SOURCE: "Vilfredo Pareto, 1848-1923," in Ten Great Economists: From Marx to Keynes, Oxford University Press, 1965, pp. 110-42.

[In the following essay, which first appeared in the Quarterly Review of Economics in 1949, Schumpeter focuses on P areto 's economic theories.]

In a volume devoted to Pareto's life and work, Professor Bousquet relates that the obituary article devoted to Pareto in the socialist daily, Avanti, described him as the ' bourgeois Karl Marx. ' I do not know that a man can rightly be called ' bourgeois' who never missed an opportunity to pour contempt on la bourgeoisie ignorante et lache. But for the rest, the analogy conveys very well the impression that Pareto had made upon his countrymen: they had in fact raised him to an eminence that was unique among the economists and sociologists of his time. No other country erected a similar pedestal for his statue, and in the Anglo-American world both the man and the thinker have remained strangers to this day. There was, indeed, a short Pareto vogue in this country that followed upon the translation of his sociological treatise. But it died out soon in an uncongenial atmosphere. Moreover, so far as the small circle of pure theorists is concerned, Pareto came to exert considerable influence on Anglo-American economics in the 1920's and 1930' s, that is, after the publication of Professor Bowley's Groundwork. But both in England and the United States, Marshallian and post-Marshallian economics offered enough in the line in which Pareto excelled to prevent him from gaining much ground of his own even before other tendencies took away whatever he had gained.

This might seem surprising owing to the fact that several important developments in theoretical economics are now seen to stem from him. But it is not difficult to explain. Pareto was the product of a sector of the Franco-Italian civilization that is far removed from English and American currents of thought. Even within that sector his towering figure stood almost alone. Pareto cannot be pigeon-holed. He paid court to no ' ism. ' No creed or party can claim him as its own, although many creeds and parties appropriated fragments of the vast intellectual realm over which he held sway. He seems to have taken pleasure in running counter to ruling humors and slogans. Votaries of extreme laissez-faire may cull plenty of passages from his writings in support of their views. Yet there was nothing he despised so thoroughly as the ' pluto-democracy' or ' plutocratic demagogy' of liberalism. Socialists are under obligation to him for rendering, as we shall see, a very important service to socialist doctrine, and also for his protests against the anti-socialist measures that the Italian government took in 1898. Yet he was not only an anti-socialist but one ofthat type whose criticism derives sting from contempt. French Catholics might thank him for his attacks upon the persecution of the French clergy that was so unedifying a sequel to the Dreyfus affair. Yet he attacked the ' laicist' policies of the Combes ministry because he was a gentleman, and not because he believed either in the mission of the Catholic Church or in her teaching.

A gentleman of such independence and pugnacity who is in the habit of dealing vigorous blows right in the midst of arguments that might in themselves be agreeable to some party or another has little chance of being popular. By now he is a figure of the past. But even in the epoch of his prime the political and social slogans with which we are all familiar controlled official phraseology, the press, party programs, and popular literature including its economic sector. The wrapping in which he presented his strictly scientific results were then not much more popular than they would be now. One has only to imbue one-self with the spirit that pervades an American textbook and then open Pareto's Manuel in order to realize what I mean: the naïve lover of modern social creeds and slogans must feel himself driven with clubs from Pareto's threshold; he reads what he is firmly resolved never to admit to be true and he reads it together with a disconcerting wealth of practical examples. Therefore it seems that the problem is not to explain why Pareto did not exert influence more widely; the problem is rather to explain how he came to exert as much as he did.

Could we confine ourselves to Pareto's contributions to pure theory, there would be little need for glancing at the man and his social background and location. But into everything that was not a theorem in the pure logic of economics the whole man and all the forces that conditioned him entered so unmistakably that it is more necessary than it usually is in an appraisal of scientific performance to convey an idea ofthat man and of those forces. I shall make an attempt to do so first (I). Then I shall briefly survey Pareto's work in pure theory (II). And I shall end up with a glance at his conception of society that has found so inadequate an expression in his General Sociology (III).

I. THE MAN

Pareto's father, the Genoese Marchese Raffaele Pareto, seems to have been a typical product of the Italian Risorgimento of the first half of the nineteenth century, an ardent adherent of Mazzini—perhaps more from national than from social reasons—an uncompromising enemy of all the governments that barred Italy's way toward national unity, and a revolutionary in this if in no other sense. Accordingly, he exiled himself to Paris where Vilfredo, the subject of this memoir, was born of a French mother: if General Galliéni once described himself as ' Francese ma anche Italiano, ' Vilfredo Pareto might have described himself as ' Italiano ma anche Francese. ' He was taken to Italy in 1858 and there went through the usual course of studies that issued in a Doctor's degree in engineering in 1869. He immediately embarked upon engineering and industrial management as a profession and after various other appointments rose to be manager general—we should say ' president'—of the Italian Iron Works. It was only in 1893 that he was appointed successor to Walras in the University of Lausanne, although he may be considered as a full-time economist a few years before that. Thus, the span during which he was primarily engaged in economic research extends from about 1892 to about 1912—practically all his later work is sociological in nature. He resigned his chair in 1906 and then retired to his home, a country place on the lake of Geneva, to grow in the course of a vigorous and fertile old age into the ' lone thinker of Céligny.'

