The Economic Unconscious
Posted by Graham Joncas
“One approach to determining whether the disincentive effect of income taxation on labor supply is important is to study the behaviour of workers in different countries with different income tax systems and see what we observe. In an article in the Minneapolis Federal Reserve Bank Quarterly Review,1 Edward Prescott does just that. Prescott studies the G-7 countries, which are Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States, and examines two periods of time, which are 1970-1974 and 1993-6. One key feature of the data that Prescott focuses on is that, in the earlier period, labour supplied per person was about the same in France, Germany, and Italy as it was in the United States. However, by the 1990s, Americans were working about twice as hard (in hours worked per person) as their counterparts in France, Germany, and Italy.
What could account for this difference in labour supply behaviour between the United States and Europe? Possibly Europeans just like vacations more than Americans do (indifference curves are different in the U.S. from Europe), or market wages (before taxes) are higher in the United States than in Europe. Prescott argues that the major reason for this difference in labour supply behaviour is differences in marginal tax rates. That is, in real-world income tax systems, typically the tax rate varies with the level of income. For example, in Canada and the United States the poor are taxed at a lower rate than are the rich. The marginal tax rate is the tax rate paid on the last dollar of income that a person earns, and this is the tax rate that matters for that person’s behaviours. In France, Germany, and Italy, in the period 1993-1996, the marginal tax rate for the average person was about 60%, compared to 40% in the United States. For this to matter for labour supply behaviour, it would have to be the case that there is a very large substitution effect[*] of a change in the real wage on the quantity of labour supplied, relative to the income effect[**]. Prescott measures these effects and finds that, indeed, the measured substitution effect is very large.
How could Prescott’s findings matter to us as economists? One example he uses relates to social security systems. […] The type of social security system in place in the United States is a ‘pay-as-you-go’ system, which funds payments to retirees from taxes collected from the working-age population. If Prescott’s results are correct, economic efficiency could be improved greatly could be improved greatly by moving from pay-as-you-go to an alternative social security system that does not tax workers at such high rates, thus increasing the quantity of labour supplied. The Canada Pension Plan (CPP), which is Canada’s social security system, is actually a hybrid system that incorporates features of pay-as-you-go and so-called fully funded plans…. Prescott’s analysis would tend to imply that the CPP is more economically efficient than U.S. Social Security, as it does less to discourage labour supply.”
Williamson, S. (2010). Macroeconomics, 3rd Ed. Toronto: Pearson Education Canada, pg. 108-9.
*Substitution Effect – The effect of a change in price of a good or service on the quantity bought when the consumer (hypothetically) remains indifferent between the original and the new consumption situations―that is, the consumer remains on the same indifference curve.
**Income Effect – The effect of a change in income on consumption, other things remaining the same.