[Karunamon]

If you had stuck with that raw data I wouldn't have taken any issue with what you said. It's that "preferred" that stuck in my craw.

All this data says is exactly what the last line of your last post does. There is no "preference" stated here. The reasons why are not explored (and would probably be out of the scope of this paper anyways).

Keep in mind, our unemployment rate right now is absurdly high right now. I somehow doubt that's because people "prefer" to remain poor.

[yummyfajitas]

There is no "preference" stated here.

Without compelling evidence otherwise, I tend to subscribe to the theory of revealed preferences - our actions reveal what we really want.

But you are logically correct, there are some possibilities, e.g., the poor would prefer to have a job but irrationally don't look for one.

I somehow doubt that's because people "prefer" to remain poor.

You are looking at the wrong choice set. The choice set is not [ poor, middle class]. The choice set is [ (poor, leisure), (middle class, hard work) ]. The theory I'm pushing is that for some people, utility(leisure)-utility(hard work) > utility(middle class) - utility(poor). See my blog post for more details.

Also, this is not caused by the current recession. The numbers have been similar since 1996 (the year of the earliest report I can find with a quick google search).

http://www.bls.gov/cps/cpswp96.htm

http://www.bls.gov/cps/cpswp2002.pdf

http://www.bls.gov/cps/cpswp2006.pdf

(Tweak URLs to get data for intermediate years.)

[Karunamon]

>Without compelling evidence otherwise, I tend to subscribe to the theory of revealed preferences - our actions reveal what we really want.

It doesn't make sense to assume that people who are not working are doing so just out of a desire to avoid work.

>But you are logically correct, there are some possibilities, e.g., the poor would prefer to have a job but irrationally don't look for one.

Or there are no jobs within distance that pay the bills (i.e. what's the point of flipping burgers for minimum wage - it won't pay the rent and feed the kids). In which case it stops being an argument of utility and one of common sense - it makes none to work at a job which would be a pointless treadmill in which you stand a zero chance of ever getting back on your feet.

What bothers me is that you're assuming preference from raw data. This is a problem when the data does not support (or even attempt to) support any such conclusion - does it take into account local conditions?

If not, then any such analysis (including yours) is baseless.