The Macroeconomics of Downward Nominal Wage Rigidity : a Review of the Issues and New Evidence for Canada
Inflation definitely has costs, but they remain difficult to quantify for rates below 10-15 per cent. Aiming for a low rate of inflation, as Canada has done in the last 20 years, carries with it various risks, such as debt deflation, reduced flexibility of interest rates, and downwardly rigid nominal wages. In this paper, I focus on the latter problem. As James Tobin pointed out in 1972, economy-wide downward nominal wage rigidity generates a long-run negative relation between inflation and unemployment, implying that permanently low inflation can be bought only at the cost of permanently high unemployment. In contrast, the classical assumption of full wage flexibility implies that low inflation carries no permanent unemployment costs. The literature has produced compelling evidence that firm and worker resistance to nominal wage cuts is fierce, extensive and persistent in advanced economies, including Canada. But the macroeconomic significance of this phenomenon is still disputed. After presenting the details of the competing classical and Tobin views on wage flexibility or rigidity, I review the theoretical and empirical objections to the macroeconomic relevance of downward nominal wage rigidity that can be found in the literature. I then present new macroeconomic evidence on the matter based on Canadian macrodata for the 56-year period 1956-2011. This evidence suggests that the real world is more Tobin-like than classical, implying that the permanent unemployment costs of aiming for a low rate of inflation such as the current 2-per-cent target could be significant. According to the results I obtain, sticking to this target would keep the national unemployment rate between 1.0 and 2.7 percentage points in excess of the asymptotic minimum rate that would be attainable at a somewhat higher inflation rate.