The conventional interpretation of Keynes's General Theory follows Hicks in postulating rigidity of money wages; variations in output and employment then occur as changing demand alters the price level and so the real wage. There is no natural rate of unemployment; rather, unemployment will settle wherever demand settles. In Discussion Paper No. 262, Research Fellow Patrick Minford reinterprets this conventional wisdom. He claims that there is in the Keynesian model a form of natural rate of unemployment. Keynes's main policy prescription was that money demand should be expanded until unemployment reaches its `minimum' rate; this occurs when benefits have fallen so much in real terms that they are no longer an attractive alternative to work. Nevertheless, Minford argues, in the absence of such an expansion in money demand, the resulting unemployment can be considered a natural rate conditional on the real value of benefits.
If this interpretation is correct, then there is a basic continuity of thinking about the determinants of unemployment which runs from the Classical economists such as Pigou through Keynes and right up to present-day natural rate analyses. Minford attempts to integrate these ideas into a common framework, focusing on three main strands of previous work: first, the expectations-augmented Phillips curve due to Friedman and Phelps, which yielded the nonaccelerating-inflation rate of unemployment; second, the `cost-push' literature which stressed the role of union power; and third, work stressing the role of unemployment benefits in prolonging job search, especially Benjamin and Kochin's investigation of interwar UK unemployment.
The natural rate literature, Minford notes, requires above all a gently upward-sloping supply curve of labour. A flat-rate unemployment benefit entitlement, indefinitely available during periods of unemployment, is an ideal candidate to generate an increase in long-run labour supply elasticity. For as long as the average real wage falls relative to this benefit, people will withdraw from labour supply indefinitely. The benefit system is the basic cause of rigidity and hence of unemployment, Minford argues, since although explanations of unemployment can literally involve every permanent shock to the economy, the benefits system is the only necessary condition for positive labour supply elasticity.

Minford then investigates whether this analysis can explain the high and persistent unemployment in many European countries during the 1980s. Arguing that there are not yet convincing studies of many countries, he focuses on Britain, Germany and Belgium. Work on UK unemployment has identified significant effects on real wage costs from union power and taxes on employers. The contribution of benefit and income tax rates depends on whether actual productivity is included in the wage equation, but existing analyses reach broadly similar conclusions. In earlier work, Minford estimated the natural rate in Germany to be 1.2 million (about 5%) in 1980, with significant contributions from employer costs, union power and unemployment benefits. Estimates for Belgium reveal a natural rate of 340,000 or 8%, close to the actual rate for 1980; tax rates, union power and benefits are all significant.
Minford concludes that what has changed since the 1930s is not the theoretical framework of economic analyses of unemployment, but the institutional constraints; as these constraints have changed, policy recommendations have necessarily changed with them. The `monetary excesses' carried out in Keynes's name by postwar governments could thus be attributed to a failure to see through his policy recommendations of the 1930s to the basic mechanism beneath.

Wages and Unemployment Half a Century on Patrick Minford

Discussion Paper No. 262, August 1988 (IM/ATE)