor car, or to invest in interest-bearing securities" [37, 126], and Hawtrey, who argues that "savings are directed to . . . two alternative uses, idle balances and active investment" [11, 441]. Baird [3, 133] is an exception when he argues that "The paradox of thrift is based on the assumption that the only alternative to consumption is the hoarding of money."

3. Remarkably, Samuelson's Economics, which has been one of the most effective propagators of the paradox of thrift argument around the world, no longer discusses it in the fourteenth edition. Instead, the text [48, 30-31] now restates the classical emphasis on saving to promote economic growth, although without identifying the argument as such: "If people are willing to save - to abstain from present consumption and wait for future consumption - then society can devote resources to new capital goods. . . . forgoing current consumption in favor of investment adds to future production possibilities." However, Nordhaus says they will re-insert the paradox of thrift proposition if they hear that it is missed! [Private communication, April 1992.]

4. Keynes [22, 84-85] himself illustrates the argument in terms of the ability of individuals to change the quantity of cash they hold but the impossibility of a community to change it independently of the quantity supplied by the "banking system."

5. There are hints of Keynes's paradox of thrift proposition in the Treatise [21], although Keynes himself minimizes the consistency between arguments there and in the General Theory, e.g., [22, 77-79]. Robertson [39] attempts rather ineffectively to dispute the necessary connection Keynes makes between saving and hoarding in the Treatise, arguing that "Saving does not necessarily involve Hoarding. But neither does Hoarding necessarily involve Saving" [39, 400].

6. This is a fundamental classical proposition which Malthus, after whose arguments Keynes believed he was fashioning his critique of classical growth theory, also accepted. Malthus [28, 314] also used the principle to criticize Lord Lauderdale: "It is certainly true that no permanent and continued increase of wealth can take place without a continued increase of capital; and I cannot agree with Lord Lauderdale in thinking that this increase can be effected in any other way than by saving from the stock which might have been destined for immediate consumption, and adding it to that which is to yield a profit; or in other words, by the conversion of revenue into capital."

7. Indeed, because it is income which is saved through the medium of money, the classics were at great pains to point out that money (cash) serves mainly a conveyance function, transferring the purchasing power of income saved into the hands of investors, e.g., Smith [50, 1:374], Ricardo [38, 3:93], Mill [32, 2:55]; also see Pigou [36, 121]. Proof of the correctness of the classical argument can be found in the fact that annual income as well as the volume of savings (financial assets) at any point in time is typically several times greater than the quantity of money (cash). Contrast this clarification with, e.g., Chick [9, 184-85].

8. Ricardo [38, 2:331] also illustrates the point very well when he argues that "capital is not bought and sold, it is borrowed at interest, and a great interest is given when profits are high."

9. An anonymous reader of an earlier version of this article fails to find citations of "saving" in the index of "either the Wealth of Nations or Ricardo's Principles," and concludes that "'saving' did not figure as a technical term in the discourse of the early classical writers," hence Keynes could not have misinterpreted them. But quotations from the classics equating savings with "capital" in this article, and which exclude hoarding, show otherwise.

10. Evidence of Keynes's misinterpretation of the term "capital" by the classics as well as by Marshall is amply illustrated in the General Theory, esp. 186-93.

11. It does not matter whether cash takes the form of specie or paper money for its holding to constitute hoarding since neither form yields interest or profit income to its holder. Moreover, the holding of cash in whatever form withholds income from being spent on goods and (non-cash) services. I elaborate this point below.

12. The purchase of investment goods directly by households or sole proprietorships out of their incomes involves saving or abstinence and also amounts to (real) capital accumulation. But such purchases do not involve borrowing and lending, hence are excluded from [Delta]FA. They also do not enter directly into the theory of interest rate determination discussed in this article. For elaboration, see Ahiakpor [1, 510-11].

13. The specification, pY, is similar to Marshall's formulation of the demand for cash, [H.sub.d] = kY, where k is the proportion of income the community, including the business sector, desires to hold in cash, H is the stock of money (cash), and Y is nominal income. But p is less than k.

