mercoledì 8 settembre 2010

Are average wages rising during the bust?

Stefan Karlsson on his blog claims that wage cuts would help speed up recovery and reduce unemployment, and that wages have been rising in the last years, despite the recession.

Data from Fed Fred, the datasets AHECONS, AHETPI, AHEMAN, e UNEMPLOY (average hourly earnings for construction, private sector and manufacturing, and total unemployed, respectively) show that the wages have really been on the rise.

However, there may be a large component of sample selection bias in the data.

People may have been fired in larger numbers among low-skilled workers than skilled ones, so that the average salary may have risen not because people are gaining more, but because the distribution of earnings has changed. This truncation in the data may cause an artifact.

Arguments in favor of my analysis:

  • It is usually the case that unemployment increases more among the poorer, or younger, workers. Maybe because of minimum wage laws, maybe because of the specific human capital embedded in skilled workers, which would hamper the company's productivity.
  • The effect has been stronger on construction workers, despite the crisis, than on average: but construction workers have also shrank in numbers much more than in other sectors, making the sample bias potentially more relevant.
  • Most of the increase has occurred when unemployment was rising, and the sample bias problem was thus stronger. Now the increase has settled.

However, the effect remains, so that real wages, instead of shrinking, are constant (or slightly rising: CPI growth is lower), and unemployment has become higher, and, what's worse from the worker's point of view, it lasts longer on average.

So, we are probably experiencing a little bit of Hooverian "don't cut wages" non-sense. But why? There shouldn't be strong unions in the US up to now, except in sectors such as steel or automotive. May it be that many workers have now, after the recession, lower than minimum wage productivity?