Substantially, this suffices for our purpose: we have to underline a few of these facts rather than to add others. First, theorists will note that owing to his training as an engineer—and he seems to have cultivated theoretical aspects—he acquired at an early age command of mathematics on a professional level. Second, it is worth while to notice that, to a degree quite unusual with scientific economists, Pareto was thoroughly familiar with industrial practice—familiar in a sense which is quite different from the kind of familiarity that may be acquired by die means available to the academic economist, the public servant, the politician. But, third, it was his passionate interest in the current issues of economic and general policy, presently to be commented on in another connection, which made him something of an economist long before he started his own creative work. Francesco Ferrara was then at the height of his fame and influence, and the frosts had not yet fallen upon a theoretical structure glorified by uncritical liberalism. His writings, especially his famous introductions (prefazioni) to the classics published in the Biblioteca dell' economista, served Pareto as well as, or better than, any of the university courses could have done that were available in his student days. His way to Walras, however, was chalked out later on by Maffeo Pantaleoni.

None of the facts above will account completely for Pareto's vision of society and politics, or even for his attitudes to the practical problems of his age and country. Nor do I believe for a moment that the deep pool of personality can ever be drained so as to show what is at the bottom of it. But there is the patrician background which, I am sure all who knew him will agree, meant much more in his case than it usually does. In particular it prevented him from becoming a brother in spirit to the men—and a fully accepted member of the various groups—with whom life threw him. It also prevented him from establishing emotional relations with the creations of the bourgeois mind, such as the twins that are called democracy and capitalism. Acting upon this background, his financial independence—a bare independence at first, something like affluence later on—helped to isolate him still further by offering the possibility of his isolating himself.

Still acting upon this background, his classical scholarship worked the same way. I do not mean that part of it which he shared with every educated person of his time, but that part which he conquered himself through incessant study of the Greek and Roman classics during his sleepless nights. The ancient world is a museum, not a laboratory of applied science, and he who trusts too much to the wisdom to be gathered there is bound to stray from every group of men that was in existence either in 1890 or 1920. Isolation was made complete by the result of his participation in the debates on the policies and politics of his country—so complete that he had decided to emigrate to Switzerland even before he received the call to Lausanne. And isolation had its effects—soothed only late in life by a second marriage that brought domestic peace (see the dedication of the Trattato)—upon a fiery temperament that was not really made to stand it.

But why should he have left his country in wrathful exasperation—the country that he loved from the bottom of his heart and whose national rebirth he had not only yearned for but witnessed? The detached observer is all the more likely to ask this question because it will seem to him that the new national kingdom did not do so badly in the thirty years that preceded Pareto's emigration. Besides progressing economically at a considerable rate and growing out of financial disorders—pace our Keynesians—it took its first steps in social legislation and established itself successfully as one of what then were called the great powers. Looking at things in this light, our observer will develop a good deal of respect for a régime such as that of Agostino Depretis. And, considering the difficulties incident to the beginnings of the new national state, he will make allowances for the less exhilarating parts of the picture. But Pareto made no such allowances. He saw nothing but incompetence and corruption. He fought with impartial ferocity the governments that succeeded one another, and it was then that he became known as an ultraliberal—in the nineteenth-century sense of uncompromising advocate of laissez-fire—and that he helped to create, among the German New-Dealers of that period, the impression that marginal utility was just a wicked trick with which to thwart reformers. Possibly this is all there is to say about Pareto's attitude in matters of economic policy and the strong traces it left upon his scientific writing before 1900. But even then there was something in that ultraliberalism of his that points in a direction exactly opposite to the creeds and slogans of official liberalism. He certainly was anti-étatiste, but for political reasons rather than for purely economic ones: unlike the English classics, he did not fight against government activity per se but against the governments of parliamentary democracy, of the very parliamentary democracy that commanded the fervent allegiance of the English classics. Viewed from this angle, his type of laissez-faire acquires a connotation that is entirely at variance with the laissez-faire of the English type. And once we realize this, the rest is easy to understand.

Toward the end of the nineteenth and during the first two decades of the twentieth century, an increasing number of Frenchmen and Italians began to voice dissatisfaction that varied from mere disappointment to violent disgust at the manner in which the cotillon of parliamentary democracy functioned and at the results it produced in France and Italy. Such sentiments were shared by men as different as E. Faguet and G. Sorel, and they were not confined to any one party. This is not the place to analyze let alone to pass judgment upon them. All that matters for us is their existence and the fact that the later Pareto stands out from this current of thought only because he himself stood out from his contemporaries and because he wrote a sociology that was—along with those of Sorel and Mosca—to rationalize it.

Englishmen and Americans, oblivious of the particular and historically unique circumstances that have developed in their minds an equally particular and unique attitude toward parliamentary democracy, have wondered about the possible meaning of Pareto's attitude toward Fascism. But this attitude is not problematical in the least. No theory is necessary in order to explain it. The events of 1914-22 had called him back to the arena of political debate. The masterly analyses he published on the origins of the First World War, on the miscarriage of Versailles, and on the futility of the League are among his strongest performances although they found no answering echo outside of Italy. But above all he witnessed with something like horror the social disorganization in Italy which it is necessary to have seen in order to believe. Attributing all the troubles of those years to the weakness of the political system of a decadent bourgeoisie, the student of Roman history may have thought of the formula by which, in republican Rome, the senate, in order to deal with an emergency, used to direct the consuls to appoint an officer of practically unlimited though temporary power, the dictator: videant cànsules ne quid detrimenti res publica capiat. But there was no such provision in the Italian constitution, and it would not have done any good if there had been one. So the dictator had to appoint himself. Beyond this and beyond approval of the success with which Mussolini restored order, Pareto never went. Mussolini honored himself by conferring senatorial rank on the man who kept on preaching moderation and who stood throughout for the freedom of the press and of academic teaching. But to his last day Pareto refused to embrace this ' ism' as he had refused to embrace any other. There is no point whatever in judging his action—or, indeed, any action or sentiment of his—from the standpoint of Anglo-American tradition.

