Robinson 2
James Buchanan
Clark, J. B.
Taylor, Harriet

Presidential addresses

Erich W. Streissler


Not At All A “Slightly Depraved Entertainment”. 

The optimist proclaims that we live in the best of all possible worlds; and the pessimist fears this is true.
James Branch Cabell, The Silver Stallion

Query: For “the pessimist” may one not read “the historicist”?

An economist is to cross an estuary from one shore to the other. Should he just swim across? But there might be sharks. Should he hollow out a log and use the dugout as canoe? But the waters of the estuary may be too choppy; in fact, there may be breakers and going in a canoe may be at best a very wet business and at worst fatal. Should he go by sail boat? But then there may be sudden squalls and the boat might capsize. Should he finally take a great detour in time and cost and build a bridge across the estuary and then just walk over it? Better still, drive across to his destination on the other shore in a fast car, which of course would require a very costly road surface on the bridge? And what will the bridge with a road of one type or another cost in the end? Will the pylons subside because the ground is not firm enough? Perhaps, as a final question after the seemingly final one: would it be even better to build an airport and fly across? 
I am sure you expect a witty answer to our economist’s decision problems. Actually there are two answers – and the problem I set is not at all a joke but rather one of those old-world morality questions or rather a simile.
The first of the two answers is the neoclassical one. The answer is easy and predetermined: You choose the most modern solution. Possibly you have to dither and have an infight about what is the most modern solution: building the road bridge for fast cars or building the airport and flying across. But one of these it has to be. Now this typically neoclassical solution may be quite correct, but only under three conditions. Or, more precisely, under at least three conditions; for it is beyond our understanding as economists to specify precisely all the conditions under which our conclusions will fully hold. Condition one: The road bridge or the airport – and thus the neoclassical paradigm – will only be the correct solution if not only we and a few of our friends and pupils will cross to the other shore, but rather a vast throng of “superfluous economists”, as Joan Robinson so aptly called them – using the term “superfluous”, of course, “in the Shakespearian sense”(1). Condition two: Physically speaking, our estuary must be basically the same as all other estuaries where car bridges or airports have been built so that it is once more the best solution. And condition three says that it does not matter whether we wait for the solution – a few years or so, until either bridge or airport is built. In other words: The technically most complicated solution presupposes perfect homogeneity of the conditions of time and place, and that over a considerable stretch of time. 
This simile is already an explanation for an important precondition of economic analysis: American economics is largely neoclassical; or at least much more so than European economics and other non-American traditions. This is to be explained exactly along the relativistic lines of the History of Economic Thought: In Europe we have much more to do with as yet unanalyzed problems substantially shaped by the historical conditions of our respective small countries and, furthermore, there are far fewer economists around to do the analysis. Our less firm beliefs in the ubiquitous correctness of neoclassical economics are themselves historically predetermined, and rightly so.
Those of you who do not much like neoclassical economics may heave a sigh of relief to hear from me that I consider neoclassics only as a – historically – “very relative truth”, although you will not like to hear that, to me, your various favourite types of economics equally appear only now and then as the best tools. But I am willing to listen to each and every type of economic argument, and so should you. Let a thousand flowers bloom – each, though, on the soil best suited to it.
I turn to the second possible answer to my problem. The question posed by the historian of economic thought is quite different from the preconceptions of the neoclassicist. The historian of thought does not ask what is the most modern way – and usually also the most costly way – to get across the estuary. He poses the true economic question: What is the most efficient way to get across? If there are actually no sharks in the water, if the water is warm and the sea calm, probably the most efficient and, of course, least costly way to cross the bay is simply to swim, particularly if you are a good swimmer. Paradoxically, it is opting always for the neoclassical method- probably, because it will receive the highest applause by the unthinking others and is most likely to get our scientific contribution accepted in a professional journal – paradoxically, I started to say, it is opting always for the neoclassical method that is the uneconomic solution for the problem of finding the best explanation for an economic question. Invariably favouring the neoclassical option is an instance of the technological fallacy, the mere cult of the modern, against which the economist is taught to rebel. Opting for a review of the history of economic thought would be the economic (qua efficient) solution. This entails examining above all the preconditions for solving a problem before solving the problem itself.

