Pasinetti on Ricardo ISTITUTO DI TEORIA ECONOMICA E METODI QUANTITATIVI - Università Cattolica del Sacro Cuore
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Università Cattolica del Sacro Cuore ISTITUTO DI TEORIA ECONOMICA E METODI QUANTITATIVI Pasinetti on Ricardo Enrico Bellino January 2012 ISBN 978-88-343-2180-5 VITA E PENSIERO
Università Cattolica del Sacro Cuore ISTITUTO DI TEORIA ECONOMICA E METODI QUANTITATIVI Pasinetti on Ricardo Enrico Bellino January 2012 VITA E PENSIERO
Enrico Bellino, Istituto di Teoria Economica e Metodi Quantitativi, Università Cattolica del Sacro Cuore, Milano enrico.bellino@unicatt.it ist.temq@unicatt.it www.vitaepensiero.it All rights reserved. Photocopies for personal use of the reader, not exceeding 15% of each volume, may be made under the payment of a copying fee to the SIAE, in accordance with the provisions of the law n. 633 of 22 april 1941 (art. 68, par. 4 and 5). Reproductions which are not intended for personal use may be only made with the written permission of AIDRO, Corso di Porta Romana n. 108, 20122 Milano, e-mail: segreteria@aidro.org, web site www.aidro.org Le fotocopie per uso personale del lettore possono essere effettuate nei limiti del 15% di ciascun volume dietro pagamento alla SIAE del compenso previsto dall’art. 68, commi 4 e 5, della legge 22 aprile 1941 n. 633. Le riproduzioni effettuate per finalità di carattere professionale, economico o commerciale o comunque per uso diverso da quello personale possono essere effettuate a seguito di specifica autorizzazione rilasciata da AIDRO, Corso di Porta Romana n. 108, 20122 Milano, e-mail: segreteria@aidro.org e sito web www.aidro.org © 2012 Enrico Bellino ISBN 978-88-343-2180-5
Acknowledgments. I am grateful to Luigi Pasinetti for the conversa- tions about his interest for Ricardo since its genesis. I also thank Luciano Boggio, Pierluigi Porta and Rodolfo Signorino for the comments on a previous version of this note. As usual, the responsi- bility of the whole content is my own. Entry written for the Companion to David Ricardo, ed. by Heinz Kurz and Neri Salvadori, Edward Elgar, forthcoming. 3
4
1. Introduction The mathematical formulation of the Ricardian system proposed by Pasinetti (1960) has become the primary reference for all scholars interested in Ricardian economics ever since its publication. It was written in the Academic year 1957-58 when Pasinetti was at Harvard University attending a series of seminars for graduate students orga- nized by Franco Modigliani, who was at Harvard that year on leave from North Western University. Modigliani asked his students to chose their favourite economist and express his theory in mathemati- cal terms. Pasinetti chose Ricardo. His interest for Ricardo had arisen from his previous Academic year spent at Cambridge, U.K.,1 where he had the opportunity to read the Kaldor (1956) paper, attend his lectures and read the Introduction to Ricardo’s collected works writ- ten by Sraffa (1951). Pasinetti recalls that in its first formulation, his paper dealt separately with first the case of just one commodity (‘corn’), naturally following Kaldor’s diagrams which were put into equations, and then with the case of two commodities (‘corn’ and ‘gold’). When the work was later submitted to Sraffa for discussion, the latter expressed strong disagreement with such a distinction and convinced Pasinetti to immediately start with the case of two com- modities.2 A basic overview of Pasinetti’s article is provided in Sec- tion 2. The article formalizes the distribution and value theory con- tained in Ricardo’s Principles but in Section 3 we shall see how Pa- sinetti’s model actually formalizes the distribution theory contained in Ricardo’s early writings, which after Sraffa’s rehabilitative read- ings provided in (1951) and (1960) – see Section 4 – can be regarded as the ‘core’ of the Ricardian distribution and value theory. 1 Pasinetti spent the Academic year 1956-57 at Cambridge, U.K., as a PhD student; in 1957-58 he was at Harvard University , where he was also regis- tered as PhD student. At the end of that year, he was asked to chose only one option, and he decided to return to Cambridge, U.K. I got these details in private conversations with Luigi Pasinetti, who is gratefully acknowl- edged. I apologize for any possible error or inaccuracy. 2 Later on, in his Lectures (Pasinetti, 1977, chap. 1) he re-proposed this dis- tinction, essentially as a didactic device. 5