Everything else is at the bottom of the pool.

II. THE THEORIST

Any appraisal of Pareto's contribution to economics must first of all give due credit to a feat of leadership. He never taught in Italy. The Faculty of Law in the University of Lausanne did not make very favorable headquarters for a campaign of intellectual conquest. The country house in Céligny looked like a buen retiro. Yet he did what Walras had not been able to do: he formed a school in the full sense of the word. An inner circle of eminent economists, a wider circle of followers of less eminence, and beyond this a broad fringe of more or less definite adherents emerged soon after 1900. They cooperated in positive work. They cultivated personal contact. They stood for one another in controversy. They recognized One Master and One Doctrine.

This school was specifically Italian. As has been pointed out already, there were but few foreign adherents, though individual pieces of Paretian teaching eventually gained acceptance both in England and in the United States. Nor did the Pareto school ever dominate Italian economics. No school ever does dominate its own country. Impressions to the contrary, e. g. the impression that the Ricardo school ever dominated English economics, are due to nothing but unrealistic historiography. Several other Italian leaders, like Einaudi, held their own ground entirely, and others, like Del Vecchio, while recognizing Pareto's eminence and adopting this or that of his doctrines, thought and wrote pretty much as they would have done had Pareto never lived. Still, there remains the fact that was a school did emerge on the basis of a theoretical structure that was inaccessible not only to the general reader but, in some of its most original parts, also to students of economics, students moreover who had never heard or seen the master.

But once we have duly recognized and thereupon discarded this feat of leadership, we see a theorist who continued the work of Walras. Nobody, of course, ever denied this, not even the most ardent disciple and, least of all, Pareto himself. Difference of opinion on this point is inevitably confined to the extent to which he surpassed the great pioneer and to the relative mental statures of the two men. There are several reasons why disciples will never agree on this either with outsiders or among themselves. One of these reasons must be noticed at once. Walras presented his immortal theory in the garb of a political philosophy that is extra-scientific in nature and, moreover, not to everyone's taste. I am afraid that there is no better way of conveying what that philosophy was than to call it the philosophy of petty-bourgeois radicalism. He felt called upon to preach an idéal social that hails from the semi-socialist French writers of the first half of the nineteenth century or, as we may say with equal justice, from utilitarianism. He looked upon the nationalization of land as an essential item in his teaching, and he was a monetary reformer whose plans have a strikingly modern ring. All this was gall and wormwood to Pareto. It was just metaphysical speculation and meta-physical speculation of a very unsympathetic kind. Their common ground was confined to pure theory and specifically to Walras' equations of equilibrium. But in every other respect they were as different as two men can be, and even their companionship in arms in the fight for mathematical economics and Pareto's obligation to Walras in the matter of the Lausanne professorship did not prevent their deep-seated mutual dislike from asserting itself or even from spilling over in conversation with third persons. While their pure theories are cast in the same mould, their systems of thought taken as wholes and their visions of the social process are not. And all those economists who are not disposed to neglect a man's philosophy and practical recommendations completely, that is to say the majority of the profession, will, for this reason alone, consider the Paretian structure to be something completely different from that of Walras.

In any case—we are neglecting sociology for the moment—it was, with one exception, in pure theory alone that he made scientific history. Let us note this exception first. In the Cours and also in a separate memoir of 1896 Pareto published a highly original pioneer achievement in econometrics that first established his international reputation and, under the title of ' Pareto's Law, ' created what may be fairly called a whole literature devoted to its critical discussion. Call N the number of income receivers who receive incomes higher than x, and A and m two constants; then Pareto's 'Law' asserts that

Chapter 7 of the Manuel contains Pareto's most mature interpretation of this generalization. We must confine ourselves here to noticing the two classes of problems which it raises. There is, first, the question of fit. Numerous investigations have been made, some of which were held by their authors either to refute the Law completely or else to establish the superiority of other methods of describing the inequality of incomes. The reader will observe that the central issue turns on the approximate constancy of the m. However, by and large, the ' Law' has stood fire rather well as the fact that it is sometimes used by competent statisticians even now suffices to prove. But there is, second, the question of interpretation. Granted that up to quite recent times the distribution of incomes according to brackets has been remarkably stable, what are we to infer from this? This problem has never been attacked successfully. Most participants in the discussion, Pigou among them, have confined themselves to criticizing Pareto's own interpretation—which, to say the least, was in fact open to objection at first—and, like so many of our controversies, this one has petered out without yielding a definite result. Few if any economists seem to have realized the possibilities that such invariants hold out for the future of our science. Viewed from this standpoint, Pareto's ' Law' is path-breaking in the literal sense even though in the end nothing whatever is left of its particular form.

I take this opportunity to dispose of another matter. In the Manuel, Pareto dealt with his ' Law' of Income Distribution in the chapter on Population. So far as the topics are concerned that are usually dealt with under this heading, this chapter does not contain much that would call for notice. But it contains a number of other things which, like the ' Law, ' are not usually included in the theory of population, and it is these items which liven up this chapter and impart to it its freshness and originality. Pareto's theory of the circulation of the élite is an example (see below, section III). Most of them are sociological rather than economic in nature, and some of them bring out sharply, indeed almost naïvely, certain prejudices that sat so incongruously upon the great analyst of human prejudice.

