Formalism and macroeconomics - a post-keynesian perspective

ARGUMENTA OECONOMICA No 1 (12)2002 PL ISSN 1233-5835

I. ARTICLES

FORMALISM AND MACROECONOMICS

- A POST-KEYNESIAN PERSPECTIVE

Much o f today’s economic research is characterized by a uniform methodology: the use of formalism. Although it is unquestionable that economics as a social science has gained scicniific strength through history by use of a formalistic approach, noi all relevant economic information can be measured quantitatively and put on a mathematical mode of expression. Economics is not quite like the natural sciences. Much econom ic evidence is qualitative in nature. So one ought to discuss the use o f formalism more critically than is common practise today. And the scientific community should accept that a more methodological pluralistic approach could further enhance the progressive status o f economics.

INTRODUCTION

In Blanchard (2000) the development of macroeconomics from the pre- 1940s until the end of the century is evaluated and highlighted. And the story that Blanchard tells is a happy one. From the very beginning we are told that “progress in macroeconomics may well be the success story of twentieth century economics” and that economic science should be characterised by the fact of “a surprisingly steady accumulation of knowledge”, (Blanchard 2000 p.

1375).

L o o k in g a t m o d e m m a c r o e c o n o m i c s to d a y w e a re to l d t h a t m o d e m theory is solidly g ro u n d e d in a general e q u ilib riu m stru ctu re. M odern m o d e ls ch aracterize the econom y a s b e in g in tem porary e q u ilib riu m , given the im p lic a tio n s o f the past, and the an ticip atio n s o f the future. T hey p ro v id e a n in terp retatio n o f flu c tu a tio n s as the result o f shocks w o rk in g th e ir w ay through p ro p a g a tio n m echanism s. M u c h o f th e current w ork is focused o n th e ro le o f im p erfe ctio n s (B la n c h a rd 2000, p. 1402).

And much of the earlier debates within macroeconomics seem to have vanished almost completely. Perhaps as a consequence of, as Blanchard puts it, the a-ideological character of modern macroeconomic research. But can

’ Department of Environmental and Business Economics. University o f Southern Denmark, Esbjerg

ideology really be put aside completely? Is economics no more a true social science? Or has economics developed to become too much similar to natural sciences?

However impressing the presentation by Blanchard is, one ought to discuss also perhaps in some detail the mainstream methodology of economics. In what follows, I present a personal, somewhat critical view, inspired by Keynes’s attitude towards economics as a social science and the views held by the modern camp of Post-Keynesians, on the ruling methodology of today’s macroeconomics which I find have gone too far in its excessive use of formalism. The main purpose of the article is primarily to put forward some important questions which modern economists should bear in mind, rather than to present some clear-cut and simple answers.

ON THE USE OF FORMALISM

Ever since the introduction of marginalism in economic theory and the victory of the neoclassical paradigm, macroeconomics has become more and more mathematical in its substance and in its mode of expression dealing with a formal-logical modelling of fundamental socio-economic phenomena. As a natural consequence of this nowadays there seems to be less room for the more qualitative aspects in macroeconomics. Our business has become more quantitative in its representation. O f course, we acknowledge that real life is not all about phenomena that can be accurately measured. When we address the public and between ourselves discuss economic problems of the day we also take, one must hope, the relevant more qualitative aspects into consideration. But looking upon the academic life on campus we hardly get any credit and certainly no merit for our efforts within what often is called the soft side of economics.

Perhaps the majority of professional economists find this very good, to be sure most of us have learnt to live under these conditions but still some (Post- Keynesians in particular), although not many, would argue that economics is and has always been a social science that has also to address and discuss matters of a more qualitative character. According to Blaug (2001) some even leave mainstream economics and go to more heterodox fields for research as a consequence of this. Or as Blaug (2001, p. 147) expressed it:

If yo u a re p h ilo so p h ically in clin ed - an intellectual ra th e r th a n a technocrai - but are a ttra c te d to eco n o m ics becau se o f its p o licy relevance o r th e b e lie f that society rests essen tially o n eco n o m ic fo u n d atio n s, y o u may well find y o u r s e lf d riftin g tow ards history o f e c o n o m ic th o u g h t... h is to ry o f eco n o m ic th o u g h t is a h a v e n for heterodoxy.

a h etero d o x y w h ic h no doubt has m an y s o u rc e s but at its fo u n d a tio n ta k e s its departure ... from a c e rta in type o f m ind, a c e rta in c o n g e n ia l style o f th in k in g .

