Zero Marginal Product Workers?

Tyler Cowen note the zero marginal product worker idea (and here) is "gaining wider currency" with an article from the Richmond Fed, which effectively seems to be saying 'See! These workers are unemployable, so easing monetary policy via the Fed won't help!'.

I confess, this disturbs me because I don't think the ZMP idea makes much sense for several obvious reasons.

First, it seems targeted to describing unemployment in skilled workers, but the lion's share of unemployment is amongst unskilled workers (estimates are running at 5% unemployment for experienced skilled workers, about 10% for new graduates and 13-15% for unskilled workers if I remember correctly).

Second, the whole idea of MP of labor is tied to marginal cost of labor. In English, workers are hired not based on what they produce alone but what it costs to get that production. And costs, especially for the unskilled workers, have increased dramatically, with effective minimum wages estimated to reach over $13/hr as the health care law is phased in.

Thus, it seems like the lion's share of unemployment can be explained by the obvious elephant in the room, lack of demand, coupled with a greatly increased marginal cost for hiring cheaper workers.

At the very least, it puts paid to this argument from the Fed. If the Fed can raise demand, it will make it more profitable for companies to hire workers, and decrease the relative costs of doing so. Hiring is always a relative calculation, and right now costs are relatively high and benefits aren't obvious since companies aren't exactly swimming in work.

Anecdotal to this, I do see structural problems, but I don't know if this or any comprehensive theory describes it well yet.

• I see lots of comments from the unemployed but highly skilled blaming HR departments and I think there's some truth to that. On the other hand, that also looks like something that could easily be pushed by the wayside if there were a truly high demand out there. If companies had projects to work on, they'd hire.
• I think we underestimate just how many skilled workers for private companies effectively work exclusively on government contracts. These are not entrepreneurial businesses in the sense we think of as economists. Our basic model is that a company develops and idea and then sells it. A car company, for instance, hires designers to build a car, and employs workers to build it. It's a speculative venture, and when (and if) the business takes off, more employees are hired. A privately held bureaucratic business that wins long term contracts from the government doesn't do this.  They employ a few people to get the work along with a core group of people, but there's not really cause for much additional hiring if they don't win a contract.
• Similarly, the requirements put on companies working closely with governments are often onerous with respect to who they hire.