Send us your blog post, blog address, address of other great sites or suggestions by email.

Wednesday, July 31, 2013

Ricardo and the Ricardian system, I: macro-income distribution

While much of the Ricardian system turns out to be the creation of James Mill, perhaps most of it was due to Ricardo himself, who of course must, in any case, bear major responsibility for his own work. To continue the Marxian metaphor, in many ways the Mill-Ricardo relationship might be more of a Marx-Engels than a Lenin-Marx connection,
Ricardo was born in London into a prosperous family of Spanish-Portuguese Jews who had settled in Holland after having been expelled from Spain at the end of the fifteenth century. Ricardo's father had moved to London, where he prospered as a stockbroker, and had 17 children, of whom David was the third. At the age of 11, David was sent by his father to Amsterdam, to attend Orthodox Hebrew school for two years. At the age of 14, with only an elementary education, Ricardo began his business career, employed by his father's ‘stockbroker’ house. It must be emphasized that, with the exception of the quasi-governmental Bank of England, there were no corporations or corporate stocks in that era. Government bonds were then called ‘stocks’, and so ‘stockbrokers’ were what would now be called government bond dealers.
Seven years later, however, David married a Quaker girl, and left the Jewish faith, whereupon he was disowned by his parents. Eventually, he became a confirmed Quaker. A London bank, already impressed with young Ricardo, lent him enough money to set himself up in his own business as a stockbroker. Within a few years, Ricardo made an enormous amount of money in the bond business, until he was ready to retire to the country in his early 40s. In 1799, at the age of 27, Ricardo, bored while whiling away time at a health resort, chanced upon a copy of The Wealth of Nations, and devoured it, becoming, like so many others of that era, a dedicated Smithian.
As Schumpeter points out, Ricardo's Principles can only be understood as a dialogue with, and reaction to, The Wealth of Nations. Ricardo's logical bent was offended at the basic confusion of mind, the chaos that J.B. Say also saw in the Smithian canon, and he, like Say before him, set out to clarify the Smithian system. Unfortunately, and in deep contrast to Say, Ricardo simplified by taking all the most egregious errors in Smith, throwing out all qualifications and contradictions, then building his system upon what was left. The worst of Smith was magnified and intensified. In his basic method, all of Smith's historical and empirical points were tossed out. This was not bad in itself, but it left a deductive system built on deep fallacy and incorrect macro-models. In addition, while Ricardo's theoretical system might have been brutally oversimplified in relation to Smith, his writing style was inordinately crabbed and obtuse. The methodology of verbal mathematics is almost bound to be difficult and obscurantist, with blocks of words spelling out equilibrium mathematical relations in a highly cumbersome manner. But on top of that, Ricardo, in contrast to his mentor Mill, was undoubtedly one of the worst and most turgid literary stylists in the history of economic thought.
In contrast to Adam Smith, for whom the output, or wealth, of nations was of supreme importance, Ricardo neglected total output to place overriding emphasis on the alleged distribution of a given product into macro-classes. Specifically, into the three macro-classes of landlords, labourers and capitalists. Thus, in a letter to Malthus, who on this question at least was an orthodox Smithian, Ricardo made the distinction clear: ‘Political economy, you think, is an enquiry into the nature and causes of wealth; I think it should rather be called an enquiry into the laws which determine the division of the produce of industry amongst the classes who concur in its formation.’
Since entrepreneurship could not exist in Ricardo's world of long-run equilibrium, he was left with the classical triad of factors. His analysis was strictly holistic, in terms of allegedly homogeneous but actually varied and diverse classes. Ricardo avoided any Say-type emphasis on the individual, whether he be the consumer, worker, producer or businessman.
In Ricardo's world of verbal mathematics there were, as Schumpeter has astutely pointed out, four variables: total output or income, and shares of income to landlords, capitalists, and workers, i.e. rent, profits (long-run interest) and wages. Ricardo was stuck with a hopeless problem: he had four variables, but only one equation with which to solve them:
Total output (or income) = rent + profits + wages
To solve, or rather pretend to solve, this equation, Ricardo had to ‘determine’ one or more of these entities from outside his equation, and in such a way as to leave others as residuals. He began by neglecting total output, i.e. by assuming it to be a given, thereby ‘determining’ output by freezing it on his own arbitrary assumptions. This procedure enabled him to get rid of one variable – to his own satisfaction.