In the field of pure theory, properly so called, Pareto's thought developed slowly and in fact retained certain preParetian features to the end. In addition to the early influences of Ferrara and of the English and French economists of the ' classic period, ' he had Walras' equations of static equilibrium to start from—after having realized, not without considerable initial resistance, that they were in fact the keys to everything else. He was further stimulated by all the suggestions that no competent theorist could have helped receiving in the decade from 1885 to 1895. Finally, he was acutely aware of the technical shortcomings and other limitations of his immediate predecessors. Thus his own theoretical work was cut out for him—most of it, indeed, by Walras himself. But his earlier work, such as his' Considerazioni sui principi fondamentali dell' economia politica pura' (Giornale degli Economisti, 1892-3), never went beyond the range of the Walrasian signposts. This is also true, and emphatically so, of his Cours. Some economists who respected Pareto but were not strictly Paretians have paid him the dubious compliment of calling the Cours his masterpiece. It is, indeed, a striking performance enlivened throughout by a strong temperament that imparts sparkle even to conventional passages. But Pareto was right in refusing to sanction a reprint or a second edition. For, so far as pure theory is concerned, there was nothing specifically Paretian about it. It was only after 1897 that he rose to heights of his own. The first major publications that testify to his progress are the ' Sunto di alcuni capitoli di un nuovo trattato di economia pura' (Giornale degli Economisti, 1900) and the Résumé of his Paris course. The Manuale or rather, because of the appendix, the Manuel (1909), then marks the point of highest elevation that he reached.

The structure of the tower he erected on that spot is far from faultless. Many things that are essential in a comprehensive treatise received but scant attention. I do not mean merely that Pareto's work cannot stand comparison with Marshall's in those qualities that are ordinarily looked for in a ' manual. ' Much more serious is it that important parts of the theoretical Organon are inadequately thought out. Pareto's theory of money, for instance, is on the whole inferior to that of Walras. His theory of capital and interest derives all its merits from that of Walras. As regards interest he seems to have been content to rely for explanation on the fact that items of physical capital, hence their services, are not free goods. His theory of monopoly cannot, I believe, be salvaged by even the most generous interpretation. In spite of all this, the adverse judgment arrived at by some critics is completely wrong. For it neglects not only many individual strong points but, much more important, the essence of the achievement. The most important of those strong points, the theories of value and of production, will be discussed presently. But first we must try to define that achievement itself of which these two theories were but applications.

The first idea that must occur, from a purely theoretical point of view, to anyone who has mastered Walras's system is to raise it to a still higher level of generality. When we follow Walras and, indeed, all the marginal utility theorists on their progress through the phenomena of exchange, production, and so on, we discover that they are trying to solve problems that in ultimate logic reduce to one only: all their problems—not only the problems of production—are problems in the transformation of economic quantities and formally alike, the differences consisting merely in the different restrictions to which economic action is subject in different fields. Suppose we decide to do what we do in all the sciences, that is, to separate out the common core of all economic problems and to build a theory of this common core once and for all. The point of view of ' mental economy' (E. Mach's Denkàkonomie) will justify this endeavor to utilitarians. A theory of this kind will work with quite general indices, such as ' tastes' and ' obstacles,' and need not stop at the specifically economic meanings that we may assign to these words. We may transcend economics and rise to the conception of a system of undefined ' things' that are simply subject to certain restrictions and then try to develop a perfectly general mathematical logic of systems. Stretches of this road should be quite familiar to economists who have for generations used primitive devices, such as our venerable friend Crusoe, for the purpose of displaying certain features of economic logic. Pareto simply did the same on a much higher level and on a much broader front. But in these alti tudes it is difficult to breathe and still more difficult to gain ground. Critics as competent as the late A. A. Young have been of the opinion that Pareto achieved nothing but ' arid generalizations.' But only the future can tell whether this is so. Meanwhile we should recognize the greatness of the attempt.

An example will show that such a ' rush for generality' may produce not only logical stones but also economic bread, though it suffers from the weakness that it still moves on a relatively low level of generality and, indeed, dates from the Cours. As everyone knows, Marx's work is an analysis of the capitalist process, no doubt geared to the purpose of showing that this process will issue in a socialist society but entirely free from any attempt at blocking out the economics of this society. And there are a number of Marxist and Neo-Marxist contributions to the latter problem that can be described only as complete failures. As everyone also knows by now, the service to socialist doctrine that Marxist theories have been unable to render has been rendered by E. Barone, whose famous paper on the subject (' Il Ministro della produzione nello stato colletivista,' Giornale degli Economisti, 1908) has been surpassed by modern writers only in secondary details. But the essential idea of Barone's argument is clearly indicated in the second volume of Pareto's Cours and in his Manuel, namely the idea to lift, as it were, the logical core of the economic process above the ground of the institutional garb in which it is given to observation. The reader will observe how easily this idea suggests itself, as a special case, once we place ourselves on the standpoint of Pareto's general theory of tastes and obstacles, although it also suggested itself to Wieser.

In this special case, Pareto has nearly lost his claims to priority—at least among Anglo-American economists—although he had not only posited the problem but also pointed out the way to its solution. In other cases, he lost them completely because he confined himself to mere suggestions. Thus, assisted by hindsight, we may discern in the Manuel many pointers toward the later economic dynamics. However, none of them, such as his reference to a form of adaptation similar to a courbe de poursuite (dog-and-his-master problem, see e.g. p. 289) and to the presence of vibration continuelle, was put to any use other than the negative one of showing that the economic system's tendency toward a unique and stable ' solution' (i.e. a unique set of values that will satisfy its conditions) is a much more doubtful matter than the economists of that period, including Walras, supposed. No positive use was made of these suggestions, and no method was indicated for attacking these problems. I think therefore that we should not hesitate to describe Pareto's work as static theory, and that substantial justice is done if we add that he, more than others, was aware of its limitations and of the call of the problems beyond.

