It might be this one from the International Labor Organization (via Tim Fernholz):

The key chart is the one on the left. Since 2007, the ILO notes, very few developed countries have seen both strong job growth and an increase in "job quality." Instead, most have been stuck with one or the other.

The nations that have made the most inroads in reducing their unemployment, such as Germany or Israel, have often done so by adding lots of lower-quality jobs, including part-time gigs and jobs with lower benefits. "Average hourly wages decreased by 2.5 per cent in Israel; in Germany spending on social benefits decreased by a full percentage point between 2007 and 2011."

Meanwhile, the countries that saw an increase in the average quality of their jobs mainly did so because of higher unemployment. In Spain and Ireland, this happened because so many temporary and part-time jobs vanished, leading to soaring jobless rates. The United States is also grouped in this category — the average quality of jobs might be a bit higher than in 2007, but that's counteracted by the fact that there are fewer jobs to go around.

The outliers among advanced nations here are South Korea, Norway, and Poland, who have somehow managed to get both strong job growth and a boost in wage and benefits. No one else seems to have figured out their secret. (Not surprisingly, Greece got the worst of both worlds — average job quality has deteriorated significantly and unemployment soared.)

That all comes from the ILO's new "World of Work" report (pdf). And it's worth reading Tim Fernholz's discussion of the full report. Among other things, he points out that wage polarization is growing just about everywhere.

Update: Dean Baker has a lengthy criticism of this post, writing, "The intended take away is that if we want to have jobs then workers will have to take lower pay and fewer benefits."

No, that wasn't the intention. The chart, to me, suggests that the recovery has been dismal for most advanced nations, even those that have managed to bring down their unemployment rates—but there's simply not enough information to conclude too much more beyond that. Still, you can read Baker's points here.