Abstract
The author replicates three wage equations suggested by McCallum (197a,b), by Henry, Sawyer and smith (1976), and by Johnston (1972), using quarterly British data 1951(I)–1977(IV). The latter pair of equations yields fair results when tested by the usual methods. However, the application of a test of post fitting period parameter stablity leads to their rejection. (The first equation performs so badly by conventional criteria that the stronger test is not required.) The results point to the need (a) for more replication and (b) for testing the power of models to predict outside the fitting period.