2. The mathematical formulation Consider an economic system composed of three classes: land- owners, capitalists and workers. Capitalists organize the production process by employing workers and the lands rented by land-owners. Two categories of goods are produced, which in the mathematical formulation are reduced to one good for each category: ‘necessary’ goods (say ‘corn’) used as wage-goods for worker, and ‘luxury’ goods (say ‘gold’). Both productions cycles take exactly one year and capital is constituted just by the wage-good advanced to workers. Lands used in corn production may differ in fertility. Capitalists be- have rationally and organize the production of corn on the various plots of land in order of decreasing fertility: lower quality land is used as the production of corn is increased. This technology can be represented by the function, Qc = f(Nc), (1) where Qc is the output of corn and Nc are the workers employed in corn production. Let us suppose that: i) f(0) ≥ 0, ii) f ′(Nc) > 0, iii) f ′′(Nc) < 0 and iv) f ′(1) > 0.3 Assumption i) is trivial; ii) and iii) to- gether reflect the fact that as corn production is extended the addi- tional product decreases as additional workers are employed on less and less fertile plots of land. We thus have decreasing returns to scale due to an extensive use of lands of varying quality.4 Assump- 3 Assumption ii) was not explicitly introduced by Pasinetti. Moreover, in (1960), assumption iii) takes the form f ′(0) > 0: “at least when the econom- ic system begins to operate and workers are employed on the most fertile piece of land, they must produce more that what is strictly necessary for their support”: Pasinetti, 1960, p. 82; the formulation of iii) used here is bor- rowed from Pasinetti (1977, equation (I.3.8-b)). 4 From the formal point of view, conditions ii) and iii) can also be inter- preted in intensive terms, according to the usual assumption of decreasing marginal returns of factors (see Pasinetti, 1977, p. 10, fn. 8). This interpre- tation of conditions ii) and iii) is not however very relevant for Ricardian analysis. Partially connected to this point, Morishima (1989, pp. 50-1) ob- jected that a unique corn production function, summarizing the input-output 6
tion iv) is a ‘viability’ condition: when the first worker is employed on the most fertile plot of land he must be able to produce more that his subsistence. Given the different qualities of lands, the owners of cultivated lands will be able to claim payment from capitalists, the rent. On each plot rent will at most be equal to the difference be- tween the corn produced on that plot and the corn produced on the least productive cultivated plot – called ‘marginal land’ – given by f '(Nc). Capitalists would have in fact the alternative of cultivating on the marginal land or worse lands, where cultivation is free. Total rents (R) are thus given by the difference between the total quantity of corn produced, f(Nc), and the corn that would be produced if all lands had the same fertility as that of the marginal land, Nc f '(Nc): R = f(Nc) – Nc f '(Nc). (2) Gold is produced by labour under constant returns to scale; its tech- nology is described by Qg = αNg, (3) where Qg is the output of gold, Ng are the workers employed in gold production and α is the quantity of gold produced by one worker. Also, the unit wage x is fixed at the subsistence level x on the basis of the Malthusian principle: x=x ; (4) x is not what is physiologically considered as the necessary mini- mum needed to support workers. It is the level that in a given country and in a specific stage of society keeps the population constant. Total wages (W) are: W = (Nc + Ng)x. (5) Capital (K) consists just in wages advanced to workers: K = W; (6) relation of all lands, could be used “to explain the rent of a land as the sur- plus which it yields”; Kurz and Salvadori (1992, § 3) argued in favour of Pasinetti’s approach. 7
