Studies in Contemporary Economics

Transcription

1 Studies in Contemporary Economics Bert Hamminga Neoclassical Theory Structure and Theory Development An Empirical-Philosophical Case Study Concerning the Theory of International Trade Springer-Verlag Berlin Heidelberg New York Tokyo 1983

2 Editorial Board D. Bos G. Bambach B. Gahlen K. W. Rothschild Author Bert Hamminga Katholieke Hogeschool Tilburg Postbus 90153, 5000 LE Tilburg, The Netherlands ISBN I3: e ISBN I3: : / This work is subject to copyriltlt All ris;l1ts are reserved, v.tlether the v.t1oie or part of the material Is concerned, speciiicajly thoee of transiatlol\ reprinting, re-use 01 DIusIratIons, broedcasiing, reprcduciion by photocopying machine or alnilar means. and storage in data banks. Further, storage or utilization of the described PI'OIJ'8Im1S on data processing Installattons Is forbidden without the written permission of the author. l.klder 54 of the German ~ Law where copies are made for other than private use, a fee Is payable to "Verwartungsgeselischaft Wort", Munich. C by Springer-Verlag B e ~ Heidelberg i n

3 PREFACE There are quite a lot of recent books on the methodology of economics, but all proceed from one or more of the following three fundamental assumptions: 1) Meta-apriorism: it is a priori believed that the results of the philosophy of science of the past decades, associated with the names of Popper, Kuhn and Lakatos, yield conceptual frameworks that can be used to describe the product of the economist's endeavours. 2) Synchronism: it is believed that the inspection of a "theory" of economics, the presentation of some economic belief at a point in time, reveals its logical structure and ipso facto the "method" by which the "theory" is constructed. 3) CreduZity: it is believed that what economists themselves say on their own methods is true, or at least of primary relevance. In this book, I endeavour to show that these assumptions are false. First, the philosophies of science by Popper, Kuhn and Lakatos did not succeed in constructing a conceptual framework capable of describing theory development in economics. Secondly, this can only be seen as soon as not Zogic but history is taken to be the referee judging the adequacy of metatheories (diachronism). Thirdly the results of metaempiricaz (not-meta-aprioristic), diachronic (not-synchronic) research reveals that even economists themselves turn out frequently to have inadequate metatheories upon which it is dangerous to rely (suspicion instead of credulity). This book presents an empirical metatheory of theory structure and theory development in neoclassical economics. The metatheory is tested and illustrated by means of a (diachronic) case study of the neoclassical (Heckscher-Ohlin) application programme of international trade. The method must be named "empirical" because no evaluative or conceptual principles originating from philosophical doctrines are used to select and present the scientific achievements of economists. Instead, the evaluative behaviour of economists themselves is revealed, by means of systematic study of the interrelationships between important publications in an historic sequence of a research tradition. "Importance" is defined in harmony with the economist's, not the philosopher's, evaluation behaviour. On the basis of this study, a completely new metatheory is constructed to correctly describe the hard facts of theory

4 development in neoclassical economics. This new metatheory turns out to illuminate the causes of some metatheoretical controversies in economics, and the conceptual frictions that prevent some doctrines in the philosophy of science, most notably those of Popper, Kuhn and Lakatos, to see what neoclassical economics really is like. The cradle of the book stood in Groningen, where I was a student in the Philosophy Department of the University of Groningen. I owe a lot to Dr. Th.A.F. Kuipers and Prof. L.W. Nauta, not only for stimulating, but especially for jamming the brackes on me at the right moments. Later on, I have been a member of the Economic Institute of the University of Utrecht, the Philosophy Department of the Katholieke Hogeschool Tilburg and the Netherlands Institute for Advanced Studies. All these institutes owe my gratitude for providing a very agreeable research environment even in the face of the on-going bureaucratic strangulation of research in the Netherlands. I received many valuable conjectures and stimulating refutations, among which at least those from Dr. W. Balzer, Dr. N.B. De Marchi, Drs. J.H. von Eye, Dr. Th. Junius, Prof. G. Pikkemaat and Drs. J.J. Siegers should be mentioned. Thanks for accurate typing are due to Marthy Sacci-Oostveen, Henny stabel-groenewout and Jacqueline Suos-Wayers. Errors that remained are, of course, my responsibility alone. Tilburg, June 1983 IV