I have to confess that I wrote about Adam Smith because I found him a fascinating character; I wrote about the mercantilists in order to understand better what Adam Smith thought or rather why he would wilfully misquote mercantilist authors; I wrote about Karl Marx at a time when my students were particularly interested in Him – with the, to me, unexpected and positive side effect that my students no longer dared discuss Marx with me because I knew Him too well. I started to write about the Austrian School when I was called to the main chair of that former School, in Vienna, now 34 years ago; and then when I was asked to write something like the tenth article on the Austrians I drifted into a completely unexamined field of inquiry, the question why the Austrians of the first two generations quoted so many of the German authors so much, while later generations did not. I found the Germans of the two middle quarters of the 19th century to have been completely neglected by researchers. Being elderly by now, I shall confess my grave sin: As I have an interest in history as such, for me historical research may have been in part what Kenneth Boulding once called “a slightly depraved entertainment” (2). But why should not a scientist sometimes do what he finds entertaining when there is a decided demand by the scientific community for such efforts? After all, all that I wrote was published. 
But then, this talk is called: Not at all a “slightly depraved entertainment”. For I wish to plead with you – from long experience – for the overbearing usefulness of such an historicist approach. I shall plead that studying the history of economic thought increases one’s skill in answering the really NEW and therefore vital questions in economics, the questions for the answer of which, because of their newness, there is a high intellectual premium. 
As historians you will find that new questions are frequently just old problems in new garb, problems which have been solved once before – a hundred, two hundred or three hundred years ago; or even a mere twenty years ago when the (now) Young, who tend to think everything is completely different from what their parents experienced, were not yet grown up.
Permit me to state that one of my greatest public successes was a mere calculation in 1990, according to neoclassical growth theory (of the 1950s), of the amount of capital that would be necessary to bring Eastern Germany, newly acceding to the West, up to an income level close to that of the West (3). Nobody believed me at the time – in fact, a junior (West) German minister was sent to Vienna to contradict my conclusions. For nearly all Germans had convinced themselves that when they wanted a miracle to happen a miracle was bound to occur. 
Economists, however, are taught that miracles are rare, and history should prove it definitely to them. The history of economic thought may be a pleasure to some. But all the same it is a very useful, in fact a necessary endeavour. This I have found repeatedly to be true. And what better can an economist say about any endeavour than to praise it on utilitarian grounds
I have to confess something else: By training, I am a lawyer or rather a jurist, and sometimes I have a suspicion that I am more of a lawyer than of an economist. I hold the oldest chair in economics in the German speaking area, a chair founded in 1763, and, of course, this chair was originally in the Vienna Law School and was still there when I, a Doctor of Law, was called to it in 1968. Legal studies are strongly historicist and typically not monistic in method, much rather multi-paradigmatic and thus in contrast to the prevailing method of American economics. (Being a lawyer, i.e. trained to see the different sides of a case, I hasten to add that the highly honorific Clark Medallists of the American Economic Association are by now frequently non-neoclassical thinkers, the typically legal-minded Andrei Shleifer of 1999 being a case in point; and even North American Nobel-Prize winners often show a strong non-neoclassical bent. Only the run of the mill American economist tends to be neoclassicist pure and simple.) Anyway, as a trained and occasionally also a practising lawyer I am much more historicist and certainly much more relativistic in my methods than many “mere” economists. Adversarial lawyers know that the truth is complex and that to employ exactly the opposite methodological strategy from one’s adversary may lead to success. Taking the opposite tack from your adversary may often be truth-serving because there is a chance that your opponent, for reasons of his own, follows a common rut of argument which, by now, has lost relevance.
In Austria, the common economics behind many legal practices is either mercantilist or what one might term Bourgeois Marxist, a kind of vulgar Marxism practised in the interest of the privileged. Under these circumstances, a neoclassical argument may cause surprise and be successful because it reveals an unsuspected truth, as I have very recently found to my personal advantage. It is a deep historical insight that in one country or in one field of intellectual practice neoclassical arguments may be extremely old hat and rife for replacement while in another field of social argument the neoclassical revolution may be just dawning. As a social phenomenon intellectual life is varied and diverse, and the deepest historical truth is: “It all depends.” Historicism tells you that in every single case you have to search for the right method at the right time.