the total amount of corn that can be advanced to workers is given: K= K . (7) Profits are determined as a surplus, that is, as the difference between the gross product (net of rents) and the ‘necessary consumption’ needed to repeat the production process year after year at least at an unchanged scale. In both industries, necessary consumption consists just of wages. Profits in the corn industry (Pc) are determined in physical terms, as the difference between homogeneous quantities of the same commodity (corn): Pc = (Qc − R) − Ncx. (8) On the contrary, profits in the gold industry (Pg) must be calculated in value, being the difference between amounts of heterogeneous commodities: Pg = pg Qg − pcNgx, (9) where pc and pg are the prices of corn and of gold. To explain what determines pc and pg, we must elaborate a theory of value. Coher- ently with the analysis developed in the Principles, Pasinetti resorts to a pure labour theory of value, according to which the value of each commodity (net of rent, if any) is equal to the quantity of labour required to produce it:5 5 Quite similar steps are proposed in (Pasinetti, 1977, p. 14): instead of in- troducing the labour theory of value through equations (10) and (11), Pasi- netti states: “Ricardo argues that what fundamentally determines the ‘value’ or ‘natural price’ of produced commodities is their cost of production”, that is, “wages plus profits at the ruling rate of profit” (see Pasinetti, 1977, p. 14). In formal terms: pc(Qc − R) = pcxNc(1 + r) (10′) pgQg = pcxNg(1 + r). (11′) Given these equations, Pasinetti just writes pgQg /Ng = pc(Qc − R)/Nc , 8
pc(Qc − R) = Nc, (10) pgQg = Ng. (11) The rate of profits of the system (r) is given by: pc Pc + p g Pg r= . (12) pc K We have thus 12 equations in 13 unknowns: Qc, Qg, Nc, Ng, R, x, W, K, Pc, Pg, pc, pg, r. The remaining degree of freedom is closed by formulating a theory of expenditure. Like Ricardo, Pasinetti supposes that rents are entirely spent on luxuries (with the exception of a neg- ligible part, not considered here). Hence pgQg = pcR. (13) This equation implicitly entails that the output of corn equals the in- comes of the other two classes taken together, i.e. profits plus wag- es.6 The configuration described by equations (1)-(13) is called by Pasinetti the ‘natural’ equilibrium of the Ricardian system. This configuration is actually a ‘moving’ equilibrium. In fact, besides a theory of distribution, a theory of value and a theory of demand, it is quite easy to enucleate a theory of growth from the economic system here considered. As capitalists save (the main part of) their profits and accumulate them into the stock of corn K available at the begin- ning of each period, a higher number of workers can be employed in which is obtained by dividing (10′) and (11′) by Ng and Nc respectively and by equalizing the left-hand side members. This equation (which equalizes the value of the product per worker – net of rents – of the two industries) together with pgα = 1 (which coincides with equation (11) after substituting (3)), replace equations (10) and (11). 6 Substituting equations (8) and (9) into (12) we obtain: rpcK = pc(Qc − R − xNc) + pgQg − pcxNg. Thanks to (5) and (13) one gets rpcK + pcW = pcQc. 9
both industries as time goes by. This entails the cultivation of more and more plots of land of increasingly lower quality. Rents thus augment, squeezing profits from a certain point on, till they are ze- roed out (or below that level which induces capital accumulation), with total wages increasing proportionally. The capital accumulation process thus stops and the system reaches the (Ricardian) stationary state. A simple way to study this process is to study the sign of the derivatives of the natural equilibrium value of the endogenous va- riables of the model with respect to K.7 Pasinetti obtains thus that all these magnitudes increase as a consequence of capital accumulation with just two exceptions: the rate of profits which decreases, and to- tal profits which increase at the beginning of the process of capital 7 More in general, Pasinetti points out that there are at least four dynamic processes at work in Ricardo’s system: i) capital mobility, which tends to equalize the rates of profits of the industries by channeling capital towards the most profitable industry; ii) a demographic dynamics, determined by the Malthusian mechanism, which pushes the wage rate to its natural level; iii) capital accumulation, as just described in the text, and iv) technical progress, which delays – without subverting – the convergence to the statio- nary state through shifts of the corn production function. In Pasinetti we do not find an explicit analysis of dynamics i) and iii): beyond temporary oscil- lations, a uniform rate of profits is supposed as permanently achieved on average, while technical progress is supposed by Ricardo to not alter the main conclusions of his analysis. More emphasis is devoted to processes ii) and iii) which