The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957

This paper tests, using nearly a century of British data (1861-1957), the hypothesis that the rate of change of money wage rates can be explained by the level of unemployment and the rate of change of unemployment, except in or immediately after years of a sufficiently rapid rise in import prices. Phillips reasons that when demand for labour is high and unemployment low, employers compete for scarce workers and bid wages up quickly, while when demand for labour is low and unemployment high, workers are reluctant to accept less than prevailing rates so wages fall only slowly – making the relation “highly non-linear” – and that wages also respond to whether unemployment is rising or falling, not just its level, since employers bid more vigorously in a year of improving business activity than in a year with the same average unemployment but no improvement. Fitting a curve of the form y + a = bx^c to a scatter of wage-change against unemployment for 1861-1913 (excluding the war and immediate post-war years), the paper finds this relation holds closely across most individual trade cycles between 1861 and 1913, largely holds up (with some cost-of-living-driven deviations attributable to import price shocks) through the disrupted 1913-1948 period, and – once a roughly seven-month lag between unemployment and wage response is introduced – again holds closely for 1948-1957, including a notable match to the sharp wage deceleration during the 1925-1929 return to the gold standard. Decomposing 1948-1957 wage changes into a “demand pull” component (predicted from the fitted unemployment relation) and a “cost push” component (from contemporaneous retail-price inflation, itself often import-price driven) identifies 1951-52 as a clear case of cost-push inflation following the 1949 sterling devaluation and the Korean War import-price shock, and 1950 and 1953-57 as episodes of “pure demand inflation” matching the fitted curve closely. Phillips concludes that, aside from rare years with a sufficiently sharp rise in import prices, the fitted relation implies unemployment of a little under 2.5 per cent would be consistent with stable product prices (given 2 per cent annual productivity growth), while unemployment of about 5.5 per cent would be consistent with stable wage rates, and that because the curve is strongly convex at low unemployment, holding unemployment constant at a given level yields a lower average rate of wage increase than allowing unemployment to fluctuate around that same level. He explicitly describes these conclusions as “tentative” and calls for further research linking unemployment, wages, prices, and productivity.

A. W. Phillips | Macro Paper Warehouse

The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957