We now proceed to a brief discussion of Pareto's work in the fields of value and production keeping in mind that, from the standpoint adumbrated above, they really merge into a single theory.

Most modern theorists, although not all, will agree that the historical importance of the utility and marginal utility theory of Jevons, Menger, and Walras rests mainly upon the fact that it served as the ladder by which these economists climbed up to the conception of general economic equilibrium, although this conception was much more clearly perceived and much more fully developed by Walras than it was by either the Austrians or Jevons. In other words, the utility and marginal utility theory was one of several possible avenues to the thing that really mattered and, besides offering an excellent method for demonstrating in an easily understandable way the relations that hold the economic system together and, in fact, make a unified system out of the mass of economic phenomena which departmentalize so easily, had no great importance in itself. Or, to put it still differently, utility theory was an extremely useful heuristic hypothesis and nothing more. But neither Walras nor the Austrians were of this opinion. On the contrary, for them the utility theory was nothing less than ultimate truth, the discovery of the key to all the secrets of pure economics. In consequence, they placed an emphasis upon it that in turn induced Pareto and the Paretians to place undue emphasis upon their renunciation of it. Authors in the English-speaking world, particularly Professors Allen and Hicks, followed suit and very generously congratulated Pareto on what to them also seemed a new departure of first-rate importance. In fact, there is a widespread opinion to the effect that this new departure constitutes Pareto's main contribution.

There are indications in the Cours that Pareto was not quite satisfied with the Walrasian theory of value from the first. But his amendments, either insignificant or not original, remained within the precincts of the principle itself. Of the insignificant amendments, we merely mention the introduction of the term ophélimité in the place of the term utility (ophélimitéélémentaire for marginal utility or Walras' rareté) on the ground that the latter carried too many misleading associations. Of those that are not original with Pareto I mention the conception of utility and marginal utility as functions of all the commodities mat the consuming unit possesses or consumes in an appropriately chosen period of time, instead of Walras' conception of total and marginal utility of every commodity as function of the quantity ofthat commodity alone. This obvious improvement is due to Edgeworth, but I confess to some doubt whether Edgeworth was fully aware of the theoretical difficulties this improvement would cause, for it turns the final degree of utility that was simply an ordinary differential coefficient with Jevons, Walras, and also with Marshall, into a partial differential coefficient and this greatly increases the mathematical difficulties we encounter when trying to prove the determinateness of the economic system even in its most simplified form.

Before long, however, and certainly before 1900, the year in which he delivered his Paris lectures that made his change of standpoint publicly known, Pareto realized that, for his purposes at least, the concept of measurable utility (cardinal utility) might be safely abandoned or that, in any case, it would have to be abandoned for reasons that were first stated exactly in the second part of Irving Fisher's Mathematical Investigations into the Theory of Value and Price (1892). To save the situation he appealed to the indifference and preference curves that had been first introduced by Edgeworth. But, whereas Edgeworth still started from a measurable total utility from which he deduced the definition of these lines, Pareto inverted the process. He took the indifference lines as the given thing to start from and showed that it was possible to arrive from them at the determination of economic equilibrium in pure competition and also to proceed to certain functions which might be identical with utility if it exists. In any case, it was possible to obtain (ordinal) indices of utility or what Pareto called index functions.

I wish to bring out two points quite clearly. The first is that Pareto, though he may have adapted an invention of Edgeworth's to his own use, imparted to the indifference varieties a meaning that they do not carry in Edgeworth's Mathematical Psychics. They are quite divested of any utility connotation, and what the concept of utility had done for me theory of economic equilibrium was now to be done by certain assumptions about the form of these indifference curves. The new idea was to replace utility postulates by postulates about observable behavior and thus to base economic theory on what seemed to Pareto to be more secure foundations. It might be urged of course that in spite of several attempts nobody has as yet succeeded in carrying out such observations and that it is difficult to indulge in the hope that we might construct them from objective data in their whole extent so as to derive a complete empirical indifference map. Let us call them, therefore, potentially empirical or, to misuse a Kantian term, ' referring to possible experience.' In any case, their introduction for a purpose entirely foreign to Edgeworth's might be called a truly original achievement were it not for the fact that this achievement had been, as Pareto recognized, foreshadowed in Fisher's work mentioned above.

The second point is that Pareto's own argument brings out the difficulty he experienced in disentangling himself entirely from the old utility theory. He always kept an eye on the cases in which it might be possible to speak of utility and even of cardinal utility, the existence of which—hence the question of integrability—continued to interest him very much. And his index functions bear after all a pretty close similarity to the old concept. In fact, as has been pointed out by Allen and Hicks, he never succeeded quite in disentangling himself entirely, and he continued to use concepts such as the Edgeworthian definitions of rivalry and complementarity that do not go well with his fundamental idea. This fundamental idea, let us add, was developed and defended as early as 1902 by P. Boninsegni. By 1908, Enrico Barone, in the paper mentioned already, definitely went beyond Pareto by confining his fundamental assumptions in the matter of value theory to what he called the fact that, confronted with given prices of products and productive services, every individual distributes his receipts from the sale of his services between expenditures on consumption goods and saving in a certain unique manner ' of which we are not going to investigate the motives.' This, so he pointed out, does away with any concept of either utility or indifference functions. The rest of the story is too well known to detain us. I shall merely mention the papers of Johnson and Slutsky that for the time being remained practically unnoticed; the important reformulation of Bowley in his Groundwork that was more influential; and the work of Allen and Hicks, Georgescu-Roegen, Samuelson, and H. Wold. If we accept the present situations as ' provisionally final,' we must indeed salute either Fisher or Pareto as the patron saint of the modern theory of value.

