The Long-Run Behavior of Velocity: The Institutional Approach Revisited

In this paper we provide evidence using annual data for the period 1880 to 1986 that institutional variables are significant determinants of velocity in the United States, United Kingdom, Canada, Sweden and Norway. This evidence supplements our earlier findings (Bordo and Jonung, Cambridge University Press, 1987) for annual data ending in the early 1970's. We present eVidence that several proxies for institutional change in the financial sector are significant determinants of the long-run velocity function; that for the majority of countries the long-run velocity function incorporating institutional determinants has not undergone significant change over the last 10 to 15 years; and that out of sample forecasts over the last 10 to 15 years based on our institutional hypothesis are superior to those based on a benchmark long-run velocity function for a number of countries. these results suggests that failure to account for institutional change in the financial sector such as may be captured by our proxy variables may well be one factor behind the recently documented instability and decline in predictive power of short-run velocity models incorporating dynamic adjustment and higher frequency data.