The degree to which economics is or should be more or less qualitative is not a new discussion among economists. Looking upon history John Maynard Keynes expressed himself very clearly in his correspondence with Roy Harrod in 1938 where he discussed the role o f econometrics as a new field of economic research. Keynes was working on a review of the early works of Jan Tinbergen, which appeared in The Economic Journal in 1939. Although sceptical in his views on econometrics, Keynes recommended Tinbergen nevertheless to continue his efforts and to follow up his early work. In his review of Tinbergen he pointed out that: ‘‘the main p rim a facie objection to the application of the method of multiple correlation to complex economic problems lies in the apparent lack of any adequate degree of uniformity in the environment”, (Keynes 1939, p. 316).

Or as he expressed himself in the correspondence with Harrod:

In c h e m is try and physics and o th e r n a tu ra l scien ces the o b je c t o f experim ent is to fill in the a c tu a l values o f the v ario u s q u a n titie s and factors a p p e a rin g in an equation or form ula; an d th e w ork w hen d one is o n c e , an d for all. In e c o n o m ic s th at is not the case, and to c o n v e rt a m odel into a q u a n tita tiv e fo rm u la is to d e s tro y its usefulness as an instrum ent o f th o u g h t ... e c o n o m ic s b e in g a m oral s c ie n c e ... it deals with in tro sp ectio n an d with values ... it d e a ls w ith m otives, e x p e c ta tio n s , psychological un certain ties . . . eco n o m ics is a s c ie n c e o f th in k in g in term s o f m o d e ls ... w hich are relevant to th e co n tem p o rary w orld ... b e c a u s e , unlike the n a tu ra l sc ie n c e , the m aterial to w hich it is a p p lie d is ... not h o m o g e n e o u s th ro u g h tim e (K e y n e s 1938).

But this is not to say that micro- as well as macroeconomics does not deal with more measurable empirical matters. Quantitative phenomena can and should be measured in the best way possible. But not all economic phenomena are measurable. You have to allow methodology to be pluralistic, which is exactly K eynes’s point. In some respect, the scientific approach of the natural sciences can be used effectively. But it cannot be used successfully in handling all kinds of economic problems. Or in the words of Joan Robinson (Robinson 1976, p. 26): “Without the possibility of controlled experiment, we have to rely on interpretation of evidence, and interpretation involves judgement; we can never get a knock-down answer” . Later on Robinson becomes even more sceptical about the economic science as she writes not without irony:

... la c k in g the experim ental m e th o d , e c o n o m ists are not s tric tly en o u g h com pelled to red u ce m e ta p h y sic a l concepts to fa ls if ia b le term s and c a n n o t c o m p e l each other to agree as to w h a t has been falsified. S o e c o n o m is ts lim ps a lo n g w ith o n e foot in untested

h y p o th eses an d the other in u n te s ta b le slo g an s. H ere ou r task is to so rt out as best we may this m ix tu re o f ideology and s c ie n c e (R o b in so n 1976, p. 2 8 ).

So Robinson too is sceptical if the use of mathematics in economic theory becomes too extensive and according to Davidson (1991, p. 23) she once should have said about her own research “I never learned to use mathematics to develop theory; therefore I had to learn how to think about problems”.

