populationist doctrine was France. The second step toward a solution is therefore to find out whether there was not something in the economic and political situation of France that might, ‘objective’ opportunities notwithstanding, suggest pessimism as regards the economic future of the country and thus explain that change in attitude. As a matter of fact, there was. During practically the whole of the eighteenth century France was fighting a losing battle with England. Many of her leading spirits began to accept this defeat by 1760 and to discount the opportunities for national expansion. Moreover, the outworn institutional pattern of the last half century of the monarchy was not favorable to vigorous economic development at home. Thus, thought turned from bold venture to the possibilities offered by agriculture, from dreams of evolution to the picture of a ‘mature’ or quasi-stationary economy. The third and final step, then, is to explain why anti- populationist sentiment gained a hold on the English mind in spite of the fact that exactly the opposite state of things prevailed in England. In order to understand this we have to realize that the long-run trend of an evolution is one thing, and the sequence of short-run situations through which it fights its way is quite another thing. Thus, the English populationists of the seventeenth and eighteenth centuries may have been quite right in considering rapid increase as motor, condition, and symptom of economic development, and equally right in worrying at the same time, as most of them actually did, about the short-run vicissitudes, the unemployment in particular, which accompanied that development; this does not convict them of contradiction either in their analysis or in their recommendations. But in the Industrial Revolution of the last decades of the eighteenth century, these short-run vicissitudes grew more serious than they had been before, precisely because the pace of economic development quickened. And some economists—as will be pointed out in a moment, a minority only—were so impressed by them as to lose sight of the trend. The resulting anti-populationist mood then produced the set of analytic propositions that came to be known, in the nineteenth century, as the Mal-thusian principle or theory of population. Before considering its early history, we must attend to another matter. [(b) Growth of Factual Knowledge.] In the United States the first census was taken in 1790; in England in 1801. In Canada and in some countries of continental Europe there had been earlier ones, but it was only in the first decades of the nineteenth century that reliable information about the numerical facts of population became available at regular intervals. The writers of the seventeenth and eighteenth centuries, therefore, theorized about population in ignorance of statistical facts. All they had to go on, if we except the rare cases in which local observation yielded definite results, were untrustworthy indications and vague impressions: thus, it was possible for English writers to disagree on such a question as whether the English population had increased or decreased during the century between 1650 and 1750. Hence, the investigations that were undertaken in order to dispel this fog and the resulting controversies exemplify a peculiar type of theory. Ordinarily, theoretical analysis is concerned with facts that are, or are supposed to be, known: it marshals, interprets, explains, establishes relations between, and generalizes from given facts or ‘data.’ This, of course, is also what the theory of population was to do in the nineteenth century. But in the seventeenth and eighteenth centuries, the main task of research on population was History of economic analysis 242 not to analyze given facts but, so far as possible, to find out what the facts actually were: it was the kind of theory that, unlike other kinds, retreats before advancing factual knowledge and must eventually be replaced by it. But the work done by those investigations—first by the Political Arithmeticians—also laid the foundations of the later theory of population. For many of the considerations that were originally developed in order to form an idea of the facts, served later on to interpret them. This is why examples of those controversies are presented below. Sir William Petty’s Essay concerning the Multiplication of Mankind (2nd ed. rev. and enl., 1686) is the standard example of seventeenth-century speculation about the facts. Sir Matthew Hale’s Primitive Origination of Mankind (1677; partly republ. in 1782 under the title of Essay on Population; on the author see J.B.Williams, Memoirs of the Life, Character and Writings of Sir Matthew Hale, 1835) may also be mentioned. Both authors infer facts, on scanty observations, mainly from ‘laws’ derived from general considerations. Of eighteenth-century controversies we shall first notice one that arose from Montesquieu’s statement in the Lettres persanes that the ancient world was more populous than was the Western world of his time. In his essay ‘Of the Populousness of Antient Nations’ (Political Discourses, 1752), Hume proffered reasons for the opposite opinion that were criticized by Robert Wallace in the Appendix to his Dissertation on the Numbers of Mankind (1753), in which he upheld Montesquieu’s thesis. Wallace found a follower in William Bell, who expanded the discussion on numbers into a discussion of causes and effects: