| Abstract: | This paper expresses econometric qualms about Bordo and Jonung's [1981] analysis of long-run velocity. They did not recognize that, for U.S. and Canadian data, the log of velocity has a unit root. Hence, estimation of a log level regression may produce spurious regressions.
When Bordo and Jonung's velocity equation is reestimated in rate of change form, permanent income is significant, contrary to their earlier conclusion. Moreover, using this approach gives a stronger result for one of the institutional variables in the velocity function, in the sense that the remaining variables become more significant. |