Of cause, there are similarities as well as differences between different kinds of sciences. Perhaps scientific breakthroughs in the natural sciences has more of “a once and for all” character than those of the social sciences. The natural sciences is at least in some respect a-historical because a given experiment can be made over again and again right until one is certain that the relationship one has found actually is a correct and a stable one. The scientific achievements within the natural sciences are in this respect independent of the economic, the social, and the psychological environment in a way that is not the case with economic research. Ours is the case o f interdependency. Economic problems have to be seen in a given historical context:

...e c o n o m ic ideas are alw ays an d in tim ately a product o f th e ir tim e and place; they can n o t b e se e n a p a rt from the w o rld th ey interpret. And th at w o rld ch a n g e s - is, indeed, in a c o n s ta n t p ro cess o f tra n sfo rm a tio n - so econom ic id e a s, if they are to retain relev an ce, m u s t also change (G alb raith 1987, pp. 1-2).

With inspiration from the natural sciences we might have invented Homo Economicus, but as Thaler (2000) has pointed out, this is not a very good representation of the living modern Homo Sapiens.

Also Payson (1997) is critical about the scientific status of modern economics. Has economics the right scientific status as it ought to have when judged as a social science discipline? Although much modern macroeconomics, for instance by the acceptance o f the rational expectation hypothesis, is characterized by the heavy use o f mathematics and thereby have become properly scientific, as many economists would argue, is such an approach really always helpful in understanding the true nature of economic phenomena? Perhaps formalism is quite convenient to use when you make a theoretical analysis but one should be aware as Payson points out that:

... m a th e m a tic s does not u n d erlie e c o n o m ic p h en o m en a - h u m a n and institutional b eh av io r d o , and that involves p sy c h o lo g y and sociology ... th e o n ly explanation for the fact th a t sc ie n c e is not used v ery m u c h in econom ics is th a t m o s t eco n o m ists today arc sim p ly c a rry in g out the only fu n c tio n s they have e v e r le a rn e d to perform : high- p o w ered m a th e m a tic s, neoclassical sy n th e s is , and “sc h o la rsh ip ” in the gam e o f getting p ublished (P a y s o n 1997, p. 260 & p. 2 7 3 ).

A sim ilar line of critical argumentation can be found in Boulding (1971). In this, (Boulding 1971, p. 233), he argues that the job o f many mainstream economists is in danger of becoming more and more: “an endless modification of variables and equations in regions of strongly diminishing returns in the knowledge function, and still sharper diminishing returns in the significance function” . These views are in good accordance with the attitude presented in Patinkin (1976). Like Payson, he is also worried that economics is in danger of losing scientific power if, as Patinkin fears, it should manifest itself as a discipline where:

... sy m b o lic p se u d o -m ath e m atical m eth o d s o f fo rm a liz in g a sy stem o f econom ic analysis ... w hich allow the a u th o r to lose sight o f th e co m p lex ities and in te rd e p e n d e n c ie s o f the real w o rld in a m aze o f p re te n tio u s a n d unhelpful sym bols (P atin k in 1976, p. 512).

And scepticism concerning the use of formalism in economics is far from a new phenomenon. Even Alfred Marshall - one of the fathers of the neoclassical paradigm - advocated the case of thoughtfulness when using mathematics in economic theory as is shown in his correspondence with his friend A. L. Bowley. In 1906 Marshall wrote what later has become a very famous quotation:

... I had a g ro w in g feeling in th e la te r y ears o f my w o rk at th e su b ject that a good m a th e m a tic a l th eo rem dealing w ith e c o n o m ic h y potheses w as v e ry un lik ely to be good e co n o m ics: an d I w ent m ore and m o re o n th e rules - 1) U se m a th e m a tic s as a shorthand lan g u ag e, ra th e r than as an en g in e o f in q u iry . 2) K eep to th e m u n til yo u have done. 3) T ra n sla te in to E nglish. 4) T hen illu s tra te by ex am p les th at a re im p o rta n t in real life. 5) Burn th e m a th e m a tic s. 6) I f yo u c a n ’t su cceed in 4), b u rn 3). T h is last I did often (quoted fro m L an d reth & C o la n d e r 1 9 9 4 , pp. 2 9 0 -9 1 ).