in his dissertation What Causes principally Contribute to Render a Country Populous? And what Effect has the Populousness of a Nation on its Trade? (1756) he presented the theory that the development of manufacture and trade, by diverting resources from the production of foodstuffs, tends to produce a decrease in population (which he took to be a fact and of which he disapproved); accordingly, he advocated fostering agriculture and an equal distribution of land among farm families. This tract called forth another, A Vindication of Commerce and the Arts (1758) by W.Temple (a clothier, not to be confused with Sir William Temple, the seventeenth- century statesman and writer). No great importance attaches to either Bell’s or Temple’s works. They are mentioned here because of a similar discussion on a similar subject that took place half a century later and is much better known: opinions not unlike those of Bell, having been reasserted by Thomas Spence, elicited a reply by James Mill that established his reputation as an economist. Another controversy was more interesting. In 1779, Richard Price, now mainly remembered on account of his proposal to establish a sinking fund that would extinguish the national debt, published an Essay on the Population of England in which he stated that population had decreased by one-fourth since the revolution of 1688 and that urban agglomeration was responsible for it. Naturally this was attacked by a number of writers, Population, returns, wages, and employment 243 especially by W.Wales (An Inquiry into the Present State of Population in England and Wales, 1781), John Howlett (Examination of Dr. Price’s Essay…1781), and others, A.Young among them. Howlett’s contribution is the most interesting one, not only be cause it is a good example of the art of reasoning on inadequate facts but also because, like Bell, he launched out into an analysis of related economic phenomena. In particular, he interpreted enclosures as a consequence of the increase in population and as a ‘cause’ of some of those improvements in agriculture that were called for because of that increase theory in which there was an important element of truth. [(c) Emergence of the ‘Malthusian’ Principle.] The theory of population as understood in the nineteenth century, however, that is to say, a theory of the factors—or ‘laws’—that determine numbers and rates of increase or decrease, emerged much earlier than that. 3 Divested of nonessentials, the ‘Malthusian’ Principle of Population sprang fully developed from the brain of Botero in 1589: populations tend to increase, beyond any assignable limit, to the full extent made possible by human fecundity (the virtus generativa of the Latin translation); the means of subsistence, on the contrary, and the possibilities of increasing them (the virtus nutritiva) are definitely limited and therefore impose a limit on that increase, the only one there is; this limit asserts itself through want, which will induce people to refrain from marrying (Malthus’ negative check, prudential check, ‘moral restraint’) unless numbers are periodically reduced by wars, pestilence, and so on (Malthus’ positive check). This path- breaking performance—the only performance in the whole history of the theory of population to deserve any credit at all—came much before the time in which its message could have spread: it was practically lost in the populationist wave of the seventeenth century. But about two hundred years after Botero, Malthus really did no more than repeat it, except that he adopted particular mathematical laws for the operation of the virtus generativa and the virtus nutritiva: population was to increase ‘in geometric ratio or progression’—that is, in a divergent geometric series—food in ‘arithmetic ratio or progression.’ 4 But the ‘law of geometric progression,’ though not in Botero’s work, was suggested by Petty in his Essay concerning the Multiplication of Mankind (1686), by Süssmilch (1740), by R.Wallace (1753), and by Ortes (1774), so that, within this range of ideas, there was nothing left for Malthus to say that had not been said before. Of the eighteenth-century authors who, 3 See especially: René Gonnard, Histoire des doctrines de la population (1923); J. Bonar, Theories of Population from Raleigh to Arthur Young (1931); C.E.Stangeland, Pre-Malthusian Doctrines of Population (1904); J.J.Spengler, French Predecessors of Malthus…(1942); F.Virgilii, Il Problema della popolazione (1924). Readers are referred to these works for details of a story that cannot be presented here. 4 If an initial value be denoted by a and another constant by b, then a geometric series runs like this: a, ab, ab 2 , ab 3 ,… The series is divergent, i.e. the sum of its elements soars above any figure we care to name, if b is equal to, or greater than, unity. An arithmetic series runs like this: a, a+b, a+2b, a+3b… It is always divergent. History of economic analysis 244 without committing themselves to this particular mathematical form, stated that population will always increase to the limit set by the supply of means of subsistence, it will suffice to mention Franklin 5 (1751), Mirabeau (1756)—who expressed himself in his picturesque manner: men will multiply to the limits of subsistence like ‘rats in a barn’— Sir J.Steuart (1767), Chastellux (1772), 6 and Townsend (1786). 