Question

Describe the main explanations for the downward rigidity of wages in the modern macroeconomy. Evaluate their probability of being correct and important.

Answer 1

In the rest of this paper, I study the type of money illusion envisaged by James Tobin, that is, downward nominal wage rigidity. I will be mainly concerned with its presence and relevance in the Canadian economy. There are three questions to be addressed: 1) why workers and firms strongly resist money wage cuts, 2) how important and persistent this resistance actually is in advanced economies, and 3) how macroeconomically significant this phenomenon is, particularly in Canada. Answers to the first two questions will come easily, but there is more uncertainty concerning the problem of macroeconomic relevance. I will therefore spend more time on this third question and look for additional evidence in the Canadian context. The first question to be dealt with is why money wage rollbacks should be relatively rare. Building on previous research (e.g., Kahneman, Knetsch, and Thaler 1986) and adding his own survey evidence, Truman Bewley (1999) has provided a clear answer to this question. Puzzling over the fact that nominal wages had rarely been cut during the 1991 recession in the United States, Bewley interviewed hundreds of managers, labour leaders, professional recruiters and advisors to the unemployed in 1992 and 1993, and asked them why. The reason most often reported to him was that workers regard pay cuts as totally unfair. As a result, firms would be reluctant to cut wages and to replace existing workers by others, willing to work for less, for fear of losing their good employees and suffering a serious drop in morale and productivity among their remaining employees. They would rather cut jobs than wages. The second question is how extensive and persistent downward nominal wage rigidity actually is in advanced economies. Again, a clear answer is on hand. It comes from the International Wage Flexibility Project, which involved 29 researchers from 14 countries in the early 2000s (Dickens, Goette, Groshen, Holden, Messina, Schweitzer, Turunen and Ward 2007). The participants scrutinized 360 annual wage change distributions from 31 different datasets covering 16 OECD countries over the period 1972-2003. The Project’s estimate of labour market heterogeneity relied on the average standard deviation of measured nominal wage changes across all datasets. After correcting for measurement errors, it was found to be 7.7 percentage points.