In Keynes’s biography on Marshall, he quotes M arshall’s views on the need to study political economy and what he found out about economics throughout his long life (as he him self wrote in retrospect about the year 1917):

I ... re g a rd e d m y se lf as a w a n d e re r in th e land o f d ry fa c ts ; lo o k in g forw ard to a speedy re tu rn to the luxuriance o f p u re thought. B ut the m o re I studied econom ic science, the sm a lle r appeared the k n o w le d g e w hich I had o f it, in p roportion to the k n o w led g e th a t I needed; and n o w , at the end o f nearly h a lf a cen tu ry o f alm o st e x c lu siv e stu d y o f it, I am c o n sc io u s o f m ore ignorance o f it th a n I w as at the beginning o f the stu d y (K e y n e s 1924, p. 171).

Therefore, according to Marshall theory could never stand alone. It always had to go hand in hand with empirical evidence. If one does not work along these lines one runs the risk of getting a theory constructed that is too far away

from real life economic phenomena. As Beed & Beed (2000) have pointed out one of the dangers of giving to much way to pure formalism is that:

’T h e o r y ’ is no longer seen as p ro p o s itio n s purporting to d e s c rib e , explain, or predict the real w o rld . It is becom ing s c h e m a ta d e sc rib in g how th e re a l w o rld m ig h t look if people b e h a v e d in the w ay the th e o ry su g g e ste d . E co n o m ic th e o r ie s describing how peo p le a c tu a lly behave are less c o m m o n .

But such an approach has not always been easy to pursue in economics as the history of economic thought tells us. On the contrary, it has often been seen by many as a long defeated research strategy. As Keynes stated in his biographic essay on Thomas Malthus, the blame for this ought perhaps at least partly to be ascribed to David Ricardo, as he - in clear opposition to Malthus, who in this respect is the hero of Keynes - came to dominate completely the methodology of economic theory in its early years: “ ... that ... the complete domination of Ricardo for a period o f a hundred years has been a disaster to the progress o f economics”, where Malthus, in a much better way than Ricardo, blended theory with empirical evidence as he worked to give:

... fo rm a l th in k in g to the co m p le x c o n fu s io n o f the w orld o f d a ily ev en ts ... so as to penetrate th ese e v e n ts with u n d e rs ta n d in g by a m ixture o f in tu itiv e selectio n and form al principle a n d th u s to interpret the p ro b le m an d propose the re m e d y (K e y n e s 1935, p. 98 & p . 107).

Of course, Keynes’s sympathy towards Malthus’s approach reflects the way he himself tried to work scientifically. Economics is concerned with problem solving and some of these problems are empirical in nature. So why not try to blend inductive information about the way the economy is supposed to work with deductive logic? The development of society and the development of macroeconomic theory cannot and should not be seen apart from one another (this approach of Keynes’s has been termed a realytic approach by Landreth & Colander (1994, p. 463) as they state: “a realytic theory is contextual”). Although perhaps it is easy to understand the relevance of this approach it is not always a simple task to follow such a research strategy for the economist. It can indeed be very difficult to change ones preconceived opinions about the accepted state of affairs as Keynes himself very elegantly pointed out in the preface to T he General Theory:

T h e id e a s w h ich are here e x p ressed so lab o rio u sly are e x tre m e ly sim p le and should be o b v io u s. T h e d ifficu lty lies no t in th e n e w ideas, but in e s c a p in g from the old ones, w hich ra m ify , fo r those b rought up a s m o s t o f us have been, in to e v e ry co rn er o f our m inds (K e y n e s 1936, p. xxiii).

Another problem related with too much use of formalism in macroeconomics is the problem o f how to incorporate in the right way historical time in economic models. As John Hicks has argued, one has to acknowledge that: “Economics is in time, in a way that the natural sciences are not. All economic data are dated ... time is a device which prevents everything from happening at once” (Hicks 1979, p. 41). Economists ought to remember that economic activity, be it production or consumption, consists of very time consuming processes. And if you accept this, then what about the concept of equilibrium and the universal use o f general equilibrium models and rational expectations which only allow economic agents to make stochastic errors in modem macroeconomics? Can GE-models really handle the problems of incorporating historical time successfully? What if economic behaviour is not characterized to a significant part by simultaneous action, perfect competitiveness, rationality or perhaps more importantly institutional stability? If: “economic theory cannot be static when its object of investigation is fundamentally historical” (Davis 1989, p. 436), should a non-contextual approach then be accepted as the predominantly right one? According to Davis one has to take account of these aspects and very critically discuss the advantages and disadvantages of general equilibrium models before using them without hardly any methodological concern at all:

