Though the labour market has always been central to macroeconomics, policy has usually had no instrument for intervening directly in the wage-setting process. But in the mid-twentieth century, economists commonly believed that there should be such an instrument. In 1950s Australia, it seemed that the arbitration system could potentially be used as such. This article uses Trevor Swan’s contemporary model of Australian policy in the 1950s to understand the tensions facing policy and explain why wage control seemed to be a solution. The arbitration judges began to consider macroeconomics in their decisions, and the unions adapted by presenting macroeconomic arguments of their own. In full employment conditions, labour had considerable economic power outside the tribunal, and the limitations of arbitration as an instrument of policy raised the shadow of unemployment as an alternative disciplining device.