14. It is also not necessary to separate the proportion of income held in cash into "active" and "idle" balances in order correctly to define savings. Any amount of income held in cash is not available for lending for as long as it is held, and does not earn interest or profit income. Smith [50, 1:340-41] makes this point well, arguing: "That part of his capital which a dealer is obliged to keep by him unemployed, and in ready money for answering occasional demands, is so much dead stock, which, so long as it remains in this situation, produces nothing either to him or to his country." [Emphasis added.] However, note that this argument does not deny the benefits of holding cash, but how much more could be achieved by employing the credit system of payment through the institution of banking.

15. In discussing the role of the banking system, Keynes comes pretty close to accepting this argument. He disputes the claim that "a depositor and his bank can somehow contrive between them to perform an operation by which savings can disappear into the banking system so that they are lost to investment, or, contrariwise, that the banking system can make it possible for investment to occur, to which no savings corresponds" [22, 81]. But on the next page Keynes changes the sequence so that it is increased income due to bank credit creation which leads to increased savings.

16. Pigou [37], Robertson [40; 42, esp. ch. 1], Viner [51], Hawtrey [11], and Hazlitt [12, ch. 8] are among those who have invoked this positive role of savings to criticize Keynes's argument that increased saving depresses investment. Even Joan Robinson [43, 47-48], while defending Keynes's arguments, affirms the validity of this point when she argues that "it is thriftiness which makes investment possible" and that "in an age of expansion, thriftiness appears as a cause of investment."

17. The term "forced saving" may be understood this way. The inflation of credit by a central bank (an increase in high-powered money or cash) or commercial banks through their reduction of economic or excess reserves causes the supply of money to exceed its demand ([H.sup.s] [greater than] [H.sup.d]). This causes the price level to rise and reduce the real income of fixed-income earners. Such reduction in the purchasing power of their income to the benefit of residual-income (profits) earners or finns is what constitutes forced saving - a transfer of purchasing power to investors without the reward of interest payments. This is a concept whose clear meaning Keynes [22, 79-81, 183, 328] had considerable difficulty in recognizing. To him, the fact that "the new money is not 'forced' on anyone" [22, 328], makes the concept meaningless. Also contrast my clarification with, e.g., Chick's [9, 190-91, 236-39], which restates Keynes's position.

18. Also see Smith [50, 1:359] and Marshall's Pure Theory of Domestic Values, p. 34, quoted in Keynes [22, 19]. Mill's clarification shows that Keynes misinterprets the classical definition of saving when he understands them as having argued that "any individual act of abstaining from consumption necessarily leads to, and amounts to the same thing as, causing the labour and commodities thus released from supplying consumption to be invested in the production of capital wealth" [22, 19; emphasis added]. Keynes's interpretation also incorrectly suggests a failure of the classics to recognize the existence of the demand for money (cash).

19. See Smith [50, 1:359 n. 1]. Blaug [5, 55] argues that "Smith's way of putting the matter is positively misleading," but also notes that "Properly understood, the [saving-is-spending] dictum attacks the popular fallacy that saving necessarily destroys purchasing power; this is why it appealed to Smith's followers." The logic of the classical argument also may be explained as follows. Labor may be employed in wealth creating ('productive') or non-wealth creating ('unproductive') activities. Saving transfers purchasing power from 'unproductive' to the 'productive' sectors of the economy. Thus, saving does not reduce total demand or consumption, but only transfers spending between groups. Also see, for example, Ricardo [38, 2:309, 326-27; 8:181; and 10:408].

20. In a fractional reserve banking system, as explained above, such new credit could arise from a reduction in the level of economic reserves (or hoarding) by banks. But note that without having borrowed from a central bank, such reserves must first have been deposited by savers.