A x io m a tic G en eral E q u ilib riu m T h e o ry , how ever, e m b o d ie s a th eo ry o f concept d e v e lo p m e n t th a t is inescapably a h is to ric a l. T h a t the full e la b o r a tio n o f its key notions p roceeds a p rio ri through a c o n c e p tu a l a n a ly s is tied to form al d e m o n s tra tio n o f m arket- clearing an d o p tim a lity m eans that th e s e n o tio n s cannot a c c o m m o d a te an y developm ent o f the e c o n o m ic p ro cess w hich fails to re f le c t these results. T h e T h e o ry itself, that is, is a closed lo g ic a l stru ctu re out o f tim e , w h ic h m ust a c c o rd in g ly b e representationally in ad eq u ate . . . A x io m atic G eneral E q u ilib riu m T heory is th e n . . . sim ply unrealistic (D avis 1989, p p . 4 3 6 -3 7 ) .

The argument of Kaldor (1972, p. 1237) runs along sim ilar lines as he sums up his views on the use of the concept of equilibrium economics. According to Kaldor this

has b e c o m e a m ajo r ob stacle to th e d e v elo p m en t o f e c o n o m ic s as a scien ce - m eaning by th e term science a b o d y o f th eo rem s based o n a ss u m p tio n s that are e m p ir ic a lly d e riv e d (from o b s e rv a tio n s ) an d w hich e m b o d y h y p o th e se s that are capable o f v e rific a tio n both in re g a rd to th e a ssu m p tio n s and th e p re d ic tio n s.

This very critical assessment o f the relevance of using GE-models was immediately questioned by others as being too hard and too narrow-minded, e.g. by Hahn (1973).

And yet, although Davis’s concern is both understandable and important, do not all micro- as well as macroeconomic theories lack realism at least to some degree? To get a theory operational and working do we not have to give up the ideal of perfect realism? But of course, the trade-off between realism and operationalism is a very crucial one. As Keynes warned us in 1936, one ought not to assume all difficulties away just to get a handy and smooth running theory. “It may well be that the classical theory represents the way in which we should like our economy to behave. But to assume that it actually does so, is to assume our difficulties away” (Keynes 1936, p. 34).

But although the use of formalism in economics should give reason to critical and methodologically inclined assessment formalism also brings about many advantages to the economist. For instance, it can give us a more structured presentation and a better general view of a perhaps very complicated theory than just qualitative arguments alone. And you can build a model, collect data, and make estimations and simulations upon which you should try to falsify or to corroborate the proposed theoretical hypotheses and statements. And perhaps the econometric efforts can result in not only better prognoses but also give room for a more qualified debate on economic affairs and even help the politicians to conduct economic policy in a more accurate way. Although one should be careful not to overestimate the blessing o f econometrics as Keynes warned us already in the 1930s. If Hutchison (1994) is right in his conclusion arguing that the economies of today are more dynamic and volatile in behaviour and institutional structure than in the past, then the assumed stability that lies behind much econometric work breaks down. Then Rodrik (2000, p. 177) may still be right in claming that “economists rank second only to astrologers in their predictive abilities” . Aside from this, formalism can be a very handy pedagogical tool when one generation of economists tries to pass on the scientific knowledge gained through history to the next generation. Used with care and insight, formalism can make science grow progressively to use a Lakatosian term.

