Geographic dispersion

Exchange-market integration has at least two dimensions: geographic and temporal. The geographic approach to integration is based on the decentralized nature of the early American foreign-exchange market. The exchange rate is compared by space and over time, with enhanced integration reflected in reduced dispersion of the rate in the various cities of quotation. Investigations of this nature, obvious though they may be, do not exist in the literature. The approach can be applied to the exchange-rate series developed in chapter 6 to derive an important conclusion.

The most fruitful comparison is between the Baltimore (White) and Boston series, for which a continuous overlap runs from 1795 to 1829. The mean of the absolute difference between the two series (from table 6.12), by period, is as follows: 1795–1800, 1.39 percent of parity; 1801–10, 1.21; 1811–20, 1.22; 1821–9, 0.36. The value for the 1820s is less than 30 percent of that in earlier decades. The implication is that there was a strong movement to exchange-market integration in the 1820s.

Further, in that decade the absolute level of integration was high, and uniformly so across all port cities. The mean absolute difference between both the White versus New York and the White versus Philadelphia exchange-rate series is 0.28 percent of parity.