Next, on to wages. Here, Ricardo took from Mill the hard-core, or ultra-Malthusian, view that ‘wages’ – all wages – are always and everywhere pressing on the food supply to such an extent that they are always set, and determined, precisely at the level of the cost of subsistence. Labour is assumed to be homogeneous and of equal quality, so that all wages can be assumed to be at subsistence cost. While briefly and dimly acknowledging that labour can have different qualities or grades, Ricardo, like Marx after him, drastically assumed away the problem by blithely postulating that they can all be incorporated into a weighted quantity of ‘labour hours’. As a result, Ricardo could maintain that wage rates were uniform throughout the economy. In the meanwhile, as we have seen, food, or subsistence generally, was assumed to be incorporated into one commodity, ‘corn’, so that the price of corn can serve as a surrogate for subsistence cost in general.
Given these heroic and fallacious assumptions, then, ‘the’ wage rate is determined instantly and totally by the price of corn, since the wage rate can neither rise above the subsistence level (as determined by the price of corn) nor sink below it.
The price of corn, in its turn, is determined according to Ricardo's famous theory of rent. Rent served as the linchpin of the Ricardian system. For, according to Ricardo's rather bizarre theory, only land differed in quality. Labour, as we have seen, was assumed to be uniform, and therefore wage rates are uniform, and, as we shall see, profits are also assumed to be uniform because of the crucial postulate of the economy's always being in long-run equilibrium. Land is the only factor which miraculously is allowed to differ in quality. Next, Ricardo assumes away any discovery of new lands or improvements in agricultural productivity. His theory of history therefore concludes that people always begin by cultivating the most fertile lands, and, as population increases, the Malthusian pressure on the food supply forces the producers to use ever more inferior lands. In short, as population and food production rise, the cost of growing corn must inexorably rise over time.
Rent, in Ricardo's phrase, is payment for the ‘use of the original and indestructible powers of the soil’. This hints at a productivity theory, and indeed Ricardo did see that more fertile and productive lands earned a higher rent. But unfortunately, as Schumpeter put it, Ricardo then ‘embarks upon his detour’. In the first place, Ricardo made the assumption that at any moment the poorest land in cultivation yields a zero rent. He concluded from that alleged fact that a given piece of land earns rent not because of its own productivity, but merely because its productivity is greater than the poorest, zero-rent, land under cultivation. Remember that, for Ricardo, labour is homogeneous and hence wages uniform and equal, and, as we shall see, profits are also uniform and equal. Land is unique in its permanent, long-run structure of differential fertility and productivity. Hence, to Ricardo, rent is purely a differential, and Land A earns rent solely because of its differential productivity compared to Land B, the zero-rent land in cultivation.
To Ricardo, several important points followed from these assumptions. First, as population inexorably increases, and poorer and poorer lands are used, all the differentials keep increasing. Thus, say that, at one point of time, corn lands (which sums up all land) range in productivity from the highest, Land A, through a spectrum down to Land J, which, being marginal, earns a zero rent. But now population increases and farmers have to cultivate more and poorer lands, say K, L, and M. M now becomes the zero-rent land, and Land J now earns a positive rent, equal to the differential between its productivity and that of M. And all the previous infra-marginal lands have their differential rents raised as well. It becomes ineluctably true, therefore, that over time, as population increases, rents, and the proportion of income going to rent, increase as well.
Yet, though rent keeps increasing, at the margin it always remains zero, and, as Ricardo put it in a crucial part of his theory, being zero rent does not enter into cost.
Put another way: quantity of labour cost, being allegedly homogeneous, is uniform for each product, and profits, being uniform and fairly small throughout the economy, form a part of cost that can be basically neglected. Since the price of every product is uniform, this means that the quantity of labour cost on the highest-cost, or zero-rent, land, uniquely determines the price of corn and of every other agricultural product. Rent, being infra-marginal in Ricardo's assumptions, cannot enter into cost. Total rental income is a passive residual determined by selling prices and total income, and selling prices are determined by quantity of labour cost and (to a small extent) the uniform rate of profit. And since the quantity of labour needed to produce corn keeps rising as more and more inferior lands are put into production, this means that the cost of producing corn and hence the price of corn keep rising over time. And, paradoxically, while rent keeps rising over time, it remains zero at the margin, and therefore without any impact on costs.
There are many flaws in this doctrine. In the first place, even the poorest land in cultivation never earns a zero rent, just as the least productive piece of machinery or worker never earns a zero price or wage. It does not benefit any resource owner to keep his resource or factor in production unless it earns a positive rent. The marginal land, or other resource, will indeed earn less of a rent than more productive factors, but even the marginal land will always earn some positive rent, however small.
Second, apart from the zero-rent problem, it is simply wrong to think that rent, or any other factor return, is caused by differentials. Each piece of land, or unit of any factor, earns whatever it produces; differentials are simple arithmetic subtractions between two lands, or other factors, each of which earns a positive rent of its own. The assumption of zero rent at the margin allows Ricardo to obscure the fact that every piece of land earns a productive rent, and allows him to slip into the differential as cause.