21. Keynes [22, 180-81], even with the aid of a diagram, fails to recognize this analysis in classical theory. Without determining the rate of interest from the supply and demand for money (cash), he insists, we could not determine equilibrium between saving and investment. As Ahiakpor [1] explains, Keynes failed to recognize the classical savings-investment-demand theory of interest rates mainly because he did not recognize their use of "capital" to mean loanable savings or "funds."

22. The analysis also casts doubt on the validity of Leijonhufvud's [26, 171, n. 60] assertion that "What Keynes should have taught is that loanable funds supply and demand govern the rate of interest and that it takes numerous restrictive assumptions to make saving and investment correspond to [loanable funds] supply and demand, respectively."

23. Surprisingly, Keynes [22, 343n] quotes Hume's essay on money but not on interest. And while Smith uses Hume's clarification that low interest rates do not arise from the abundance of money to criticize the views of Montesquieu, Law and Locke, Keynes [22, xxxiv, xxxv] praises Montesquieu's money (cash) theory of interest over Smith's.

24. Among the exceptions considered by Marshall is the possibility that the desire to provide for old age and family could be so strong as to cause some savers to pay borrowers for taking charge of their savings until needed in future, "and interest would be negative all along the line" [29, 582, n. 1].

25. Also see Keynes [22, 62-65]. In the Treatise, Keynes [21, 1:172] also describes saving as "the negative act of refraining from spending the whole of [one's] current income on consumption." [Emphasis added.]

26. However, Keynes [22, 107] recognizes that "a larger real consumption at a later date [being] preferred to a smaller immediate consumption" is among the motives for saving by individuals. It is his alleged negative consequences of saving for growth which he is here emphasizing.

27. Pigou [37, 126] provides an excellent rebuttal of this argument with an analogy of the decision to have ones' teeth pulled and the decision to buy false teeth. Instead of recognizing Keynes's argument as arising from his failure to appreciate the role of savings in determining the rate of interest, which does affect borrowers' spending decisions, Leijonhufvud [26, ch. 7] formalizes the argument as a failure of "intertemporal coordination."

28. Keynes [21, 1:140-41] makes a similar argument: "When a man is deciding what proportion of his money-income to save, he is choosing between present consumption and the ownership of wealth. . . . in so far as he decides in favour of saving, there still remains a further decision for him to make. For he can own wealth by holding it either in the form of money (or the liquid equivalent of money) or in other forms of loan or real capital."

29. Also see Marshall's [30, 46] use of the term "to invest" in reference to households' purchase of financial assets, consistent with his urging that economists use "language of the marketplace" [29, 60]. Richard Kahn [19, 549], for example, notes that Keynes "allowed himself to become confused by the two quite different meanings of the word 'investment' - in securities and in new capital goods."

30. But in a 1939 Times article, Keynes argues that an increase in saving reduces the rate of interest: "The clue to the solution of the Treasury problem lies in the Treasury's ability to wait until the new savings have had time to become available in an investable form. If the Treasury waits just long enough for the market to become greedy for stock the weight of savings seeking investment will force the rate of interest downwards." Quoted in Robertson [42, 24 n. 3; emphasis mine].

31. Of course, Keynes [22, 141-42] himself incorrectly denies that an increase in investment demand or "the marginal efficiency of capital" would raise the rate of interest.

32. Yet this mistaken criticism of classical analysis by Keynes appears to stand firmly in the way of explaining his error. For example, a previous anonymous reader cites it in Keynes's defense as being "at the centre of [his] critique," as if it were valid.

33. Robinson [44, 27] also restates this Keynesian view of saving when she argues: "If private saving is going on, there is a leakage of notes [cash] out of circulation into hoards." Chick [9, 57, 185] and Patinkin [34, 63, 270-72], for example, re-affirm.