require the setting of a truly dynamic system which is pre- sented by Pasinetti in the Appendix of his article. The convergence to the stationary state with a natural wage rate is formally proved there, even if from the economic point of view the description of the transient process, i.e. the sequence of ‘natural’ equilibria, is more relevant than the resting point of the process (the Ricardian steady state): “Ricardo, however, inves- tigates the properties of his system at a very particular stage of the whole movement, which he considers the relevant one. Most of the analysis is car- ried on as if the demographic mechanism has already fully worked through, while the capital accumulation process has not yet been completed” (Pasi- netti, 1960, p. 87). For this reason, the rough analysis based on the sign of partial derivatives proves to be more informative than the rigorous analysis in the Appendix. 10
accumulation and decrease when the system approaches the statio- nary state. 3. Principles versus Essay and the “early writings” System (1)-(13) is a stylized version of a Ricardian economy suitably simplified in such a way as to present the core of Ricardo’s Principles while avoiding all those complications that prevent him to provide univocal and rigorous results. The crucial assumption that makes this experiment possible is the supposition that just one com- modity (‘corn’) is both consumed and used as capital good. Not sur- prisingly, but quite interestingly, this simplification coincides with that adopted by Ricardo in the Essay (Ricardo, 1815) and in some other early writings which Sraffa (1951 and 1960) has allowed us to re-appraise for their theoretical content and insight. The crucial de- vice of these works is the substantial homogeneity between outputs and inputs (wage goods or, simply, ‘corn’) due to the primacy of agriculture. This homogeneity makes it possible to determine the rate of profits of agriculture in physical terms with the consequence that competition among capitalists will induce other industries to align their rates of profits to that obtained in agriculture. A simple re- styling of Pasinetti (1960) provides us with an analytical formulation of this distribution and value theory contained in Ricardo’s early writings which seems even more effective than the one provided by Pasinetti of Ricardo’s Principles. Let us replace the five equations (8)-(12) with the following four equations: Qc − R − xN c rc := , (8E) xN c p g Qg − pc xN g rg = , (9E) pc xN g rg = rc, (10E) pgα = 1 (11E) 11
(the letter (E) following the equation numbers stands for Essay and Early writings.) Equations (8E) and (9E) define the rates of profits of the two industries. The physical nature of the rate of profits of the corn industry emerges immediately: it is a ratio between quantities of corn, Qg, R, and xNc; after substituting (1) and (2) rc can be re- f ' ( Nc ) − x expressed as rc = . The rate of profits of the corn indus- x try can thus be known before the determination of prices. On the contrary, the rate of profits of the gold industry depends on prices. Capital mobility will tend to align the rate of profits of gold to the rate of profits of corn, as stated by equation (10E). Thus, the rate of profits of the entire system is f ' ( N c ) − x 1 − x / f '( N c ) r= ≡ , (14) x x / f '( N c ) where x/f ′(Nc) is the quantity of corn paid as wage to the amount of labour required to produce 1 unit of corn on the marginal land. This result echoes the famous ‘basic principle’ that “it is the profits of the farmer that regulate the profits of all other trades” contained both in Ricardo (1815) and in his correspondence in 1814 and early 1815 with other economists (see Sraffa, 1951, p. xxxi). This equalization takes place through suitable changes of the relative price of gold in terms of corn: after substituting (8E) and (9E) into (10E) and using (1), (2) and (3) one obtains: pg 1/ α = . (15) pc 1/ f ' ( Nc ) Lastly, equation (11E) fixes the quantity of gold produced by one worker (α) as the unit of account. Observe that 1/α and 1/f ′(Nc) ex- press the quantities of labour required to produce one unit of gold and one unit of corn on the marginal land, respectively. The labour theory of value does not enter here as an assumption, like in equa- tions (10) and (11), nor does it play a particular role in the theory here considered. It is just a consequence of the assumption that the 12