We might just as well turn Ricardo on his head and apply the differential theory to wages, and say, with Schumpeter, that ‘one pays more for good than for bad land exactly as one pays more for a good than a bad workman’.4
Third, in discussing the rise in cost of producing corn, Ricardo reverses cause and effect. Ricardo states that increasing population ‘obliges’ farmers to work land of inferior quality and then causes a rise in its price. But as any utility theory analyst would realize, the causal chain is precisely the reverse: when the demand for corn increases, its price would rise, and the higher price would lead farmers to grow corn on higher-cost land. But this realization, of course, eliminates the Ricardian theory of value and with it the entire Ricardian system.
And fourth, as numerous critics have pointed out, it is certainly not true historically that people always start using the highest-quality land and then sink gradually and inevitably down to more and more inferior land. Historically, there have always been advances, and enormous ones, in the productivity of agriculture, in the discovery and creation of new lands, and in the discovery and application of new and more productive agricultural techniques and types of products. Defenders of Ricardo counter that this is a purely historical argument, ignoring the logical beauty of the Ricardian theory. But the whole point is that Ricardo was, after all, advancing a historical theory, a law of history, and he certainly claimed historical accuracy for past and future predictions for his theory. And yet it is all a purely arbitrary, and hence largely untrue, assumption of his logical doctrine in the guise of a theory of history. Ricardo's basic problem throughout was making cavalier and untrue historical or empirical generalizations the building blocks of his logical system, from which he drew self-confident and seemingly apodictically true empirical and political conclusions. Yet from false assumptions only false conclusions can be drawn, regardless how imposing the logical structure may or may not be.
Ricardo's differential rent theory has been widely hailed as the precursor of the neoclassical law of diminishing returns, which the neoclassicals were supposed to have generalized from land to all factors of production. But this is wrong, since the law of diminishing returns applies to increasing doses of a factor to homogeneous units of other, logically fixed, factors – in this case land. But the whole point of Ricardo's differential rent theory is that his areas of land are not homogeneous at all, but varying in a spectrum from superiority to inferiority. Therefore the law of diminishing returns – as grasped by Turgot and rediscovered by the neoclassicals – simply does not apply.5
Rent, though increasing, is then effectively zero and not part of expenses or costs. Rent is disposed of in the Ricardian equation. But we have not yet finished the determination of wages, which so far we have said is precisely fixed at the subsistence level. What will happen to the costs of subsistence over time? They will rise as the cost of the production of corn rises with the increasing population, forcing the cultivation of ever more inferior lands. Over time, in the slow-moving long-run Ricardian equilibria, the cost of food will rise, and since wages must always be at the subsistence level, wages will have to rise to maintain real wage rates equal to the cost of subsistence. Now we begin to close the Ricardian circle. Rents are in effect zero, and wage rates, always at subsistence, must rise over time as the cost of food increases, in order to keep precise pace with the rising cost of subsistence. But, then -voila! – we have finally determined all the variables except profits (at least to Ricardo's satisfaction), and, since total income is ‘given’ or kept frozen, this means that profits are the residual from total income. With rents out of the picture, if wage rates have to keep rising over time, this necessarily means that profits, or profit rates, have to keep falling. Hence the Ricardian doctrine of the ever-falling rate of profit (i.e. long-term rate of interest). Note that this is not the same as Adam Smith's view that the profit rate falls over time because and in so far as capital continues to accumulate; profit was supposed to be an inverse function of the stock of capital. Ricardo's doctrine of the falling rate of profit follows by triumphant tautology from his attempt to determine the other factor shares of total income. When profits fall to zero, or at any rate to a low level, capital will cease to accumulate and we arrive at Ricardo's ‘stationary state’.
Ricardo, even more than Smith, totally leaves out the entrepreneur. There can be no role for the entrepreneur, after all, if everyone is always in long-run equilibrium and there is never risk or uncertainty. His ‘profits’, as in Smith, are the long-run rate of return, i.e. the rate of interest. In long-run equilibrium, furthermore, all profits are uniform, since firms rapidly move out of low-profit industries and into high-profit ones until equalization takes place. We then have ‘profits’ at a uniform rate throughout the economy at any given time.
