The turnover-reducing effects of minimum wage hikes might be undesirable

I have just uploaded a new working paper. Although this one is not about economic history, it does concern an issue I have to deal very often in economic history: wages and price controls.

In this working paper, I decided to tackle the issue of the minimum wage and the empirical finding whereby there are only small or even insignificant negative effects on employment from hikes in the minimum. The argument generally advanced is that the minimum wage actually solves a coordination failure. Since there are important search costs for employees and employers and important costs in training and firing, a large number of differing wage rates for unskilled workers creates a coordination failure where there are numerous unfilled vacancies and high turnover. In the presence of a minimum wage, these frictions are minimized and there are gains in employment. At least, that is the argument.

For the sake of argument, I decide to accept the empirical finding of the absence of discernible effects on employment resulting from minimum wage hikes. I argue that, if the relation between turnover and productivity is not linear, the effects of the minimum wage will be observed on productivity rather than employment.

As time passes, an unskilled worker acquires some forms of human capital through experience and soft skills. Once he has accumulated this human capital, he becomes more attractive for all firms. However, some firms might be better matches for this employee than other firms. Productivity is only optimized if this worker is matched with the employers who value him the most. Normally, employers would attract this worker by offering higher wages with the highest wage offer representing the best potential match between this employee and employer. The high wage rate, commensurate with skills and experience acquired, would incite a higher turnover in the sector with unskilled workers. However, the minimum wage increases the threshold of wage offers that other sectors must surpass to incite the employees of the first sector to change employer. Thus, the minimum wage creates a coordination failure leading to labor misallocation that generates suboptimal productivity.

In short, the argument that the minimum has small negative effects on employment does not mean that there are small overall effects on the economy. Defenders of minimum wage hikes must not rejoice at their finding as the mountain they must climb is still dangerous.

Full paper : HERE 

Unemployment in Canada after recessions: the 1920s versus the 2000s

While doing research on other issues, I found an easily compilable dataset of unemployment rates for trades on a monthly basis in Canada with provincial breakdowns. This gave me the idea to see how the different regions of Canada dealt with recessions with regards to unemployment.

The graph below is a very simple coefficient of variation of unemployment rates across provinces. They represent the months of December 1918 to December 1924 (in red) and the months of December 2007 to December 2013 (in blue). The data for the 1920s is of lesser quality than the data for the late 2000s and early 2010s, but still it gives a good idea of how recessions are “shared” in Canada.  In the 1920s, we can clearly see that the recession was felt more deeply in some regions than others. This is less the case in our days. However, the 1920s recession  was one where aggregate supply shocks provided a strong explanation (see here). Still, the difference is marked and it indicates that recessions in Canada are now more “well shared” than was the case in the 1920s.  It indicates also the provincial economies of Canada are now better integrated than was the case in the 1920s (this is a topic of debate in the literature).


In a few weeks, I will post data comparing the same things but with the 1930s Great Depression.