7 Steuart, whose priority Malthus was to acknowledge, was particularly explicit. Exactly as Botero did, he took the ‘generative faculty’ as a constant force to be compared to a spring that is held down by a weight and is certain to respond to any decrease in this weight. Townsend defined the limiting factor as ‘hunger, not as directly felt or feared by the individual himself, but as foreseen and feared for his immediate offspring.’ As far as I know, Ortes was the only writer to admit that ‘reason’ may have more influence than is implied in the anticipation of want—an influence that he illustrated by the celibacy of the Catholic clergy. Botero, then, was the first to sound that note of pessimism which was to become so famous a bone of contention in the days of Malthus: as we have seen, he associated increase of population with actual or potential misery. But most of the authors who believed that populations tend to increase without assignable limit did not share Botero’s pessimism but were on the contrary in sympathy with the populationist sentiments of their times and countries. Petty and, before their conversion to the Botero-Malthus view of he matter, Mirabeau and Paley may serve as examples. 8 This position involves, of course, no error of reasoning. For the fact that a population is physically capable of multiplying until it lacks not only food but also ground to stand on is no cause for worry unless complemented by the additional proposition that it actually will tend to do this instead of merely responding to an expanding economic environment by growing along with it (or even, possibly, by a decrease in the birth rate). In other words, population must actually tend to ‘press against’ the food supply. But even if such a tendency be admitted, it need not cause any worry about the calculable future or, what is more important for us, have any relevance to the explanation of contemporaneous phenomena. For this to be the case it is evidently not sufficient to believe that population will or may ‘press against’ food supply at some indefinitely distant time: we must believe the pressure to be either actually present or actually imminent. Unless this can be established, belief in that tendency is compatible with the opposite belief as regards any given situation or as regards the outlook ex visu of any given situation. The reader may well think that I am placing unnecessary emphasis upon these obvious distinctions, but their neglect is 5 Benjamin Franklin, Observations concerning the Increase of Mankind. Still more than others Franklin treated the case of human populations in the light of the general case of all animal species. On the other hand, he emphasized ‘room’ and ‘enemies’ as limiting factors rather than food. 6 François Jean, Marquis de Chastellux, soldier by profession, published a treatise De la félicité publique that is not without merit. 7 Joseph Townsend, see especially his Dissertation on the Poor Laws, 1786. 8 Petty listed populousness among the main assets of the Netherlands which made them such formidable competitors of England. Mirabeau, in those parts of L’Ami des hommes ou traité de la population that were published in 1756, declared that a large population is a blessing and the source of wealth: agriculture should be encouraged precisely because this would make people multiply like rats. It was Quesnay himself who induced Mirabeau to reverse the causal relation between wealth and population. William Paley (Principles of Moral and Political Philosophy, 1785, Book VI, ch. 11) held the same opinion. He was converted by Malthus’ Essay and recanted in his Natural Theology (1802). Population, returns, wages, and employment 245 responsible for the futility of many of the controversies that arose about population both in the eighteenth and in the nineteenth centuries. A work by R.Wallace 9 will, however, illustrate the way in which mere belief in pressure of population at some indefinitely remote future may after all be made relevant to economic analysis. Wallace considered equalitarian communism as the absolutely ideal form of society. Nevertheless, he rejected it. And the only reason he adduced for doing so was that in such a society there would be no check to the operation of mankind’s physical powers of multiplication, so that the career of a communist society would eventually have to end in overcrowding and misery—a standpoint that evidently did not imply any opinion about the situation that actually prevailed in Wallace’s time. Whatever we may think of the merits of the argument, it presents two characteristic features that cannot be underlined too strongly. First, if the proposition about unchecked multiplication were valid, it would evidently come near to being a ‘natural law’ in the strict sense of the term. Most of the English economists of the subsequent hundred years accepted it as such—as formulating an inexorable quasi-physical necessity. The same economists were in the habit of claiming similar necessity and universal validity, not only for those economic propositions that are nothing more than applied logic, but also for others such as their ‘law of wages.’ It is evidently not unreasonable to suspect that this habit of the English economists had something to do with their belief in that biological ‘law.’ If this be so, the question of the classic ‘eternal laws of economics’ should not be treated as a question of the philosophy of scientific method but simply as a question of the validity or relevance of an individual proposition. Second, it never seems to have occurred to Wallace to look for obstacles to human perfection other than mankind’s power of multiplication: except for the dangers that threatened from this he had no more doubt about human perfectibility than had Condorcet. This was in keeping with the superficial sociology of the Enlightenment, but it is interesting to note that Malthus and in fact all the ‘classics’ seem to have been of the same opinion. I know of only one writer who at least sounded the eugenic note. It was Townsend. In the work mentioned above he argued that provision for the ‘idle and vicious’ would put a burden on the ‘more prudent, careful, and industrious’ that would restrain them from marrying: ‘the farmer breeds only from the best of his cattle; but our laws choose rather to preserve the worst…’ The outstanding authority for the other opinion, that is to say, for the opinion that pressure of population was actually present around 1750—and is in fact an ever-present phenomenon—was Quesnay. Unlike Cantillon, from whom he broke away in this point, 10 he held not only that propagation has no other limits than those of subsistence but also that it tends always to go beyond them. The only justification he proffered for this dogmatic statement was that, always and everywhere, there are people who live in 9 Various Prospects of Mankind, Nature, and Providence, 1761. This work was criticized by Godwin and, since Malthus’ work in turn started from a criticism of the latter’s ideas, Wallace may have had more influence upon what became known as Malthusianism than any other of the writers who anticipated Malthus’ doctrine. Malthus did full justice to Wallace’s work but made it quite clear that, unlike Wallace and like Quesnay, he believed that pressure was an actual and indeed ever-present fact. 10 Substantially, Cantillon was populationist. But he touched in passing upon the problem ‘whether it is better to have a great multitude of poor people or a smaller number of more prosperous ones.’ History of economic analysis 246 poverty or want (indigence). This overpopulation theory of poverty is of the essence of ‘Malthusianism.’ But before the publication of Malthus’ Essay it had so few adherents that to this day most historians attribute it to him. Populationism did not indeed hold its own—not, at least, outside Germany and Spain. But everywhere economists refused to accept the opposite view. Most of them seem to have agreed with Bishop Berkeley, who delighted in the vision of joyfully bustling multitudes, or with Hume, who called the happiness of society and its populousness ‘necessary attendants.’ Accordingly, A.Smith summed up by reducing the principle of population to a stale truism, preserving however its character as a natural law: ‘every species of animals naturally multiplies in proportion to the means of their subsistence, and no species can ever multiply beyond it.’ (Wealth, Book I, ch. 8.) And at the same time he declared, in the spirit of the old populationists, that ‘the most decisive mark of the prosperity of any country is the increase of the number of its inhabitants’ (ibid.). Beccaria discounted both the enthusiasms and the pessimisms of economists about increasing numbers: he recognized that increase was not always a blessing to be prayed for at all times; but also that there was no reason for being afraid of it at all times. In fact he seems to have been the one authority to teach explicitly the obviously sensible view. Genovesi went further than this, however, in effecting a synthesis between the two opposites. He saw that, from the standpoint of a population living under given conditions, numbers are capable of being either too small or too great in the sense that increase or decrease would produce greater ‘happiness.’ This led Genovesi to reassert the old idea of optimum population (popolazione giusta, Lezioni, Part I, ch. 5) that was to be sponsored again by Knut Wicksell. This concept is difficult to handle and perhaps not very valuable. But it has the merit of bringing out the truth that populationism and Malthusianism are not the mutually exclusive opposites they seemed to be to so many people. 2. INCREASING AND DECREASING RETURNS AND THE THEORY OF RENT [(a) Increasing Returns.] We have seen that the populationist attitude, so far as it is economically motivated, implies a belief that increase in population will (within limits) increase per capita wealth or, as we may also put it, a belief in Increasing Returns. So does, in most cases, the protectionist attitude that went with populationism (see below, ch. 7). The idea of increasing returns in this sense—that is to say, increasing returns with reference to a national economy as a whole, and irrespective of any well-defined reason why returns should be increasing and of whether it is physical returns or returns in terms of money that are meant—is no doubt a hazy one and does not amount to more than an ‘inkling’ of any of the various meanings that the concept was to acquire. But beyond such inklings, which were of course very frequent, we also find here and there more precise arguments such as Petty’s argument that expenditure on what may be termed social overhead— expenditure on government, roads, schools, and so on—does not, other things being equal, increase proportionately with population: this puts increasing returns into the not quite equivalent form of decreasing cost per unit of service, but nevertheless identifies a