A Reuters story with the headline “U.S. states’ debt tops $4 trillion — report” caught my eye because I had never seen any analysis of the debt of states that put the number so high. The report the story referred to was issued by a mysterious non-profit organization called State Budget Solutions (SBS). The group’s website lists no physical address and says that SBS is affiliated with another non-profit, Sunshine Review of Alexandria, Virginia, whose Form 990 IRS filing declares that it has one employee. Furthermore, there is a third affiliated non-profit, Sunshine Standard, which also lists no information describing its funding or management. Considering that the three groups are non-profits focused on transparency and accountability for governments, their own behavior is pretty opaque. It’s also extremely odd that they don’t ask for donations on their websites. What non-profit doesn’t need donations?

SBS describes its mission in the following way:

“The State Budget Solutions Project is non-partisan, positive, pro-reform, proactive and anchored in fundamental-systemic solutions. The goal is to successfully engage political journalists/bloggers, state officials and opinion leaders in a new way of thinking about state government and budgets, fundamental reforms, transparency and accountability.”

These are excellent goals, but the information that SBS is disseminating seems slanted for political purposes. For instance, it breaks out the liabilities of all 50 states by type of debt and shows unfunded pension liabilities for states at about three times the more commonly cited figure of the Pew Center on the States.

To use another example, SBS lists unfunded pension liabilities as $2.86 trillion, while Pew’s total is $759 billion. That is a discrepancy of over $2 trillion. In one especially extreme case, SBS says that the pension liabilities for New Jersey amount to $144.8 billion, while Pew says they add up to $35.7 billion. It’s plausible that part of the difference arises from SBS using outdated pension data from 2010, before New Jersey reformed its pension system. If so, that would make its data much less credible, because Pew uses data from 2011. One thing I find really odd is that SBS has included the specific Pew data in its public data spreadsheet available via Google but has not made reference to it in its public statements.

Another example of widely divergent data between SBS and Pew relates to the unfunded liabilities of New York State, whose plan is generally considered one of the best funded in America. Although the plan has $150 billion in assets, SBS says that it is underfunded by $182 billion. Pew says the New York State plan is underfunded by $3 billion. Someone’s numbers are messed up.