capital of both industries is constituted by a single commodity; capi- tal intensity is thus uniform between industries, xNc/Nc = xNg/Ng = x, and the labour theory of value holds.8 For further reference let us de- note by (E) the model constituted by equations (1)-(7), (8E)-(11E) and (13). It contains 12 equations in 12 unknowns: Qc, Qg, Nc, Ng, R, x, W, K, pc, pg, rc, rg. 4. The direct connection with Sraffa (1960) Model (E) constitutes the crucial link in the chain from Ri- cardo’s early writings to Sraffa (1951 and 1960). Sraffa’s works ap- pear thus as the generalization of Ricardo’s distribution and value theory contained in his early writings.9 The unacceptable restriction that corn was the only commodity required for its own production as well as for the production of all other commodities is now totally re- moved. By introducing some additional Assumptions in the Sraffa system we can see the direct connection with system (E). Ass. 1: wages paid ex-ante; the price system become thus pT = (1 + r)(pTA + wlT), (16) where p is the price vector, A is the input coefficient matrix, l is the direct labour coefficient vector and w is the money wage rate. Ass. 2: wages are constituted by a composite commodity represented by vec- tor x; then 8 Curiously enough, the assumption that capital consists of corn anticipated to workers only, leads us also to the opposite extreme of the labour theory of value i.e. the pure capital theory of value: by multiplying both the nu- merator and the denominator of the right-hand side of (15) by x we see that the relative price of gold in terms of corn is regulated also by the ratio of the quantities of capital-wage required to produce one unit of the two goods. 9 Note that Sraffa writes: ‘It should […] be stated that it was only when the Standard system and the distinction between basics and non-basics had emerged in the course of the present investigation that the above interpreta- tion of Ricardo’s theory [presented here in Section 3] suggested itself as a natural consequence’ (Sraffa, 1960, p. 93). A criticism of this Sraffian ‘in- terpretation’ of Ricardo’s theory is expressed by Porta (1986). 13
w = pTx. (17) Thanks to (17) the price system (16) can be re-written as pT = (1 + r)(pTA + pTxlT) = (1 + r)pTS, (18′) where S = A + xlT is the socio-technical matrix. In this case the en- suing rate of profit is 1 − λ*S r= , (19) λ*S where λ*S is the dominant eigenvalue of S. Ass. 3: commodities are just required as wage goods and not as capital goods, hence A = O; Ass. 4: wages are constituted by just one commodity, say commodity 1; then x = [x1, 0, …, 0]T, matrix S is reduced to a matrix having all zero entries except for in the first row, which has components x1lm, m = 1, …, M, where M is the number of commodities. Then λ*S = x1l1, and (19) collapses into 1 − x1l1 r= (19′) x1l1 which coincides with equation (14), as x1l1 still represents the quanti- ty of commodity 1 paid as wage to the amount of labour required to produce 1 unit of commodity 1. In this way the correspondence be- tween the Sraffa system and the model of Section 3 is complete. If some of the above assumptions are relaxed and we allow – as it is normal nowadays – that commodities are employed as capital goods and that wages enter as generalized purchasing power and are ex- pressed in terms of the Sraffa Standard commodity relation (19′) finds its correspondent in 1− w r= , (20) w + 1/ R where R is the uniform physical rate of surplus of basic commodities (see, for example, Bellino, 2004). Hence, relation (20) extends to the 14
general case of C (≤ M) basic commodities the idea (conveyed by (14)) that the rate of profits of a system can be expressed in purely physical terms as the surplus of the production process of basic commodities only.10 5. Concluding remarks Pasinetti’s formulation of the Ricardian system was originally con- ceived as a mathematical presentation of the basic structure of Ricar- do’s Principles. But the assumption that capital is constituted just by one commodity has two consequences: on the one hand, it establish- es a direct connection of the model with the logical structure de- scribed by Ricardo in his Essay, rather than in its Principles; on the other hand, it renders unnecessary the introduction of the labour theory of value as an assumption, as we have no necessity to measure aggregates of commodities with different compositions in order to calculate the rate of profits. This rate emerges as a ratio of physical quantities of corn for the entire economic system. This result is the point of departure of Sraffa’s generalization to any number of basic commodities. References Bellino E., (2004), “On Sraffa’s Standard commodity”, Cambridge Journal of Economics, Vol. 28, No.1, pp. 121-132. Kaldor N., (1955-56), “Alternative Theories of Distribution”, The Review of Economic Studies, Vol. XXIII, No. 2, pp. 83-100. 