Population, returns, wages, and employment 247 definite phenomenon that can be observed in every society and in every individual firm. Before this, a general law of increasing returns in manufacturing industry, also in the form of a law of decreasing unit cost, had been stated explicitly and in full awareness of its importance by Antonio Serra, 1 much as it was to be stated in the nineteenth-century text-book. The restriction of increasing returns to manufacturing should be particularly noticed. Serra did not indeed assert that agrarian production was subject to decreasing returns. But the idea that industrial and agrarian production as such follow different ‘laws’ was as clearly expressed by him as if he had. Thus he foreshadowed an important feature of nineteenth-century analysis that was not completely abandoned even by A.Marshall. In the seventeenth and eighteenth centuries, however, most economists said nothing about this. But many implied, or even explicitly said, that increasing returns prevailed also in agriculture. We shall presently discuss the most important example of this position. At the moment let us notice that A.Smith, more than a century and a half after Serra, took a view that was closely similar to his. He clearly, though loosely, stated a law of increasing returns for manufactures: first, in connection with division of labor (Book I, ch. 1) and, second and more fully, in the digression on the ‘Effects of the Progress of Improvement upon the Real Price of Manufactures,’ which he inserted into Part III of his huge chapter on the rent of land (Book I, ch. 11), where he attributed the fact that ‘a much smaller quantity of labour becomes requisite for executing any particular piece of work’ ‘in consequence of better machinery, of greater dexterity, and of a more proper division and distribution of work.’ 2 But nowhere did he state a law of decreasing returns, though he repeatedly brushed against it, especially in Chapter 11. In fact, in Chapter 1, he merely noted a difference between agricultural and industrial production in the scope they offer for ever increasing division of labor, and his text is compatible with the interpretation that he meant to assert increasing returns also for agriculture but to a lesser degree. And this in spite of the fact that the two cases of decreasing (physical) returns, which West and Ricardo were to recognize, had been fully described before him by Sir James Steuart (1767) and Turgot (1767). 3 [(b) Decreasing Returns: Steuart and Turgot.] Steuart in his Principles (1767)—and after him Ortes in his Economia Nazionale (1774) —presented what the late followers of Ricardo were to call the case of the Extensive 1 Breve trattato (1613), Part I, ch. 3: nell’ artefici vi può essere moltiplicazione… e con minor proporzione di spesa (in manufacturing industry, output may be increased at less than proportional increase in expense). Serra does not tell us to what this fall in cost is due. It may be plausibly assumed, however, that he thought of the same facts that A.Smith was to enumerate. 2 Observe that this statement mixes up two entirely different things: ‘better’ machinery seems to point to an effect of the widening of knowledge—the Technological Horizon—that occurs in the course of economic development. Improved division of labor, on the other hand, is one of the consequences of mere increase in output and may occur within an unchanging technological horizon or an unchanging state of the industrial arts. 3 Observations Sur le Mémoire de M.de Saint-Péravy, included in the editions of the Oeuvres (mentioned in ch. 4, sec. 4). The date given in the text is not absolutely certain; moreover, it is the date of writing. We do not know how wide or narrow a circle of readers the paper reached at the time. History of economic analysis 248 Margin: as population increases, poorer and poorer soils have to be taken into cultivation and, applied to these progressively poorer soils, equal amounts of productive effort produce progressively smaller harvests. Turgot discovered the other case of decreasing physical returns, the one that the same followers of Ricardo were to refer to as the case of the Intensive Margin: as equal quantities of capital (avances)—amounts of labor would, however, do just as well in this case—are successively applied to a given piece of land, the quantities of product that result from each application will first successively increase up to a certain point at which the ratio between increment of product and increment of capital will reach a maximum. Beyond this point, however, further application of equal quantities of capital will be attended by progressively smaller increases in product, and the sequence of these decreasing increases will in the end converge toward zero. This statement of what eventually came to be recognized as the genuine law of decreasing returns cannot be commended too highly. It embodies an achievement that is nothing short of brilliant and suffices in itself to place Turgot as a theorist high above A.Smith. It is much more correct than are most of the nineteenth- century formulations—Turgot’s formulation was indeed not surpassed until Edgeworth 4 took the matter in hand. A particularly felicitous feature is the insertion of an interval of increasing returns before the interval of decreasing returns; that is to say, the recognition of the fact that decreasing returns do not prevail right from the application of the first ‘dose’ of some variable factor but set in only after a certain point has been reached. This should have disposed, once for all, of the erroneous opinion that he who asserts that extension of production will, under given circumstances, be attended by increasing returns therefore denies the validity of the ‘law of decreasing returns.’ Moreover, Turgot’s increasing returns are defined with unsurpassable neatness: they are the increasing returns that attend the application of a variable factor to one that is given in a fixed quantity—or to a set of factors whose quantities are held constant—before the optimum combination of factors is attained. Thus, Turgot may be said to have formulated a special case of what American economists around 1900 were to call the Law of Variable Proportions. 5 4 See below, Part IV, ch. 6, sec. 5b. 5 The same thing may be expressed, by means of a different concept, in a somewhat different way. This concept, which emerged toward the end of the nineteenth century (see below, Part IV, ch. 7, sec. 8), is now being called the Production Function. This function expresses the technological relation that exists between the quantity of product and the quantities of the ‘factors’ that co- operate in varying proportions to produce it. Reducing, for the sake of simplicity, the number of these factors to two, we may mark off the quantities of the product and of the two factors on the axes of a system of rectangular space co-ordinates. Every point in space that corresponds to any positive and finite values of those three quantities will then represent that quantity of product that can (at best) be produced by the corresponding quantities of factors, and the set of all these points will identify a surface in three-dimensional space, the production surface. Now let one of the factor quantities be held constant, and cut this surface by a plane at right angles to this factor’s axis and go through the point on this axis that corresponds to the constant. The curve of intersection between the surface and the plane will represent Turgot’s law of first increasing and then decreasing returns. Though Turgot did not discover either the production function or its geometric picture, the production surface as such, we may say that he discovered a property of it, viz., the form of one of its contours, and hence that he got hold of something, possession of which (with ordinary care and competence prevailing in our science) should have brought out the production function of today Population, returns, wages, and employment 249 before the eighteenth century was out. The reason why this argument is being inflicted upon the reader at this stage is that the case is so revelatory of the ‘ways of the human mind,’ which rarely discovers the obvious and fundamental first. More often it gets hold of some particular aspect of an idea and then works back to the conceptions that hold priority in logic. And finally it must be recorded to Turgot’s credit that he stated his law in terms of successive increments of product and not in terms of average product (per unit of the variable factor). This means that he actually used marginal analysis and that command of modern technique could have improved only the form of his statement. There is really nothing to criticize in it except inadequate awareness of the necessity to specify both the product for which his law is to hold and also the variable factor that is to be applied: the basketful of disparate things that hide behind his avances does not meet the latter requirement but merely dodges it. 6 To the further objection that he did not emphasize the fact that his law made sense only with a given state of technological knowledge, or a given technological horizon, or a given production function—as we should say—he would probably have replied that this goes without saying. But we are about to see that this is not so. Another point must, however, be noticed before we go on. Both Steuart and Turgot spoke of agriculture only. Fifty years ago this would not have astonished anybody, since it was then established practice to restrict decreasing returns to agriculture. But we who take it for granted that neither increasing nor decreasing returns are restricted to any particular branch of economic activity but may prevail in any branch, provided certain general conditions are fulfilled, are in a position to realize how surprising that actually was. Explanation seems to lie in the fact that, to the unsophisticated mind, there is something particularly compelling in the limitations imposed upon human activity by an inexorably ‘given’ physical environment. It takes prolonged effort to reduce the analytic importance of these limitations to their proper dimensions and to divorce them from the soil and the industry that works the soil. Yet it should not have taken so long to see that there is really no logical difference between trying to expand output on a given farm and trying to expand output in a given factory, and that if farms cannot be indefinitely multiplied or enlarged, neither can factories. The additional explanation required is provided by the belief of practically all eighteenth- century authors—a belief that carries over to the ‘classics’ of the nineteenth century— that while the factor land was given once for all, the other original factor, labor, would always increase to any amount required if allowed to do so. If we adopt this view, we shall