When I regularly tell my law students at the beginning of their economics course that I am going to present them only with wrong arguments they tend to snigger. Probably, they were already convinced beforehand that all of economics is only lies, anyhow. Actually, pointing out that all economics is necessarily wrong is just the opposite from telling lies. For lies are purposive falsehoods told against your own conviction and in order to mislead.
All economic reasoning is false in the sense that a complete analysis of all the circumstances relevant in a certain social setting is beyond our human capacity. “The Theory of Complex Phenomena” is the title of the article which Hayek dedicated to Popper in the latter’s Festschrift (4). There Hayek argued that every social phenomenon is at least as complex as the most complex brain at work in this social sphere and probably even more complex than any constituent brain. Taking it as an axiom that the instrument of analysis always has to be more complex than what it analyses, Hayek concludes that it is impossible for any human brain ever to understand any social phenomenon completely. Following Hayek I conclude that economists have to remain mere approximators.
All our social theories are only more-or-less correct and therefore strictly speaking always false; and economics is merely a part of social theorizing. Economic phenomena are far too complex to be grasped fully. We have to use simplifications of thought or, in other words, models. A model is a simplified analytical structure of our mind, constructed for a given analytical purpose. As a mere simplification it certainly cannot explain everything but, we hope, it will elucidate the particular question in which we happen to be interested. Being simplifications all models always have to be “ad hoc” constructions. 
It is a grievous fault of a certain type of economic thinker – and, I admit, in particular of that kind of economist whom I would call a methodologically absolutist neoclassicist, i.e. the man who believes that the same economic conclusion has to be valid for all times and all places – it is a grievous fault of this man to condemn a model he dislikes for its being “ad hoc”. No, to be “ad hoc” is actually the best thing that could be said of a model: It means that the model elucidates the very economic phenomenon for which (i.e. “ad hoc”) it has been built. There can be more or less appropriate models but all have to be “ad hoc” An economic theory can only be more or less correct and always only so at a specific time and a specific place. All economic truth is in that sense merely historical, and never a full or complete history either, but an historical simplification. 
A good economic model will come as close as possible to the truth; nevertheless and strictly speaking it will always embody much falsehood as it cannot be correct always and everywhere. Actually, I am saying nothing that is not embodied in standard econometric methodology: The errors-in-equations explanation of a linear econometric structure explains the coefficient of determination R2 as the amount of explanatory power of a given quantitative model. Econometrics and absolute methodological neoclassicism are in full contradiction to each other. The historical method, as far as it is quantitative, is nothing but the best choice of an econometric model. This, actually, is what the historical method, as far as it was quantitative, did mean for the founder of the Older German Historical School, Wilhelm Roscher. He pursued important quantitative research and found, e.g., that in his time British firms were much larger than German firms and thus in a given industry which showed increasing returns to scale were more cost-effective, i.e. cheaper producers. 
The historical method thus demands a search for the best explanation imaginable which, however, can never be a perfect explanation. As far as measurement is concerned and as far as one can be more or less certain that a relationship is sufficiently close to a linear one (at least a monotonic one), econometric modelling is equivalent to historical search: Econometric search is the search for a model with the highest corrected R2. Basically, this is an historical exercise. In both cases the historically minded economist has to add a further, and the most important, qualification. Theorizing from past experience can only be correct if the future may be taken to be drawing from basically the same probability system as the past was. As to this continuity assumption we can never be certain, of course. Over the economists’ hell the words are written: panta rhei. Whenever he prescribes for the future even the historian cannot overcome the analytical curse of panta rhei. A major eruption of the volcano on the Aegean island of Thera (or Santorini) brought Minoan culture on Crete to its end. There was no use then in planning for a future.
Mark Blaug has recently reminded us of the treatise by Philip Miroswki, More Heat Than Light – Economics as Social Physics: Physics as Nature’s Economics, 1989. The idea that economics is – or should be – a kind of social physics stems from early 19th century French social philosophy and has taken root once more in late 20th century America, as one American tradition of thought has always had a strong affinity to French Revolutionary scientific postulates. In an involved argument Mirowki shows that economics is based imperfectly on the physics before 1860, but contradicts basic tenets of modern physics. However: So what? 
It is, indeed, true that leading absolutist neoclassical economists try to make one believe they are just practising another type of physics. The famous German physicist Max Planck once told Keynes that at first he had wanted to study economics, “but had found it too difficult” (5). So he opted for something simpler and took up physics. Physicists in general consider this story a good joke while economists know – or, if they do not, should know – that the anecdote contains a correct evaluation. Or, to let Hayek put the difference between physics and economics in a nut-shell: Physics is the description of everything which is easily describable. Leaving metaphorics: Physics seeks a complete description of restricted problems while economics seeks an approximate analysis of highly complex phenomena. Now, why should good approximations and complete descriptions ever use the same methods and concepts? To find a good approximation one needs historical vision; it is historical problem solving.
At most, economics is in some of its applications closer to engineering (6) than to theoretical physics. Engineers use many approximating calculations, even though one might call them applied physicists or chemists. And they approximate because the problems they deal with are far too complicated to permit of exact calculation. All this is even more true of economics where the complexity is vastly greater than in the physical world – not least because the economic world is changing, and often rapidly so. Actually, even our quantitative statements are just informed guesses within an historical context. 
Economics is fundamentally different from physics because it is a historicist subject geared to finding not perfect but in a sense makeshift best explanations, best suited to the problem in time and place. This is not least true for econometric modelling which only superficially resembles the equations of physics. While in physics, since Gauss, the error term in an equation has usually been explained by the human imperfection of measurement, by errors-in-measurement, in econometrics the error term is explained by the very approximate truth of the theory in question, by the so-called errors-in-equations explanation. No physicist would dare present to his colleagues an estimated equation with the frequently low R2 or “explanatory power” of econometric exercises. Thus, methodologically, in its deep structure, econometrics is historicist.