Some economists even argue to the point that it is only through the use of more formalism in economic theory that economics as a science has gained strength. For instance, Lazear (2000) points out that economics is more scientific than all other of the social sciences exactly because of a more cogent reasoning. In economics we assume that agents behave rationally and effectively in an environment of equilibrium, and we can formulate these actions of the rational economic man in precise mathematical terms. And in doing this, we can make the best of both worlds because according to Lazaer (2000, p. 102): “economics is scientific; it follows the scientific method of stating a formal refutable theory, testing the theory, and revising the theory

based on the evidence”. But is this really a true picture of modern macroeconomics? To some it might seem a little bit too prosperous to be true. Perhaps Niehans (1981, p. 174) is not all that mistaken in saying that:

H ard ly a n y eco n o m ic theory is e v e r e m p iric a lly falsified. It ra th e r falls into disuse and is fo rg o tte n , perhaps to be re d is c o v e re d d ecad es la te r . . . e c o n o m ic doctrines are usually te s te d n o t by system atic m e th o d s , b u t by a D a rw in ia n s tru g g le for survival in the aren a o f h isto ry .

And economics is not like the natural sciences because in many important respects economics is “a science of unique events” (Niehans 1981, p. 175). If that is to be the case then we are bound to have a problem with falsification in economics. It does not come in as handy in economics as is the case with the natural sciences. But the problem of the economist is not only to make tests and try to falsify hypotheses. He must also give way to try

to “e x p la in ” th e past and, by so d o in g , learn so m eth in g th a t m ig h t be useful in the future. In th e c h a o s that reality re a lly is, he tries to create little is la n d s o f intellectual order, in th e e v e r-ch an g in g flow o f h is to ry he tries to cre a te d u ra b le and predictable patterns (N ie h a n s 1981, p. 167).

Then to Niehans and Blaug (2001) and many others, economic history and the development of economic thought matter. And the study of these two disciplines need not be very formalistic in their approach.

CONCLUDING REMARKS

Above all science is about explanation and the solving of problems. In performing our task we have to apply the right methodology. Although mainstream macroeconomics might agree upon that the methodology should be one of formalism, not all relevant economic evidence can be found or put in quantitative terms. Economics is a social science. And economic actions performed by bounded rational people in imperfect markets, in constant changing societies happen in historical time. So economic phenomena and their actual development are path dependent in a very crucial way. Or as Davidson (1991, p. 35) states;

O ur k n o w le d g e about e c o n o m ic e v e n ts occurring th ro u g h tim e is, how ever asy m m etric: alth o u g h w e m ay k n o w th e p a st, we c a n n o t b e su re th a t we have any reliable k n o w le d g e about the e c o n o m ic fu tu re. T h e future r e m a in s to be created by hum an a c tio n s an d is not m erely d e te r m in a te d by som e im m u ta b le eco n o m ic law. In other w o rd s, f o r m any im portant e c o n o m ic activ ities e sp e c ia lly lo n g d u ra tio n ones

-in fo rm atio n a b o u t future cu rren t o u tc o m e s does n o t c u rre n tly ex ist. The econom ic future is y e t to be created and is no t p re d e te rm in e d by e x is tin g ru le s o r econom ic law s.

And the expectations people form may be rational in the way that they always try to make the best expectations possible but that is not to say that they only make stochastic errors when their planned economic activity is realized. Human behaviour is not rational in the way the rational expectation hypothesis argue.

So macroeconomic theory is bound to be contextual if it wants to be successful. And in gaining scientific success formalism has its place. But formalism cannot do the job alone. You have to accept that not all economic relevant information is quantitative in its nature. You have to give way to qualitative evidence as well. And you have to realize that most scientific economic knowledge is not easily bought. Not even with advanced econometric tools and as Johnston (1991, p. 52) has warned us, we should be aware of the possible dangers of modern techniques:

It is th u s all to o possible for so m e o n e to activate an e c o n o m e tric softw are package, o f w hich he has only a dim u n d e rs ta n d in g , to apply it to d a ta o f w hose nature and p ro v en an ce h e is ignorant, and th en to d ra w conclusions a b o u t a n eco n o m ic situation, w hose h is to ric a l and institutional re a litie s he has, perhaps, not s tu d ie d in any depth.

Therefore, there should also be room for at least a minimum of knowledge of how economic thought and society developed in the past as argued by Blaug (2001).