34. A previous anonymous reader tries to absolve Keynes from his error in defining savings to include hoarding by arguing that Marshall "explicitly allowed that people might sometimes provide for the future by amassing coin [29, 192]. If Keynes is guilty of departing from an established classical usage then so is his teacher, Marshall." Besides attempting to substitute a possible exception (I could not locate the statement on the cited page in my copies of the eighth edition of the Principles) for the rule, this argument ignores Marshall's explanations of the connection between (positive) interest and saving [29, Book IV, ch. 7] which properly excludes hoarding, or Keynes's failure [22, 189] to recognize any valid theory of interest in the Principles, or to interpret correctly Marshall's arguments as cited in this article.

35. Also see Keynes [22, 174]. As Robertson [42, 16] correctly points out, Keynes's argument represents "a curious inhibition against visualizing more than two margins at once." The rate of interest balances at the margin the utility of cash (the convenience and security of cash), the return of investment income, and utility of immediate consumption. Keynes seems to recognize only the first two of these marginal conditions.

36. See, e.g., Ricardo [38, 2:301-31].

37. Keynes's statement of Say's Law of markets can be misleading if it suggests that suppliers of goods and services create the demand for what they themselves sell.

38. Also recall Malthus's explanation that saving does not mean hoarding, quoted above.

39. See, e.g., Ahiakpor [1, 523-24] for some of such arguments.

40. Fortunately, most development economists emphasize the need for increased saving to promote economic growth in the Third World, paying hardly any attention to the paradox of thrift argument. Some probably consider the argument just one of those peculiar propositions in "neoclassical" economics, which they claim do not apply to Third World economies.

41. Smith [50, 1:478] makes the remark in the context of international trade, not saving. Although consistent with Smith's position on saving, Keynes's use of the quotation to contrast with Mandeville's argument that belief in the virtues of public frugality "is an error" [22, 361], even when qualified with "probably with reference to . . . ," is yet another evidence of why one needs to be careful in accepting Keynes's interpretations of the classics.

References

1. Ahiakpor, James C. W., "On Keynes's Misinterpretation of 'Capital' in the Classical Theory of Interest." History of Political Economy Fall 1990, 507-28.

2. Bach, George L., with R. Flanagan, J. Howell, F. Levy, and A. Lima. Macroeconomics. Eleventh edition. Englewood Cliffs, N.J.: Prentice-Hall, 1987.

3. Baird, Charles W. Elements of Macroeconomics. Second Edition. St. Paul, Minn.: West Publishing, 1981.

4. Baumol, William J. and Alan S. Blinder. Macroeconomics. Fifth edition. New York: Harcourt Brace Jovanovich, 1991.

5. Blaug, Mark. Economic Theory in Retrospect. Fourth edition. Cambridge: Cambridge University Press, 1985.

6. Boyes, William and Michael Melvin. Economics. Boston: Houghton Mifflin, 1990.

7. Case, Karl E. and Ray C. Fair. Principles of Macroeconomics. Englewood Cliffs: Prentice Hall, 1994.

8. Campagna, Anthony S. Macroeconomics. New York: St. Martin's Press, 1981.

9. Chick, Victoria. Macroeconomics After Keynes. Cambridge: The MIT Press, 1983.

10. Hartfield, H. R., "The Early Use of 'Capital.'" Quarterly Journal of Economics, November 1934, 436-43.

11. Hawtrey, Ralph G., "Alternative Theories of the Rate of Interest." Economic Journal, September 1937, 436-43.

12. Hazlitt, Henry. The Failure of the "New Economics." Princeton, N.J.: Van Nostrand, 1959.

13. -----. Economics In One Lesson. Westport, Conn.: Arlington, 1978.

14. Heilbroner, Robert L. and James K. Galbraith. The Economic Problem. 9th ed. Englewood Cliffs: Prentice Hall, 1990.

15. Henderson, J. Vernon and William Poole. Principles of Macroeconomics. Lexington, Mass.: D.C. Heath, 1991.

16. Hicks, John R., "Mr. Keynes' Theory of Employment." Economic Journal, June 1936, 238-53.

17. Hume, David [1752], "Of Interest," in Hume's Writings on Economics, edited by E. Rotwein. Edinburgh: Nelson, 1955.

18. Johnson, Harry G., "The Keynesian Revolution and the Monetarist Counter-Revolution." American Economic Review, May 1971, 1-14.