Let me give a few practical illustrations from my own main fields of work. At present these are in monetary economics and international economics. As illustrations I shall use three examples from the latter field. These illustrations touch questions of considerable practical interest to present-day Europe.
(i) My first example is the theory of convergence created to explain, above all to forecast, the likely development within the European Community. When the European Community was founded careful estimation showed that utilising comparative advantage between the initial member countries to the full would raise national incomes by a mere one half of one percent. In hindsight this very low estimation was not implausible because the original EC countries were rather similar countries, especially as to their stages of development. 
Enter the Second Historical Law of Economic Policy which says that whenever an important politician is faced with an economic conclusion which does not suit him, he will find an economist who comes up with a more convenient theory. (Please note that I use the word “he” in full political correctness, because I wish to make a nasty dig at certain types of persons who, by politeness, cannot be female.) If we are interested in the truth this Second Law, however, is often in conflict with the First and Most Basic Historical Law of Economic Policy, which says: Most politicians are wrong most of the time, and some politicians are wrong all of the time. In this case, though, many economists assumed on their own that the politicians were in fact right in surmising that the advantage of European Union would not only be an additional one half percent of GDP. So, economists came up with the theory of convergence which basically said: Due to economic integration all countries will develop more rapidly over time, but those with the lowest levels of income per capita will develop most rapidly. Thus, countries will converge in their GDP levels per head. Now this theory – perhaps I should rather call it this conjecture – proved right over time both for the original members of the European Community and, in particular, to an astonishing degree, for the newly acceding members Spain and Portugal. I am sorry to report that, so far, convergence has not been showing in Greece, the relative average income level of Greece having dropped slightly in relation to the EC average. So there is an exception, so far, and, unfortunately, it is exactly our kind host country. But possibly national income accounts do not correctly measure psychic welfare and it may be that what really matters for some Greeks is that they are “in” while the Turks are “out” and, at least within the EU, Greece was spared the financial convulsions Turkey recently went through. 
So let us return to serious economics: The theory of economic convergence within Europe has been empirically quite successful – with, as I pointed out, just one exception. But it was not derived within any of the great economic paradigms. It is, in fact, merely an historical generalization. I know, of course, that you can derive convergence from theoretical models – one of the best, to my mind, being the model which the doyen of German economics, Wilhelm Krelle, presented on the occasion of receiving an Honorary Doctor’s degree at the University of Vienna (7).Krelle’s model develops mathematically the very plausible idea that countries will converge if they are close enough at the outset and diverge if they are sufficiently different then. The question is, however: When will the latter be the case, when, exactly, will they be close and when too different? My revered elder friend, the Hungarian Belá Csikós-Nagy, has developed the ingenious idea that former Yugoslavia had to develop worker self-government because this simply meant that wages for the same type of work could be different within Yugoslavia and, with relative productivity levels differing to the extent of 1:7 within the very heterogeneous country, the Yugoslavs knew that wages would never be able to converge. History has proved Csikós-Nagy right at the most drastic level: Yugoslavia disintegrated. The theory of convergence is only very conditionally true. 
