Job Duration and the Macro-evaluation of Employment Policy: The Example of UK Regional Policy

Abstract

The paper argues that in order to use macro-evaluation techniques to calculate the employment effect of a policy over some sub-period after policy commencement, it is necessary to determine the rate at which policy jobs are lost. Crucially, it is the rate at which jobs cease to be attributable to policy which is relevant, rather than the rate at which jobs are lost to firms. This point is demonstrated for the case of UK regional policy over the period 1971-81. It is shown that the measured employment effect of this policy is highly sensitive to assumptions about the rate of policy jobs loss, but insensitive to the period over which jobs exist prior to being lost. Under reasonable assumptions, it is argued that the employment effect of this policy over the 1970s was twice that previously thought.