19. Kahn, Richard, "Some Aspects of the Development of Keynes' Thought." Journal of Economic Literature, June 1978, 545-59.

20. Keynes, John M. Monetary Reform. New York: Harcourt, Brace, 1924.

21. ----- [1930]. A Treatise on Money. London: Macmillan, 1958.

22. ----- [1936]. The General Theory of Employment, Interest and Money. Paperbound ed. London and Basingstoke: Macmillan, 1974.

23. -----, "The General Theory of Employment." Quarterly Journal of Economics, February 1937, 209-23.

24. -----, "Alternative Theories of the Rate of Interest." Economic Journal, June 1937, 241-52.

25. Leijonhufvud, Axel. On Keynesian Economics and the Economics of Keynes. New York: Oxford University Press, 1968.

26. -----. Information and Coordination. New York: Oxford University Press, 1981.

27. -----. "What Was the Matter with IS-LM?" in Modern Macroeconomic Theory, edited by Jean-Paul Fitoussi. Oxford: Basil Blackwell, 1983.

28. Malthus, Thomas R. [1836]. Principles of Political Economy. Second edition. New York: Kelley, 1964.

29. Marshall, Alfred [1920]. Principles of Economics. Eighth edition. Reprinted. London: Macmillan, 1964.

30. ----- [1923]. Money, Credit and Commerce. New York: Kelley, 1960.

31. McConnell, Campbell R. and Stanley L. Brue. Macroeconomics. Twelfth edition. New York: McGraw-Hill, 1993.

32. Mill, John S. Collected Works. Edited by J. M. Robson. London: University of Toronto Press, 1965.

33. Parkin, Michael. Macroeconomics. Second edition. New York: Addison-Wesley, 1993.

34. Patinkin, Don. Money, Interest and Prices. Second edition. New York: Harper & Row, 1965.

35. -----. Anticipations of the General Theory? Chicago: University of Chicago Press, 1982.

36. Pigou, Arthur C. Industrial Fluctuations. London: Macmillan, 1927.

37. -----, "Mr. J. M. Keynes's General Theory of Employment, Interest and Money." Economica, May 1936, 115-32.

38. Ricardo, David. Works and Correspondence. Edited by Piero Sraffa. Cambridge: Cambridge University Press, 1951, 1957.

39. Robertson, Dennis H., "Saving and Hoarding." Economic Journal, September 1933, 399-413.

40. -----, "Some Notes on Mr. Keynes's General Theory of Employment." Quarterly Journal of Economics, November 1936, 168-91.

41. -----, "Alternative Theories of the Rate of Interest." Economic Journal, September 1937, 428-36.

42. -----. Essays in Monetary Theory. London: Staples, 1940.

43. Robinson, Joan. Introduction to the Theory of Employment. London: Macmillan, 1937.

44. -----. Exercises in Economic Analysis. London: Macmillan, 1960.

45. Samuelson, Paul A., "The Canonical Classical Model of Political Economy." Journal of Economic Literature, December 1978, 1415-34.

46. ----- and William D. Nordhaus. Economics. Twelfth edition. New York: McGraw-Hill, 1985.

47. -----. Economics. Thirteenth edition. New York: McGraw-Hill, 1989.

48. -----. Economics. Fourteenth edition. New York: McGraw-Hill, 1992.

49. Skidelsky, Robert. John Maynard Keynes. Vol. 1. London: Macmillan, 1983.

50. Smith, Adam [1776]. The Wealth of Nations. Edited by E. Cannan. Vols. 1 and 2. Chicago: University of Chicago Press, 1976.

51. Viner, Jacob, "Mr. Keynes On the Causes of Unemployment." Quarterly Journal of Economics, November 1936, 147-67.

52. Wonnacott, Paul and Ronald Wonnacott. Economics. Third edition. New York: McGraw-Hill, 1986.