Why am I discussing all this with you? Because in the forthcoming process of integration of Eastern European countries into the EU the differences in average productivity levels will be vast, about the Yugoslav 1:7. In several instances, they will certainly be much larger than 1:3, which some consider the border-line for likely integration. We have no convincing theory that economic convergence will occur once again. History need not repeat itself if initial conditions are sufficiently different. Thus, in trying to forecast, economics is at a methodological impasse. We have a mere historical presumption of the possibility of convergence without either relevant antecedent empirical experience or compelling theoretical argument. And if the politicians believe all is going to be well in the best of all possible worlds – well, they have often done so without good reason since Voltaire’s Candide. Historicism must counsel doubt.
(ii) My second example – if I had had the time – would have had to do with the deservedly great success of Ricardo’s trade model of comparative advantage. I shall cut it short and merely present the conclusion: This very successful model assumes that there will be no factor movements
But in the forthcoming European enlargement labour movements may very well prove important. It is not at all clear whether with open labour markets the most highly qualified workers of the newly acceding and mostly very poor countries will not migrate to the richer “older” member countries. Because of the complementarity of other supportive factors in a highly developed economy incoming workers could achieve a much higher productivity in the already highly industrialized countries of Europe and thus much higher wages than they would earn at home. Thus migration may be to the mutual advantage of both employer and employee. On the other hand, the newly acceding countries may be swamped with cheap mass-produced commodities from the West, not least in agriculture. The result might be a migration of the best to the West and increasing unemployment due to commodity competition in the outcompeted Eastern economies, as one of the best-known German labour economists has forecast, namely Michael Burda, who is himself a migrant to Germany from the USA (8). So instead of economic convergence economic divergence might take place in spite of integration. This would create very serious political problems. Remember, the Ricardian model has been a lucky winner rather than an always correct model, and its luck might just run out.
(iii) My third and last example is taken from exchange rate theory, at present my special topic of interest. Now here the economist is faced with two absolutely convincing theories which lie at the very paradigmatic heart of neoclassical analysis, to wit: the arbitrage principle. The first is the theory of purchasing power parity, apparently already known to Ricardo (9). By arbitrage the exchange rate has to equalize the purchasing power of money in two countries or, alternatively, in the change variant of PPP, the exchange rate has to change to the extent of the difference between the inflation rates. The second theorem is uncovered interest parity, already dealt with by Keynes. It says the exchange rate has to change to the extent of the difference in interest rates. 
Now these theories both follow from the arbitrage principle, and the arbitrage principle from rationality and, even if individual agents were not rational, from competition, as the rational agents would rapidly eliminate the then heavily loss-making non-rational ones. So it came as a complete shock to exchange rate economists when Meese and Rogoff showed in 1983 (10) that empirically the theories named, and also all mongrel theories constructed from them, fail completely. The best theory for explaining exchange rates is the random walk without drift, which basically is nothing but the non-theory that the best forecast of the exchange rate is always that it will not change – whatever it is and whatever its history has been up to then. The proof of Meese and Rogoff is empirical, and it is extremely difficult to fault them. This fact, saying, as it does, that there is no useful economic theory at all on the most important prices in the modern world, is so disquieting for economists that later on Rogoff himself cheerfully presented theories in evident conflict with his own most important and as yet uncontroverted historical finding (11). 
Now, this denial of all stable theory of exchange rates is not quite so surprising once you examine more closely the two paradigms named, i.e. purchasing power parity and uncovered interest parity. You will soon find

