The problem is that the marginally attached definition requires that you have actively applied in the last year.  There are a lot of people who would be in the labor force, but haven't applied in the last year.

Now this is at least to some degree a valid criticism.  A few points though:

The default outcome is two years, really, because in order to receive the two consecutive years of unemployment comp (thanks Bill!), you have to actively look for work.  The number of people who would refuse that is, I think we'd all agree, exceedingly small.

Once that runs out, you're going to run into very few people who have not applied for a job -- great ones, middling ones, lots of shit ones -- within a year-long period, because, simply put, they're going to be on the streets otherwise.  Nearly all in that scenario will wind up taking shitty part-time jobs and getting counted under the alternative measures.  Some older ones -- and this is true especially given the Boomer retirement we're going through -- will fold and take shittier retirements.  But most will be caught by the measure.

Be nice to America. Or we'll bring democracy to your country.

Eurogreen apparently had a similar thought to me.  The crisis in Europe hit the periphery, where you're going to find less development and (I suspect) more traditional, one-income households generally.  The hit to wages from the recession was also much more dramatic on the periphery than it was in the US or the EZ core.

The demographics are pretty similar.  Ignoring the crisis, you should actually see participation rising faster in the US, because stronger job growth here should induce more labor participation.  But incorporating the crisis into the analysis probably gives you a good chunk of the answer.

Be nice to America. Or we'll bring democracy to your country.