at once sympathize with the reluctance of those authors to treat labor and land alike and to apply the laws of physical returns impartially to both. Then we shall also sympathize with the lopsided analytic structure they set up. [(c) Historical Increasing Returns.] As we have seen above, asserting that, in a given situation, increasing returns prevail in a country’s agriculture, that is, that increase of input would be attended by more than proportional increase in output, does not imply denial of the validity of the law of 6 Unless the factor applied is such a definite physical thing as a fertilizer of invariant kind and quality or even labor of a given kind and quality, difficulties arise that threaten the meaning of the law. History of economic analysis 250 decreasing returns. This fact must now be brought to bear upon the interpretation of the views of those English economists and politicians who actually did make that assertion. Whether right or wrong in point of fact, their position was logically defensible if they meant no more than either or both of two things. They were all right as to logic (though possibly wrong as to their facts) if they meant that in the last decades of the eighteenth century English agriculture 7 was moving in an interval of increasing returns, that is to say, that land had not yet received its optimum complement of other factors. They were not less right in logic (and, to some extent, in fact) if they meant that there were looming in the future possibilities of improving agricultural methods of production that would materialize if additional resources (‘capital’) were made available to agriculture—in the same way in which this was actually happening in industry. Observe, however, that this is something quite different from the increasing returns we have been discussing. We can indeed, if we so choose, speak of increasing returns’ attending increased application of resources also in this case. But these spells of increasing returns, unlike the others, do not occur within the given pattern of technological practice. Like A.Smith’s improved machines they involve a change in this pattern. If we visualize Turgot’s intervals, first of increasing and then of decreasing returns, as a curve that ascends, reaches a maximum, and then descends, 8 then we see that the increasing returns in the previous sense are depicted by a section of the curve, but that increasing returns in the sense now under discussion are not. They can, however, be represented by shifting the whole curve upward (altering its shape or not as the case may require) into a new position: the old curve breaks off and is replaced by a new one that keeps a higher level (though not necessarily all along its course) but again displays both an interval of increasing returns in the previous sense and an interval of decreasing returns. The increase in returns in the new sense occurs as the curve shifts from its old to its new position. It should be added that, if the curve shifts again and again, there is no reason why the differences between these successive levels should grow smaller: there is no law of decreasing returns to technological progress. In order to avoid confusion between two entirely different phenomena, we had better restrict the term Increasing Returns to Turgot’s case. This we shall accordingly do. When we wish to retain the association between the two, misleading though it is, we shall use, for the phenomenon now explained, the phrase Historical Increasing Returns. The phrase has been chosen in order to indicate that these historical increasing returns cannot, like the genuine ones, be represented by any curve or ‘law,’ least of all by a curve on which we can travel back and forth. For new levels of technique are reached in the course of an irreversible historical process and are hidden from us until they are actually reached. 7 Those authors and politicians always spoke of returns to agriculture as a whole just as nineteenth- century economists were in the habit of doing. Strictly, however, the laws of returns in the sense of Turgot are defined only for the individual farm. It is an additional merit of Turgot that he so envisaged them. Transition to a whole industry, let alone the national economy as a whole, is not quite such plain sailing as primitive analysis assumes. 8 See footnote 5 above. To repeat, the abscissae of the curve there described represent successive equal ‘investments’ of some resource, say, labor of a given quality, the ordinates the corresponding amounts of total product. But we may also let the ordinates represent the increments in total product that successively result from each additional dose of ‘investment.’ Of course, this ‘derived’ curve (of marginal products) will reach its maximum before the other does. Population, returns, wages, and employment 251 . multitude of poor people or a smaller number of more prosperous ones.’ History of economic analysis 246 poverty or want (indigence). This overpopulation theory of poverty is of the essence of ‘Malthusianism.’. executing any particular piece of work’ ‘in consequence of better machinery, of greater dexterity, and of a more proper division and distribution of work.’ 2 But nowhere did he state a law of decreasing. thing as a fertilizer of invariant kind and quality or even labor of a given kind and quality, difficulties arise that threaten the meaning of the law. History of economic analysis 250 decreasing