But, still more definitely than patron saint of the modern theory of value is Pareto the patron saint of the ' New Welfare Economics.' The story of how, once more, he came to render a service to a cause with which he was—or would be—completely out of sympathy is not without its humor. From the very beginnings of economics, a loosely defined public welfare played a great role in the writings of economists. The familiar slogans of utilitarianism (Beccaria, Bentham) did something toward rationalizing the concept, and the utility theory of value seemed admirably qualified to implement it: in fact it was promptly harnessed to the task, e.g. in the field of taxation. The Fisher-Pareto theory of indifference varieties, destroying as it did the bases of arguments that worked with cardinal utility or even with interpersonal comparison of utility (satisfaction), should, so we might think at first blush, have done away with all this. But instead of drawing this conclusion—and in spite of his contempt for the political humanitarianism of our age—Pareto immediately went on to attack the problem of maxima of collective satisfaction afresh. The definitive formulation was left for Barone, but the main idea is again Pareto' s. He observed, first, that all changes imposed upon any given economic pattern may be said to increase welfare or collective satisfaction in a perfectly objective sense if those who gain in terms of numéraire could compensate those who lose in terms of numéraire and still have some gain left. This criterion will in fact salvage some though not all of the welfare judgments usually passed by economists. Second, Pareto pointed out that welfare judgments that cannot be salvaged in this manner must be explicitly based on extra-economic, e.g. ' ethical,' considerations. And third, he showed that the criterion may be used in order to establish that l'état collectiviste may improve upon the level of welfare that is practically attainable under perfect competition. But, barring developments, these points are pretty much what the New Welfare Economics amounts to.

That part of Pareto's welfare economics which deals with the logic of production provides a convenient transition to his second great contribution to pure theory, his theory of production. Approaching the problem from the side of the theory of choice and applying to the producer's case the general apparatus of indifference curves and derivative concepts (lignes du plus grand profit, lignes de transformations complètes et incomplètes etc.), he sketched out a comprehensive structure only parts of which are explicitly present in the literature of his time and which may be said to constitute the foundation of the mathematical theory of production of our own age or, at all events, of its statics. In particular, its very generality leaves room for all the special cases that we may wish to treat without placing exclusive emphasis on any one of them: the ' obstacles' may be anything at first, and can then assume any of the forms that occur more commonly in practice—the factors that are required in fixed quantities irrespective of output, the factors that are required in technologically determined quantities per unit of output, the ' compensatory' factors, and so on, all take their places in a theoretically complete schema of possibilities. In appraising this achievement, we must keep in mind that Pareto was primarily concerned with generalizing and otherwise improving the work of his great predecessor. Again his work may be divided into a first part that culminated in the Cours and a second part that culminated in the Manuel, though some minor touches were added in the article in the Encyclopédie des Sciences Mathématiques (Volume 1, 1911).

Originally, Walras had expounded his theory of production on the assumption of fixed coefficients of production—fixed (average) inputs per units of output—not because he believed that mis was the only or even a very important case but because he thought himself justified in adopting what he considered to be a simplification. His reply to private criticisms that poured in on him was that ' the economists who will come after me are free to insert one by one all the complications they please. They and I, so I think, will then have done everything that it was our duty to do' . So far as this goes Pareto cannot be said to have done more than take Walras' advice. In addition, when the Cours appeared, Walras had already introduced variable coefficients, on a suggestion of Barone's that reached him in 1894, though without altering the argument of the fundamental section on production. In the same year (1894) appeared Wicksteed's Essay on the Coordination of the Laws of Distribution. Finally, variable coefficients of production were no novelty in any case after all that Jevons, Menger, and Marshall had said on the subject. Pareto's Cours added only an elegant formulation and a number of reasons—not all of them convincing—why the case of compensatory coefficients should not be treated as the only or the fundamental one either.

It is of course a mere matter of terminological taste whether or not we are to confine the phrase ' marginal productivity theory' to this case. Pareto did so confine it and, in the years following upon the publication of the Cours, grew increasingly hostile to it, declaring it definitely ' erroneous.' He was evidently under the impression that he had refuted or, at any rate, outgrown it in the same sense in which he felt that he had refuted or out-grown the marginal utility theory. His brilliant theory of cost—which, among other things, withdraws from their dangerously exposed positions the textbook theorems that, in perfect equilibrium of pure competition, price should equal marginal cost and total receipts should at the same time equal total cost—permit us to test this claim. So far as productive combinations depend on economic considerations—and, after all, it is the economic considerations which it is the economists' task to clear up—the difference, as compared with straight marginal productivity theory, is not great. But Pareto does teach us how to handle the deviations from it that technological and social restrictions impose. And, here as elsewhere, he does something else: he always points beyond himself.

III. THE SOCIOLOGIST

There is nothing surprising in the habit of economists to invade the sociological field. A large part of their work— practically the whole of what they have to say on institutions and on the forces that shape economic behavior—inevitably overlaps the sociologist's preserves. In consequence, a no-man's land or everyman's land has developed that might conveniently be called economic sociology. More or less important elements that hail from that land are to be found in practically every economic treatise or textbook. But beyond this many economists, and especially those who define economics proper rather strictly, have done sociological work. A. Smith's Moral Sentiments and Wieser's Gesetz der Macht are both out-standing instances of a large genus. But few if any men in the list of great economists have devoted so large a part of their energy as has Pareto to what at first sight seems to be an extra-curricular activity, and few if any owe so much of their international reputation to what they have done in that field. But his achievement is not easy to characterize and to appraise. The enthusiastic applause of some and the hostility of others are both understandable but neither can be taken quite seriously because the non-scientific sources of both are painfully obvious in most cases. Although several minor works and a large number of newspaper articles would have to be considered in order to give a satisfactory picture, we need not go beyond the Systèmes socialistes, the Manuel (especially Chapters II and VII), and the Trattato di Sociologia Generale.