5 CONTENTS page INTRODUCTION 1. Introduction for philosophers of science 2. On the cause of fruitless methodology (an introduction for economists) 2.1. The problem of fruitless methodology 2.2. The cause of fruitless methodology 2.3. Two preliminary arguments in favour of a principle of Suspicion Aims and set-up of this study 3.1. Aim of this study 3.2. Set-up of this study A. Two central theses in the theory of international trade and their place in the economists' world view 11 A.1. The neoclassical econpmists' world view: neoclassical foundations of economic analysis 11 A.2. The effects of international trade through the economists' spectacles: comparative statics 15 A.3. Two central theses in the theory of international trade 16 A.4. Why are the theorems "interesting" 17 A.4.1. Internal explanations 18 A The classical theory of international trade 18 A The production function 20 A The production function, the classical problem, and the Heckscher-Ohlin solution 22 A Conclusion 26 A.4.2. External explanations 28 B. The emergence of interesting theorems 30 B.1. Introduction 30 B.2. The Stolper-Samuelson Exposition 31

6 B.2.1. The S&S-exposition, Part I: The Heckscher-Ohlin explanatory ideal and results already achieved 31 B.2.2. Neoclassical intermezzo I: mathematical marginalism and the introduction of a second product B.2.3. The Stolper-Samuelson exposition: Part II B The index number problem B The production structure once again: the box diagram B The production possibility curve B.2.4. Neoclassical intermezzo II: why all this is "neoclassical" B Demand structure: utility-function and indifference curves 38 B The neoclassical "General Equilibrium" FEA: a final characterization 39 B.2.5. Finale: The S&S-exposition of the effects of international trade: the S&S-result 41 B.3. Questions arising in connection with the S&Sexposi tion 45 C. Economic "normal science" centred around interesting theorems 47 C.l. Introduction 47 C.2. Some central concepts in informal symbolism 49 C.3. Examples of research centred around the S&S-theorem 51 C.3.1. The Metzler-results 51 C.3.2. Second example: the Koo-results 52 C.3.3. Third example: The SOdersten and Vind result 53 C.4. Research centred around the factor price equalization theorem C.4.1. The emergence of the FPE-theorem in V 222 C.4.2. Problems of the FPE-theorem in V 222 C No diversification cone C More than one diversification cone VI

7 C.4.3. Some examples of field extension with respect to the FPE-theorem C.4.4. Further weakening of conditions and alternative conditions results C.4.S. Some further examples of field extension with respect to FPE: m F n C.S. Summary D. Consolidation: metatheory on theory structure and theory development in the Ohlin-Samuelson research programme 66 E. The hypothesis of the plausibility strategy E.1. The "plausibility strategy": a first approach E.2. Initial explanatory power of the hypothesis E.3. Elementary plausibility E.3.1. The forced hypothesis about the existence of a sixth characteristic of the Ohlin-Samuelson programme: an SEPC E.3.2. A taxonomy of elementary plausibilism E Conditions plausibilism E Conditions plausibilism I: special conditions E Conditions plausibilism II: EI-conditions E Conditions plausibilism III: FEA-conditions E Field plausibilism E Consequence plausibilism E.4. SEPC, the enigma F. Problems of econometrics. The mutal independency thesis F.l. Identification problems F.2. Anchoring problems F.3. How these problems arise from the nature of theory development in the Ohlin-Samuelson programme F.4. The mutual independency thesis VII

8 G. Conclusion: results and relevance 102 G.l. Summary of the meta theory 102 G.2. Some concluding remarks on the relationship between theory development and plausibilism 104 G.3. Some beliefs about the generality of the meta theory 106 G.3.1. From neoclassical theory of i n t e r n a trade ~ ~ o n a l to neoclassical theory structure and development in general 106 G.3.2. Towards theory structure and theory development in economics? 109 G.3.3. Conclusion about generalizability 109 G.4. The merits and problems of other metatheories if applied to neoclassical general equilibrium research programmes 110 G.4.l. The construction of a Meta-Philosophical Language 111 G.4.2. Neoclassical theory structure and theory development restated in MPL G.4.3. The "Friedman controversy consensus" in MPL G.4.4. Carnapian inductive logic in MPL G.4.5. Popper ian falsificationism in MPL G.4.6. Lakatosian "Methodology of Scientific Research Programmes" in MPL 119 G.4.7. KUhnian revolutionary and normal science 123 G.4.8. Sneed's approach to theories of mathematical physics 128 G.4.9. Summary 131 G.5. Vice, Virtue and Rationality 134 H. Appendix: a scheme for formalization of the link between theory development and plausibilism 138 I. Appendix: the irrefutability of the Samuelson-type meaningfull theorems 146 J. Appendix: The Friedman controversy and its consensus 152 VIII

9 K. Appendix: Ricardian versus neoclassical FEA 161 L. References 164 M. Index 170 IX