• that the two fully convincing theories are actually in full conflict with each other if real interest rates between countries differ, which they do both factually and according to well-established theories;
• that uncovered interest parity is actually stated over an undefined period and furthermore can never theoretically explain the effect of any permanent change in interest rates, be it ever so small;
• that it remains an undefined problem which interest rates one should examine and whether they could be e.g. the rates of return on common stock;
• and that the relevant price level or, alternatively, the relevant rate of inflation is still to be specified.

All these specifications will be undertaken by economic agents in a time-varying way. In other words, fads have to rule the roost. So precisely when you delve more deeply into the problem you find the only theories you can hold about exchange rates are historical hypotheses and, of course, historically variable ones. 
There is an interesting methodological point as well: The very efficient markets for exchange rates of large currencies are exactly those where demand and supply equilibrium is reached every day, often in the form of an auction. But this daily equilibrium can tell us basically nothing about the equilibrium next day, with which, most likely, we would be much more concerned. 
This is a fascinating field of inquiry for my students. I have them develop better theories of one given exchange rate or a few of them during a limited period, and every other student will have the satisfaction to develop a new mini-theory by proving – with complicated econometric methods – its forecasting relevance, i.e. its improvement relative to the simple random walk. Furthermore, one has to use different methods for the very short period up to perhaps a year, on the one hand, and for longer periods, on the other. This is straight historical research, though of very recent periods. It is fascinating for the researcher because there is so much that is unknown, and exactly because a compelling theory is absent. In addition, one learns the trade of certainly the highest-paid kind of “historicist” observers, the ones in financial markets. I can also report to you the possibly interesting fact that one of my students found the behaviour of the Greek drachma to be the relatively best explainable of all currencies relative to the German mark. For this, there are two possible explanations: Explanation one would say this is so because the Greeks are the most rational of all the people examined by my students, a flattering conclusion. Explanation number two would not be quite as flattering. It would say that among all the economies examined by my students with respect to exchange rates the Greek economy was the most primitive. This would follow because for an economy which has commodity movements but not yet substantial capital movements the exchange rate has to track, at least with a lag, purchasing power parity, and a standard exchange rate theory will come through. Let me comfort the Greeks in the audience: Even the drachma was, up to recently, only explainable up to 30 percent of its variation and most other currencies perhaps up to 10 or 15 percent, so that even the Greeks are nearly unexplainable. The startling conclusion is that, with regard to exchange rates, it is exactly the most developed economies which have to be explained purely historically while good, i.e. theoretically satisfactory, economic explanations can only be found for as yet undeveloped or threshold countries. Thus, in the case of exchange rates, history proves to be high theory for the most developed nations.