10 The same asymmetry also appears in other spheres: a technical improve- ment in the production of a basic commodity (of corn) spreads its effects to the entire system through the wage profit relation (20). If the improvement concerns a non-basic commodity (gold), it merely affects the relative price of non-basic products. The same can be said of a tax on basics or non- basics. Again, it is easy to prove that the prices of basic commodities de- termine the prices of both basic and non-basic commodities, while the pric- es of non-basics do not. 15
Kurz H. D. and Salvadori N. (1992), “Review Article – Morishima on Ricardo”, in Cambridge Journal of Economics, Vol. 16, No. 2, pp. 227-47. Morishima M., (1989), Ricardo’s Economics. A General Equilibrium Theory of Distribution and Growth, Cambridge University Press, Cambridge. Pasinetti L.L., (1960), “A Mathematical Formulation of the Ricar- dian System” The Reviev of Economic Studies, Vol. XXVII, No. 2, pp. 78-98. ⎯⎯⎯ (1977), Lectures in the Theory of Production, Macmillan, London. Porta P. (1986), “Understanding the significance of Piero Sraffa’s Standard commodity: a note on the Marxian notion of sur- plus”, History of Political Economy, Vol. 18, No. 3, pp. 443- 454. Ricardo D. (1815), An Essay on The Influence of a low Price of Corn on the Profits of Stock, John Murray, London; edition used: Sraffa (1951-73), Vol. IV, pp. 1-41. Ricardo D. (1817), On the Principles of Political Economy and Taxation, John Murray, London; edition used: Sraffa (1951- 73), Vol. I. Sraffa P. (1951-73 ed.), The Works and Correspondence of David Ricardo with the collaboration of Maurice H. Dobb, Cam- bridge University Press, Cambridge. Sraffa P. (1951), “Introduction”, in Sraffa (1951-73), Vol. I, pp. xiii−lxii. Sraffa P. (1960), Production of Commodities by means of Commodi- ties. Prelude to a Critique of Economic Theory, Cambridge University Press, Cambridge. 16
Istituto di Teoria economica e Metodi quantitativi Working Papers (dal 2007) 2007 44. GIANLUCA FEMMINIS E GIANMARIA MARTINI, Spillovers, disclo- sure lags, and incentive to innovate. Do oligopolies over-invest in R&D?, giugno 2007. 45. GIANLUCA FEMMINIS, From simple growth to numerical simulations: a primer in Dynamic Programming, settembre 2007. 46. FILIPPO GREGORINI, Political Geography and Income Inequalities, settembre 2007. 47. FERDINANDO COLOMBO E ALESSANDRA MAININI, The Control of Politicians in Modern Democracies: Discipline, Selection and Rent- Shrinking, settembre 2007. 48. ICONIO GARRÌ, Political Short-termism: A Possible Explanation, ottobre 2007. 49. ENRICO BELLINO E GIANPAOLO MARIUTTI, At the foundation of economics: a look at the Pandora’s box of the labour theory of value, no- vembre 2007. 2008 50. GIANLUCA FEMMINIS E GIANMARIA MARTINI, Irreversible R&D investment with inter-firm spillovers, settembre 2008. 51. ICONIO GARRÌ, Politician’s Reputation and Policy (Un)persistence, set- tembre 2008. 52. GIANLUCA FEMMINIS E GIANMARIA MARTINI, Production ex- tended research joint ventures and welfare, dicembre 2008. 2009 53. LUIGI FILIPPINI E GIANMARIA MARTINI, Strategic Choice between Process and Product Innovation under different Competitive Regimes, gennaio 2009. 54. LUIGI PASINETTI, Prospettive e Limiti dell’Economia quantitativa, gen- naio 2009. 55. GIANLUCA FEMMINIS E GIANMARIA MARTINI, First mover advan- tage in a dynamic duopoly with spillover, luglio 2009. 17
56. DANIELA PARISI, Public and Private Stances in Economic Policies. General Historical Notes on Social Services and the Specific Case of Italy in the first half of the XX Century, dicembre 2009. 57. LUCIANO BOGGIO, VINCENZO DALL’AGLIO, MARCO MAGNA- NI, On Labour Shares in Recent Decades: A Survey, dicembre 2009. 2012 58. ENRICO BELLINO, Pasinetti on Ricardo, Vita e Pensiero, gennaio 2012 (isbn 978-88-343-2180-5). 18
19
Printed by Gi&Gi srl - Triuggio (MB) January 2012 20
Università Cattolica del Sacro Cuore ISTITUTO DI TEORIA ECONOMICA E METODI QUANTITATIVI Pasinetti on Ricardo Enrico Bellino January 2012 ISBN 978-88-343-2180-5 VITA E PENSIERO
You can also read