THIRLWALL LAW AND BALANCE OF PAYMENTS CONSTRAINED GROWTH - FURTHER COMMENT ON THE DEBATE

This paper reviews the debate that has occurred in Applied Economics over Thirlwall's Law (McCombie, 1989; McGregor and Swales, 1985; 1986; Thirlwall, 1986). The empirical evidence concerning the validity of the Law is re-assessed using a variety of tests including that suggested by McCombie (1989) and the new data set provided by Bairam (1988). The Law is overwhelmingly rejected by the evidence, contrary to the conclusions of McCombie and Bairam. The rationale for the assumption of constant relative prices which is required to derive Thirlwall's Law is reconsidered. The original 'law of one price' rationale renders the approach incoherent. The more recent rationale of sticky oligopolistic price improves the internal consistency of the model but necessitates consideration of non-price competition. However, neither the theory underlying Thirlwall's Law, nor empirical tests of the Law, properly account for this. In principle, the Law cannot be employed to distinguish between demanding constrained and supply-constrained theories of economic growth because it is consistent with both extreme Keynesian and neoclassical positions. It is argued that the supporters of the Law have not provided convincing grounds for preferring an extreme Keynesian interpretation appeal to other evidence. Finally, the relationship between Thirlwall's Law and the Harrod Trade multiplier is clarified. © 1991, Taylor & Francis Group, LLC. All rights reserved.