Comparing Minimum Wages

Jordan Weissman has a piece at Slate about Seattle's $15 minimum wage. In it, he compares it to wages elsewhere in the world, including for McDonald's workers in Denmark:

Internationally, no country has an official minimum above the $11 mark once you adjust for purchasing power, according to the Organization for Economic Co-operation and Development. It’s often reported that fast food workers in Denmark can earn $20 per hour thanks to their strong labor unions. But take into account the nation’s eye-popping cost of living, and their wages are closer to $13.

The conclusions Weissman draws from these purchasing power parity (PPP) adjustments don't actually follow. PPP was designed initially to compare baskets of GDP across countries. PPP is also used sometimes to compare personal incomes across countries. There is some trouble with using PPP to compare personal incomes across countries because people do not spend their incomes on baskets of GDP. They spend them on baskets of consumer goods. This is why, in the US, we tend to adjust personal incomes across time using the Consumer Price Index rather than the GDP Deflator. But that's an aside really for my point here.

At best, the PPP, when used to compare minimum wages, gives you some sense of the value of the minimum wages to the worker as a consumer. That's an important thing to know. But looking at the value of wages to the worker as a consumer is not the same thing as trying to compare the costs of employing the worker to the employer. Thus, using PPP-adjusted wage levels to remark on the potential disemployment effects of a minimum wage does not make sense.

Productivity

If you believe in a roughly marginal productivity view of wage-setting (i.e. the worker is going to be paid their "value added" to the employer), then it shouldn't matter to the employer that the prices of goods in the area are more expensive. All else equal, I assume that a McDonald's worker in Texas adds as much value (to the employer) as a McDonald's worker in Alaska, even though the cost of living in Alaska is much higher because of transportation-related issues specific to Alaska. What about the higher cost of living in Alaska should mean that the employer will be able to pay the McDonald's worker more money?

The Alaskan worker will need to get more income to buy the same basket of goods as the Texas worker, but what does that have to do with the way that the employer sets the wage? In theory, nothing.

Using PPP adjustments to compare the ability of employers to pay certain wages doesn't seem to make a great deal of sense then. That holds for Alaska v. Texas just the same as US v. Denmark.

Tax Schemes

There is a more serious problem than the above though. Given that PPP is meant to track costs to consumers who actually go out and buy things, it has packed into it costs like sales taxes. If tax regimes differ considerably across countries, this will unravel the entire comparison for the purposes of trying to normalize the cost of certain wage levels to employers (which again is the operative question when talking about minimum wage and disemployment).

For example, imagine we are in country A, and it has a 10% income tax and no other tax. McDonald's workers in A make $20/hour. Now imagine we are in an otherwise identical country called B. In B, the McDonald's workers also make $20/hour. B has a 10% income tax as well, but it also has a 25% Value Added Tax (VAT), which is a kind of sales tax.

Consumer prices in country B will be 25% higher than consumer prices in country A. A simplistic PPP adjustment would therefore say that McDonald's workers in B make $16/hour versus the $20/hour made by Mcdonald's workers in A.

But you could not conclude from this that, to employers, the cost of employing a McDonald's worker in B is 25% lower than the cost of employing a McDonald's worker in A. The effect of VAT on prices is somewhat helpful when talking about the meaning of those identical $20 wages to the worker, but it does not tell us that A is risking more disemployment from its high McDonald's wages. They aren't higher to employers.

I explain the problem in hypothetical terms, but it's not a hypothetical issue. Denmark, the country Weissman uses to do PPP-adjusted wage comparisons between McDonald's workers there and here, has a 25% VAT. America has a 0% VAT.

Conclusion

You probably shouldn't use PPP adjustments to make this kind of point about wage levels and disemployment.

Comments