Let us begin with two aspects of Pareto's sociology that are perfectly obvious and the reverse of difficult to characterize. First, although Pareto the economist touched upon a large number of extremely concrete and practical problems throughout his long life, his purely scientific contribution is in the realm of the most abstract economic logic. It is, therefore, quite understandable that he should have experienced a wish and, in fact, a need to erect alongside his pure theory another building that would shelter facts and reasonings of a different kind, facts and reasonings that would do something toward answering the question how the elements taken care of by his economic theory might be expected to work out in practical life. Second, we have seen that in his earlier days, at least as long as he lived in Italy, he had taken a passionate interest in the debates on questions of economic and general policy. The born thinker that he was must have been struck by the impotence of the rational argument, and the question must have intruded upon him of what it really is that determines political action and the fate of states and civilizations. Again, it is quite understandable that, so soon as he had settled down to a life of thought, this question should have emerged from the sphere of easy and superficial answers that all of us are prone to give when immersed in our daily work, and that he should have attempted to raise it to the plane of scientific analysis. This amounts to saying that primarily and fundamentally his sociology was a sociology of the political process. Of course, everything that man does or thinks or feels and all his cultural creations and his attitudes toward cultural creations are bound to come in somehow or other when we think about the political process which then becomes but a special case. But it was this special case which fascinated Pareto and for the sake of which he erected and adorned a much larger structure.

Next, still moving on ground that is relatively easy to survey, we shall consider his method. Pareto himself emphasized again and again that he simply applied the same ' logico-experimental' methods that had served him for the purposes of economic theory to the task of analyzing the ' experimentally' verifiable reality of other aspects of social life, allowing himself to be guided here as elsewhere by the example of the physical sciences. This was, of course, a complete delusion. It is easy to observe, for instance, that he made large and in part illegitimate use of psychological interpretations for which there is no analogy in the physical sciences and that his material, such as it was, was the product of observation and not of experiment—a difference which is fundamental from the standpoint of method. I am afraid that what he really meant to emphasize when trying to formulate his rules of procedure was simply the detachment of the philosopher who does not identify himself with any party, interest, or creed. The possibility of such detachment raises, of course, a very well-known fundamental difficulty and one that Pareto was the less qualified to overcome because he failed to see it. Actually he used two different analytic schemata: one that may be called a morphology of society and does invite the use of facts that are, potentially at least, amenable to observation as are the facts of anatomy or biology in a similar sense; and another that pertains to social psychology. Bom schemata are indeed illustrated or even, to some extent, verified by historical and contemporaneous instances, but neither is derived from them by anything like a ' logico-experimental' method: both are reflexes of a highly personal vision of the social process that owes much to Pareto's background, practical experience—and resentments. The affinity of the morphological schema with Darwinian selection and of the socio-psychological schema with parts of the teaching of Tarde, Durkheim, Lévy-Bruhl, and Th. Ribto is obvious. Still more so is the relation of both with the current of thought glanced at in the first section of this essay that issued in derogatory criticism of the doings of parliamentary democracy—the current that was anti-intellectualist, anti-utilitarian, anti-equalitarian and, in the special sense defined by these terms, anti-liberal. But the force of the man created from these materials something that was nevertheless specifically his own.

The morphological schema centers in the proposition that all societies consist of heterogeneous masses of members—individuals or families—and are structured according to the aptitudes of these members for the relevant social function: in a society of thieves, the ex hypothesi widely varying ability to steal would determine social rank, and hence influence upon the government of the society. Pareto seems to assume that these abilities, while capable of improvement and of decay, are substantially innate, though he makes little effort to establish this. Moreover, though distributed continuously in the population, they lead to the formation of classes, the ' higher' ones of which have and use the means of buttressing their position and of separating themselves from the lower strata. In consequence, there is in the lowest strata a tendency to accumulate superior ability that is prevented from rising, and in the topmost stratum, in the aristocracy or élite, a tendency to decumulate energy through disuse—with resulting tension and ultimate replacement of the ruling minority by another ruling minority that is drawn from the superior elements in the couches inférieures. This circulation des élites does not, however, affect the principle that it is always some minority which rules, and does not do anything to bring any given society nearer to the ideal of equality, though it does produce equalitarian philosophies or slogans in the course of the struggles that ensue. With a turn of phrase that recalls the first sentence of the Communist Manifesto, Pareto proclaimed that history is essentially a history de la succession des aristocracies. But his presentation of this part of his argument is so very sketchy and he leaves his readers with so much to interpolate that I am not at all sure that I have rendered justice to his thought. Nevertheless, I had to make the attempt. For some such argument is necessary in order to put his social psychology into its proper light.

The socio-psychological schema centers in the concept of the non-logical (not necessarily illogical) action. This concept recognizes the well-known fact—well known, in particular, to economists—that the great mass of our everyday actions is not the result of rational reasoning on rationally performed observations, but simply of habit, impulse, sense of duty, imitation, and so on, although many of them admit of satisfactory rationalization ex post either by the observer or the actor. So far there is nothing in Pareto's psycho-sociology that could be unfamiliar to anyone. What is unfamiliar, however, is his tremendous emphasis upon the additional facts that a great number of actions—and let us add at once, beliefs—are being rationalized, both by actors and by observers, in ways that will not stand up under scientific analysis and, more important, that some actions and beliefs are altogether incapable of being rationalized in any way that will. The importance of this second step for a sociology of the political process becomes obvious if we take a third one: Pareto maintained that the large majority of all the actions and beliefs that make up that process are of the type mentioned last. Take, as an instance on which we all agree, the idea of the Social Compact or, as an instance on which most of us agree, Rousseau's theory of the volontégénérale. Only, according to Pareto, practically all the actions, principles, beliefs, and so on prevailing in the collective mind of electorates belong in the same category. And a large part of the Trattato consists in illustrating this, often amusingly, sometimes instructively.

