Pension Data Porn

I enjoyed the New York Times' Friday data mashup on public versus private earnings, though it's a shame they weren't about to come up with something by way of an estimate on pension/benefit compensation,  It does occur to me, however, that there is at least one problem with measuring pension/benefit compensation, both public and private, and that is for any calculation to be meaningful, you'd have to estimate the likelihood that the promised benefits will actually be paid.  I don't think that the remotest likelihood that public entities will be able to honor their more fanciful promises to teachers, prison guards, fire fighters, etc.  Perhaps a good rule of thumb (though imperfect) would be to scale the number down to the current level of funding.  This isn't perfect: you need to make all kinds of guesstimates, not least as to the appropriate interest rate (and I sign on with those who believe tht the rates currently used most places are too high).  You'd want to do the same with private sector obligations, of course, but since the private sector has pretty much killed off its defined-benefit obligations, it shouldn't be much of a problem.