The Minnesota Teachers Retirement Association website didn't bother with finesse when it blasted a recent Bloomberg article about the funding problems of Minnesota public employee plans. Bloomberg's sin was relying on the plan's own numbers under new pension accounting rules to "paint an incomplete and misleading picture."

That's what the debate about meeting the state's pension obligations has come to. We have a big retirement system saying its own accounting shouldn't really be relied on.

The teachers' retirement system (TRA) isn't alone in the state with an apparent funding problem, of course, as Minnesota jumped into the top 10 as ranked by Bloomberg by the size of state public employee pension underfunding. And TRA wasn't alone among state plan administrators in putting a reassurance on its website. But it was the TRA's numbers that seemed to get Bloomberg's attention.

In a conversation with J. Michael Stoffel of the TRA, his argument in favor of a different, actuarial report that showed better financial health was far more sophisticated than the spin on the TRA website. But he still wanted to tamp down any cause for worry. "The accounting [report] is just a snapshot in time," he said.

Well, that's true generally of financial reports. The year-end balance sheet for UnitedHealth Group reflected the close of business on Dec. 31, that's it. But that doesn't mean it doesn't reflect a lot of careful estimates of what could happen as UnitedHealth tried to fairly report the assets and liabilities on that date, like the value of claims it will someday have to pay.

There was certainly a big change in an important assumption for the teachers' plan. It went from about 77 percent funded as of mid-2015 to less than 45 percent funded in 2016.

What happened is that the accounting for the plan used a recent standard put out by the Governmental Accounting Standards Board, or GASB, the group that makes accounting rules for governments. Applying GASB made plans like the TRA use a different discount rate, a way of measuring the value of all the promises made to employees in their retirements. It's one tool to use when trying to answer the question, on any given date, of how much is really owed.

The discount rate is so fundamental in finance it even gets taught in high school classes now. It's used to figure out a value today for money that will get paid or received down the road. A $1,000 benefit check that needs to be paid in 2027 doesn't need to be thought of today as worth $1,000. The question is how much less that debt is worth, and a key consideration is how much uncertainty there is that the whole amount gets paid.

The accounting rules for companies' pensions have evolved, too. That's one reason 3M last used a discount rate of around 4.2 percent to put a value on what it owes for U.S. pensions.

Any debt owed in 10 years is worth a lot less if discounted at 8 percent than it would be if it were discounted at 4.2 percent. The only reason to discount it by 8 percent, though, is if there are doubts about whether it's really going to be paid. That's why auditors wouldn't let 3M get away with using 8 percent to value its U.S. pension ­liability if it tried.

The higher the discount rate, the higher the perceived uncertainty and the lower the current value.

There isn't much reason to think the state would let pensions go unpaid in 10 years, just like 3M seems good for its pension obligation. Yet that's not what the discount rate used in the past would tell us. The teachers retirement ­system in its 2015 report used 8 percent. Applied to the whole system's obligation it led to a far lower number than what that obligation seems to be really worth.

Not far enough

So GASB changed the standard, but unfortunately not quite in alignment with the corporate practice. The new discount rate assumption for the teachers plan, the one that helped drive Minnesota plans higher on the Bloomberg ranking of troubled state public pensions, was about 4.7 percent. The plan got to that rate with a methodology that seems impossible to describe in a newspaper. But it sure changed the picture, as the value of the liability shot from $26.6 billion to more than $43 billion.

What was already a sizable financial hole turned into a $23.9 billion crater.

You would think advocates for greater government transparency would be happy, yet Mark Haveman of the nonpartisan Minnesota Center for Fiscal Excellence said, "GASB absolutely blew it." He made a convincing argument that units of government still have far too much wiggle room under GASB for unrealistic pension plan assumptions.

There was even a government transparency worst-case scenario getting discussed among pension ­policy nerds last week, that this plan and others like it could have a couple of big investment years and be permitted, under GASB rules, to go back to using the unrealistic discount rates of the past, like 8 percent.

Yet there's hope that such a scenario won't come to pass, based on a quick conversation with Myron Frans, the commissioner of Minnesota Management and Budget, a principal policymaker for pensions.

"We recognize that GASB applies a different view," he said. "From the state's perspective, we want to encourage all the plans to move toward a reduced discount rate to more accurately reflect the liabilities and exposure we have going forward."

Based on years of watching state government, even a well-managed one like Minnesota's would likely ease into a change like this over the course of a few years to make changes in values less dramatic.

A good place for policymakers to start on the road to greater transparency is asking their plan administrators to please stop calling their own accounting incomplete and misleading.

lee.schafer@startribune.com • 612-673-4302