If you accept that economics is a true social science and give way to more methodological pluralism than just the enhanced chase after the virtues of formalism alone and overcome the uniformities of many o f today’s academic circles within economics, then the future for economic science looks very prosperous indeed. With a slight rewriting of a well known economic statement we could then have it that macroeconomics does matter and so do perhaps also economists in the future to come as was the case in the past.

REFERENCES

Blanchard, O. (2000): What Do We Know about Macroeconomics that Fischer and Wicksell Did Not?, The Quarterly Journal of Economics, pp. 1375-1409.

Bced. C. Beed. C. (2000): Intellectual Progress and Academic Economics: Rational Choice and Game Theory, Journal of Post Keynesian Economics 1999-2000, pp. 163-85.

Blaug, M. (2001): No History o f Ideas, Please, W e’re Economists, Journal of Economic Perspectives, Winter 2001, pp. 145-64.

Boulding, K. (1971): After Samuel son, Who Needs Adam Smith?, History o f Political Economy 1971, pp. 225-37.

Davidson, P. (1991): Controversies in Post Keynesian Economics. Aldershot: Edward Elgar. Davis, J. B. (1989): Axiomatic General Equilibrium Theory and Referentiality, Journal of Post

Keynesian Economics 1989, pp. 424-38.

Galbraith, J. K. (1987): A History o f Economics - the Past as the Present. London: Hamish Hamilton.

Hahn, F. H. (1973): On the Notion o f Equilibrium in Economics. Cam bridge University Press. Hicks, J. (1979): Causality in Economics. Oxford: Basil Blackwell.

Hutchison, T. W. (1994): Ends and M eans in the Methodology o f Econom ics in: New Directions in Economic Methodology, R. Backhouse (ed.), pp. 27-34, London: Routledge.

Johnston, J. (1991): Econometrics: Retrospect and Prospect. The Econom ic Journal 1991, pp. 51-56.

Kaldor, N. (1972): The Irrelevance o f Equilibrium Economics. The Econom ic Journal 1972, pp. 1237-55.

Keynes, J. M. (1939): Professor Tinbergen’s M ethod, Collected W ritings Vol. XIV, pp. 306-18, London: Macmillan St. Martin’s Press 1973.

Keynes, J. M. (1938): Two Letters to Roy H arrod, in: Philosophy o f Economics. An Anthology, D. M. Hausman (ed.), Cambridge: Cam bridge University Press 1984, pp. 300-302.

Keynes, J. M. (1936): The General Theory o f Employment, Interest and Money, London: Macmillan Cambridge University Press 1973.

Keynes, J. M. (1935): Thomas Robert M althus, Collected Writings Vol. X, pp 71-108, London: Macmillan St. M artin’s Press 1972.

Keynes, J. M. (1924): Alfred Marshall, Collected Writings Vol. X. pp 161-231, London: Macmillan St. Martin’s Press 1972.

Landreth, H.. Colander, D. C. (1994): H istory o f Economic Thought, Boston: Houghton Mifflin Company 1994.

Lazear, E. P. (2000): Economic Imperialism, The Quarterly Journal of Economics 2000, pp. 99-145. Niehans, J. (1981): Economics: History’, Doctrine, Science, Art, Kyklos 1981, pp. 165-77. Patinkin, D. (1976): Keynes and Econometrics: On the Interaction Between the Macroeconomic

Revolution o f the Interwar Period, in: John Maynard Keynes - C ritical Assessments, J. C. Wood (ed.). Vol. I, pp. 510-545, Croom Helm Ltd. 1983.

Payson, S. (1997): Regardless o f Philosophy, Economics Will Not Be a Science Until It Is Based on Science, Journal of Post Keynesian Economics 1996-97, pp. 2 57-74.

Robinson, J. (1976): Economic Philosophy. Harmondsworth: Pelican Books.

Rodrik, D. (2000): How Far Will International Economic Integration G o?, Journal of Economic Perspectives, Winter 2000, pp. 177-86.

Thaler, R. H. (2000): From Homo Econom icus to Homo Sapiens, Journal of Economic Perspectives, W inter 2000, pp. 133-41.