Karl Smith offers some useful comments, stressing that Japanese unemployment does in fact seem high. My view is twofold:
1. Japanese output performance, while hardly stunning, is not as bad as the numbers make it appear. In a TGS age, the numbers will basically overstate the performance of health care + education economies (lots of rent-seeking counted as output) and understate the performance of export economies. Most exports are “real stuff,” and with some exceptions, such as arms sales, the buying is not driven by rent-seeking or agency-ridden third party payment. Japan of course is an export-based economy. Consider the possibility that the U.S. and Japanese growth rates have not been as different as they look in the published numbers.
By the way, regarding output per man hour, Japan has been about 70 percent of the U.S. level since the 1990s and not falling.
2. As Interfluidity points out, it is precisely the inefficient sectors of the U.S. economy which are the sources of whatever employment growth we have. If those sectors are smaller in your economy, reemployment will be correspondingly tougher.
Put this together and you get a common picture. The U.S. for a while has been more willing to absorb its displaced workers in the rent-seeking sectors and thus it looked more different from Japan than it really was. Some of that willingness has gone away, as voters have sought or tolerated cuts in state and local government spending,and other areas. Our published growth numbers decline and our measured unemployment increases, and so we look more like Japan, but it’s not as big a shift in regime as it might appear.
Oddly, it is Japan which comes off looking like the more transparent society.
What is the alternative to explanations along this track? That Keynesian nominal stickiness holds across time horizons of twenty years and up? Let’s turn to Scott Sumner again:
If you look at the Japanese unemployment rate you do see the normal ups and downs of the business cycle. You also see no change since 2000. There is no monetary model that I know of that suggests tight money could slow economic growth without raising unemployment. Thus although Japanese tight money might have slowed growth in the 1990s (when the unemployment rate trended upward), the recent slow growth should be due to non-monetary factors (unless the data is wrong.)
Japan is in any case full of puzzles.