The Triple-Parity Law

Scientists and epistemologists generally agree that a scientific law must be (a) relatively simple and (b) not contradicted by the available evidence. In this paper we propose and test one such law pertaining to international economics, the triple-parity law. It integrates three well-known equilibrium conditions, which are shown to prevail in the long run, on aver-age and ex post: (i) uncovered nominal interest rate parity (UIP); (ii) relative purchasing power parity (PPP); (iii) real interest rate parity (RIP). Using a cross-section of annual mean values or trend growth rates for 18 OECD countries in the post-Bretton-Woods/pre-EMU floating rate period (1976-1998) and employing a variety of single-equation and system estimation methods, we present robust evidence that the triple-parity law ultimately holds for large and diversified economies. For a few, mostly small and specialized countries, its working is however affected by some significant financial or real comparative (dis)advantages, for which estimates are provided. The law says nothing about short-term dynamics, but it can provide useful benchmarks in this context too, insofar as measures of the speed of convergence to long-run equilibrium are estimated. The triple-parity law, finally, illustrates another, rather fundamental point: if we look beyond short-term fluctuations and vagaries, economic laws do exist in the long run, just as economists used to think in the days of Marshall, Fisher, Walras and Pareto.