To come back to my initial simile of an economist facing the estuary and pondering how best to get across it: It is the historical method to look for the best way by weighing all the pros and cons, not least among them the likely costs. Only the historical method can find a rational solution. Those who will not avail themselves of it, will forge ahead anyhow – and may easily be drowned.

(1) Joan Robinson, “The Second Crisis of Economic Theory”, American Economic Review, Papers and Proceedings, 62 (1972), pp. 1–10 (Ely Lecture), here p. 1.
(2) Kenneth E. Boulding, “After Samuelson Who Needs Adam Smith?”, History of Political Economy, (1991), 3:2, pp. 225–237. Quoted after Mark Blaug, “No History of Ideas, Please, We're Economists”, Journal of Economic Perspectives (Winter 2001), 15:1, pp. 145–164, here pl. 149, note 169.
(3) Erich W. Streissler, “Deutschland, Deutschland über alles?” and “Deutschland, Deutschland über alles? Teil II”, Wochenpresse, No. 14, 6 April 1990, pp. 48–53 and No. 15, 13 April 1990, pp. 42–45.
(4) F. A. Hayek, “The Theory of Complex Phenomena”, in: The Critical Approach to Science and Philosophy, Essays in Honor of Karl R. Popper, Mario Bunge (ed.), Glencoe IL, 1964, pp. 332–349.
(5) J. M. Keynes, “Alfred Marshall” [1924] in id., Essays in Biography (1951 edition), p. 158, footnote.
(6) Philip Mirowski, More heat than light – Economics as social physics: Physics as nature's economics, Cambridge 1989, seems to misunderstand economics. Though there are highly abstract, conceptional problems to be solved now and then, as in theoretical physics, most of the subject is much more “applied", corresponding to engineering. There is a qualitative difference in the degree of complexity to be dealt with in both cases of intellectual endeavour, and in both fields. Mirowski says: “It comes as a shock to learn that it has been estimated that 90 percent of all engineering systems cannot be treated by currently available methods of statistical mechanics, and are instead resolved by resort to classical thermodynamics or phenomenological procedures” (p. 66). A shock to whom? Apparently to Mirowski, who should rather seek for the corresponding “shock” (?) in economics.
(7) Wilhelm Krelle, “Dynamics of the Utility Function”, in: J. R. Hicks and W. Weber (eds.), Carl Menger and the Austrian School of Economics, Oxford 1973, pp. 92–128.
(8) Michael C. Burda, “Mehr Arbeitslose – Der Preis fŸr die Osterweiterung?”, in: Erweiterung der EU, Jahrestagung 1999, Schriften des Vereins fŸr Socialpolitik N.F. 274, Berlin 2000, pp. 79–102.
(9) It would seem so from the paragraph preceding the remark in “On Foreign Trade”, chapter VII: “This I believe to be the only two causes which regulate the comparative value of money in the different countries of the worlds”. See The Works and Correspondence of David Ricardo, Piero Sraffa (ed.), vol. I, Cambridge 1977, p. 145. In the index, vol. XI, Sraffa mentions “comparative costs and foreign trade, I, 133–141”. 
(10) Richard A. Meese and Kenneth S. Rogoff, “Empirical Exchange Rate Models of the Seventies: Do They Fit Out of Sample?”, Journal of International Economics, 14 1/2 (Feb. 1983), pp. 3–24.
(11) See e.g. Kenneth S. Rogoff, “The Purchasing Power Parity Puzzle”, Journal of Economic Literature, 34 (1996), pp. 647–668; Maurice Obstfeld and Kenneth S. Rogoff, “Exchange Rate Dynamics Redux”, Journal of Political Economy, 103/3 (June 1995), pp. 624–660.