A plausible insight into Ricardo's habitual confusion of long-run equilibrium and instantaneous adjustments with the real world has been offered by Professor F.W. Fetter. Fetter points out that Ricardo's practical familiarity was not with business and industry (as was, we might note, J.B. Say) but with the bond and foreign exchange markets. Ricardo ‘usually assumed that even in industry and agriculture, adjustment took place on the basis of as small price differences, and almost as quickly, as did arbitrage in government securities and in foreign exchange’.6
To return to the Ricardian world: note that Ricardo does not say that the cost of corn rises over time because rents keep rising on corn land. He must get rid of the rent variable, and he can only do so by assuming that rent is zero at the margin, and therefore never forms any part of costs. Rent, then, is effectively zero. Why then does the cost of corn rise? As we have indicated, because the quantity of labour needed to produce corn, and hence the cost of producing corn, rises over time. This brings us to Ricardo's theory of cost and value. Rents are now out of it. Wages are not costs either, because a key to Ricardo's system is that rising wages lead only to lower profits, and not to higher prices. If rising wages meant that costs increased, then Ricardo, who as we shall see had a cost-theory of value and price, would have to say that prices rose rather than that profits would necessarily fall. Wages he treated as uniform, since Ricardo, like Marx after him, maintained that labour was homogeneous in quality. Not only did that mean that wages were uniform; but Ricardo could then treat, as the crucial part of its labour cost, the quantity of labour embodied in any product. Differences in quality or productivity of labour could then be dismissed as simply trivial and as a slightly more complex version of the quantity of labour hours. Quality has been quickly and magically transformed into quantity.
We have reached the edge of the Ricardian – and Marxian – labour theory of value. So far we just have a labour-quantity theory of cost. Ricardo vacillated at this point, between a strict labour theory of cost, and a labour-quantity theory plus the uniform rate of profit. But, since the uniform rate of profit, presumably around 3–6 per cent, is small compared to the quantity of labour hours, Ricardo may be pardoned for dismissing the profit-rate part of cost as of trivial importance. And, since all profit rates are assumed to be uniform, and, as we shall see, Ricardo had a cost theory of value or price, he could easily dismiss the uniform and small proportion, profit, as of no account in explaining relative prices.
It is, of course, peculiar to consider profits, even profits as long-run interest, as part of the ‘costs’ of production. Again, this usage stems from eliminating any consideration of entrepreneurial profits and losses, and focusing on interest as a long-run ‘cost’ of inducing savings and the accumulation of capital.
If profits for Ricardo are always uniform, how is this uniform profit determined? Curiously, profits are in no way related to savings or capital accumulation; for Ricardo, they are only a residual left over after paying wages. In short, to hark back to our original equation of Ricardian distribution: total output (or income) = rent + profits + wages. Remarkably, Ricardo has attempted to determine all the variables with only one variable explicitly determined. Output, as we have seen, was assumed as mysteriously given, from outside the Ricardian system. Wages (‘the’ uniform wage throughout the economy) is the only explicitly determined variable, determined completely to equal the cost of subsistence, embodied in the cost of producing corn. But that leaves two residuals, rents and profits, to be determined. The way Ricardo tries to get around that problem is to dispose of rents. Rents are the differential between the lands in cultivation and the least productive, zero-rent, land in use. The cost of producing corn is equal to the quantity of labour hours embodied in its production. Since rents are zero at the margin, they do not enter into costs, and are passively determined; at the no-rent margin, labour and capital's shares exhaust output. And since wages are supposedly determined by the cost of raising corn, this means that profit can only be a truistic residual of wages, otherwise the variable would be overdetermined, and the system would evidently collapse.
The alleged historical laws follow from the model. Since increasing population forces more and more inferior land into cultivation, the cost of labour in producing corn (i.e. the quantity of labour hours needed to produce it), must keep rising. And since price is determined by cost, supposedly boiled down into the quantity of labour hours to produce the good, this means that the price of corn must keep rising over time. But since real wage rates are fixed always at the cost of subsistence, and this is assumed to be the price of corn, money wage rates must keep rising over time (while workers remain at the subsistence level), and therefore profits must keep falling in the course of history.
Adam Smith believed that the rate of profit, or the long-run rate of interest return, is determined by the quantity of accumulated capital, so that more capital will lead to a falling rate of profit. While this theory is not fully correct, it at least understands that there is some connection between saving, capital accumulation, and long-run interest or profit. But to Ricardo there is no connection whatever. Interest on capital is only a residual. By a series of fallacies, and holistic, locked-in assumptions, trivial conclusions are at last ground out, all with a portentous air, allegedly telling us conclusive insights about the real world. As Schumpeter scornfully puts it: propositions such as ‘profits depend upon wages’, and the falling rate of profit, are excellent examples of ‘that Art of Triviality that, ultimately connected with the Ricardian Vice, leads the victim, step by step, into a situation where he has got either to surrender or to allow himself to be laughed at for denying what, by the time that situation is reached, is really a triviality.

Austrian Perspective on the History of Economic Thought (2 volume set)

No comments:

Post a Comment