LONG HOURS, HIGH PAY? Greg Mankiw points out a new study showing that, in 2002, the top income quintile was twice as likely to work long hours as the bottom quintile. "That is," he writes, "wages and hours worked went from being negatively correlated to being positively correlated. This may be an important piece of the puzzle of rising income inequality." Possibly so. Of course, the bottom quintile are low-wage workers in jobs that rarely pay benefits and often keep employees in a sort of part-time twilight so they don't qualify for health care -- that may be a piece of the puzzle as well.

Also interesting, however, is recent research by Tom Hertz of American University who found that "[h]ouseholds whose adult members all worked more than 40 hours per week for two years in a row were more upwardly mobile in 1990-91 and 1997-98 than households who worked fewer hours. Yet this was not true in 2003-04, suggesting that people who work long hours on a consistent basis no longer appear to be able to generate much upward mobility for their families." So while the upper quintiles may work longer, it's not clear that more hours positively associate with income increases from workers further down the ladder. Laboring longer to get ahead may only work for those who are already there. Indeed, the really relevant data here may need to be broken down by profession. High-paying professions may force longer hours while lower-paying occupations don't reward -- or often even allow -- such work at all.

--Ezra Klein