Paul Krugman and Dean Baker took the Washington Post editorial page to task yesterday for stating that unfunded state and local pension liabilities amounted to $3.8 trillion. They accuse the page of misquoting a study in which the total was cited as only $1 trillion.
The WaPo editorial page did misquote the study, but that doesn’t change the fact that the $1 trillion is a completely mismeasured and fictional number. Unfunded state and local liabilities are around $4 trillion when the liabilities are correctly measured.
Since Professor Krugman’s post links to the study in question, by the Boston College Center for Retirement Research, I am sure he saw that the authors actually discuss these measurement issues at length. The Boston College authors even provide liability estimates under what they agree is a more appropriate methodology, and find that unfunded state and local liabilities are a multiple higher than the uncorrected $1 trillion.
Currently, standard practice measures the funding status of public pensions in the US under the laughable assumption that every dollar in the pension funds will earn compound returns of 7.75% or 8% per year. That’s the basis for the $1 trillion in unfunded liabilities.
But if a state or local government promises a risk-free pension, one that will be paid regardless of how the stock market performs, then that promise is like a government bond and should be measured accordingly. That’s the way pension promises are measured in most public or semi-public plans in countries like Canada and the Netherlands. In the Netherlands, for example, discount rates of 0-4% are used. Even US companies follow this basic principle that a pension is like a bond issued by the sponsor by treating their pension liabilities as corporate bonds for the purposes of their books.