Human capital is the most closely linked to government activities, as it is mainly an output of public educational investments, combined with student and parental investments. The public money we spend on teachers’ salaries helps produce high school graduates, many of whom become college graduates (indeed, mostly public-college graduates); these graduates, in turn, have been our country’s most important asset for growth since the mid-20th century.

In one leading model, gains from educational investment added $3.7 trillion to our national wealth in 2009, more than four times the value of public education as measured in G.D.P. that year. The return on educational investment is clearly enormous, yet it is nowhere to be found in the G.D.P. story of growth.

He just measured the return on educational investment, that $3.7 trillion, by evaluating the graduates' contribution to the GDP. I'm pretty sure his seemingly breakthrough logic is broken in that all of the govt's contributions to the economy-- acts like Clean Air and other nonmarket capitalization-- are intermediary actions that have no effect until what they're directed at beings to produce: graduates, manufacturers, etc. Those producers are private sector, the regular "traditional" factors in the evaluation of GDP, so I really don't think there's a mismeasurement at all. Maybe they haven't credited the government enough, which is the only claim that this evidence can back up really, but again govt investment that can be factored into the GDP always flows into the private sector anyway.

In other words:

The first problem is that much of what G.D.P. measures as personal, private spending — which counts for two-thirds to 70 percent of the economy — is highly socialized consumption. It is not financed directly from households’ private earnings, but from public sources or from publicly subsidized private sources.

Those publicly subsidized private sources are the primary factored-into's of the GDP. Therefore there is no discrepancy there, I just don't see what mismeasurement he's talking about.