It will serve our purpose to put this point strongly, more strongly than Pareto himself ever put it. The masses of thought and the conceptual structures that form the conscious surface of the social and in particular of the political process have no empirical validity whatever. They work with entities such as liberty, democracy, equality, that are as imaginary as were the gods and goddesses who fought for and against Greeks and Trojans in the Iliad, and are connected by reasonings that habitually violate the rules of logic. In other words, from a logical standpoint, they are nonsense unalloyed. This makes a political philosophy that is best described by its diametrical opposition to that of Jeremy Bentham. It should be observed, however, that this diagnosis of the political myths (Sorel) did not induce Pareto to overlook the function that this logical nonsense may fill in national life. After having gone through with an analysis that is severely positivist in nature, he refused to draw the conclusion that would seem the obvious one to the positivist. While political creeds and social religions—with Pareto there is very little difference between these two—contribute to dissolution in dissolving civilizations, they also contribute to effective organization and action in vital civilizations. This is a very curious attitude for a thorough-going positivist to take and will perhaps be cited at some future time as an outstanding example of the mentality of an epoch that destroyed one type of metaphysical beliefs while ushering in another. It reminds me of the advice which I have heard some psychoanalysts give to some of their patients, namely the advice to cultivate with a view to possible remedial effects a sort of synthetic belief in God. There is of course no contradiction between maintaining that social and political creeds have no empirical significance and admitting that some of them may make for social cohesion and efficiency. But the social philosopher who should thereupon undertake to advise the adoption of the latter would run into the same difficulty as our psychoanalyst: so long as his analysis is being accepted his advice must be ineffective, for no synthetic God can be trusted to help; so soon as his advice is accepted his analysis will have to be rejected.

That tissue of creations of our imagination Pareto called dérivations. The argument adumbrated in the preceding paragraph abundantly shows that they are not without importance as factors that help to shape the historical process. It was Pareto's opinion, however, that this importance is relatively small and that substantially these dérivations do no more than verbalize something more fundamental that comes much nearer to determining actual political behavior and the sum total of non-logical actions. Now if we defined this more fundamental something in terms of group interests, and if we then went on to define these group interests in terms of the social location of groups within a society's productive organization, we should be, to say the least, very near Karl Marx's view of the matter, and there is in this point actually a strong affinity which I think it important to emphasize. In fact, if we adopted this line of reasoning, there would be only two major points of difference left between Marxian and Paretian political sociology. On the one hand, Pareto introduced explicitly an element that is only implicitly present in the Marxist analysis: the importance for the explanation of an actual stretch of history, of the greater or smaller degree of social flexibility that a given society displays, or, in other words, the importance of the fact that there exists an optimum or vertical mobility and of resistance to it that will better than others guarantee what might be termed stability of political change. On the other hand, we need only recall our sketch of Pareto's social morphology in order to realize that with Pareto the historical process is not so much the result of the conflict of comprehensive social classes as it is the result of the conflict of their ruling minorities. It is submitted that, while both differences are to the credit of Paretian sociology, they do not amount to more than corrective improvements upon the Marxist schema. I might add the fact that property relations per se are much less in evidence with Pareto than they are with Marx, and that this also constitutes a claim to superiority of the Paretian analysis. But it will be readily seen, that this point is really implied in the other two.

Actually, however, Pareto did not follow up this line of analysis. With him the link between the tissue of delusions which he called dérivations and the objective determinants of actual behavior was supplied by what he called the résidus. I am conscious of the danger of being unfair if, for the sake of brevity, I define these résidus as impulses generally found to be present with human beings that revive, and not in a very inviting manner, the old psychology of ' instincts.' We need not discuss the list that Pareto drew up—and which contains such items as an instinct of combinations, the sexual impulse, and so on—especially as Pareto himself does not seem to have been very satisfied with it. It is sufficient to point out the obvious methodological objection to any such procedure; even if Pareto's résidus and the ' laws' of their association and persistence were much more satisfactorily analyzed than they are, they would still be labels rather than solutions of problems, and call for professional investigation of a kind for which Pareto lacked the equipment. It is therefore quite understandable that Pareto's work has exerted so little influence upon professional sociology and social psychology, and that professional sociologists and social-psychologists have but rarely displayed a sense of the greatness of the structure as a whole.

But those and other shortcomings are not decisive. Pareto's work is more than a research program. Also, it is more than mere analysis. The fundamental principle that what individuals, groups, and nations actually do must find its explanation in something much deeper than the creeds and slogans that are used in order to verbalize action, conveys a lesson of which modern men—and none more than we economists—stand much in need. We are in the habit when discussing questions of policy of accepting at face value the slogans of our own and, indeed, of a by-gone time. We reason exactly as if the Benthamite creed of the eighteenth century had ever been valid. We refuse to realize that policies are politics and to admit to ourselves what politics are. We cultivate the subnormal and do our best to suppress whatever there is of strength and sparkle. In conditions such as these, Pareto's message, however one-sided, is a healthy anti-dote. It is not, like his economics, a technical achievement of the first order. It is something quite different. It is an attempt to preach a sermon.