Michael D. Bordo, Rutgers and NBER
Ronald MacDonald, University of Strathclyde

We are grateful to Pierre Sicsic and Marc Flandreau for their helpful comments on a preliminary draft of this paper. We are also grateful to Andrew Davies and Gillian Maciver for their excellent research assistance.

This paper examines the recently-noted finding that the Classical gold standard represented a credible, well-behaved, target zone system from the perspective of the well-documented failure of countries today by the rules of the game in the classical period. In particular, we test an hypothesis of Svennson (1994) that a credible target zone can confer on a country a degree of independence in the operation of its monetary policy. We propose a number of ways of testing this proposition and implement them for a newly-created monthly data base over the period 1880-1913. We demonstrate that the Classical gold standard worked in the way predicted by SvennsonÕs model. This would seem to have an important bearing on the kind of institutional framework required for a modern day target zone (such as the exchange rate mechanism) to function effectively and, in particular, to weather speculative attacks.