“potential” be defined as the “trend” or “normal” levels of factor utilisation, a time-series
measure such as the Hodrick-Prescott filter may be used (Turner et al. 1996).
However, in this case the level of potential employment (N* ) is calculated as:
N* = LFS (1 - NAWRU) (2)
where: LFS = the smoothed labour force (the product of the economically active population and
the trend participation rate);
NAWRU = the estimated non-accelerating wage rate of unemployment.
The method adopted to measure the NAWRU (Pichelmann and Schuh, 1997) essentially assumes
that the change in wage inflation is proportional to the gap between actual unemployment and
the NAWRU (Elmeskov and MacFarlan, 1993). Assuming also that the NAWRU changes only
gradually over time,3 successive observations of the changes in inflation and actual
unemployment rates can then be used to calculate a time series corresponding to the implicit
value of the NAWRU. More specifically, it is assumed that the rate of change of wage inflation is
proportional to the gap between actual unemployment and the NAWRU, thus:
D2logW =-a(U -NAWRU), a > 0
where D is the first-difference operator and W and U are the real wage and unemployment rates,
respectively. Assuming the NAWRU to be constant between any two consecutive time periods,
an estimate of a can be calculated as:
(3)
This is based on the assumption of partial hysteresis: actual unemployment feeds only partly into
future equilibrium unemployment. In this case unemployment evolves only slowly towards its
steady-state level. In such a situation, the short-run NAWRU — meaning the level of
unemployment at which there is no current upwards or downwards pressure on inflation — always
lies between steady-state equilibrium unemployment and last period’s actual unemployment. This
carries the implication that high unemployment can only be slowly reduced to its long-run
equilibrium level if temporary increases in inflationary pressures are to be avoided (Pichelmann
and Schuh, 1997).