The Prebisch-Singer Terms of Trade Hypothesis: Some (Very) New Evidence

Abstract

The chapters in this section of the book are concerned with various dimensions of an hypothesis which has become inextricably associated with the names Hans Singer and Raul Prebisch. According to this hypothesis, which was launched simultaneously by Singer (1950) and Prebisch (1950), the net barter terms of trade between primary products and manufactures have been, and could be expected to continue to be, subject to a downward long-run trend. Being in direct contradiction with the then prevailing orthodoxy, it is not surprising that the Prebisch-Singer hypothesis (P–S hereafter) attracted criticism from a number of quarters. The ensuing debate, which initially focused its attack on the basis of issues related to the treatment of transport costs and quality change, is well summarized by Spraos (1980), who showed that adjustments for shipping costs and changing quality left the hypothesis largely undented, in the sense that they failed to destroy its empirical validity. However, since the mid-1980s the debate surrounding the P–S hypothesis has shifted to the statistical arena. Indeed, such is the interest generated by the hypothesis amongst econometricians and time-series statisticians that it has established itself as one of the major test beds on which they routinely evaluate their latest methods of trend estimation!