The velocity of circulation of money : empirical evidence for the United Kingdom 1870-1991

The equation of exchange is not in itself a theory of the demand for money. It can be argued that it is no more than an identity velocity determines the value of velocity. Given certain assumptions it can be a theory of the aggregate price level. One such supposition is that velocity is a constant, or at least a stable function of a few variables. Velocity over time is far from being a constant. Friedman argues that this is mainly due to errors of measurement and deviations between actual and desired velocity. Keynes suggests that there is no reason to believe that velocity is stable, and that in periods of underemployment equilibrium it may be quite volatile. He also proposes that velocity will depend on the structure of the economy, including the state of technology and institutional arrangements. The main aim of this thesis is to employ long time series data and up to date econometric techniques to produce evidence that relate to these two opposing views. The models employed use both income and transactions velocity measures. Transactions velocity has been much neglected in the twentieth century, on the grounds that a direct statistical measure is not available. This thesis attempts to resolve this problem by using archive material, sixty variables and seven thousand observations to construct an original transactions series for the period 1870-1991. The thesis traces the historical origins of the concept of velocity, provides a comprehensive and critical review of earlier work on the subject and produces a considerable amount of empirical work based on long term United Kingdom series of observation. The reported evidence using Johansen cointegration techniques, suggests that there is a long run vector between velocity and a few economic variables. However, the dynamic relationships are both unstable and volatile over the full sample period. Only in using sub-samples can satisfactory statistical results can be achieved