Testimony to GASB Public Hearing, San Francisco, CA, 10/14/10

David Crane, Special Advisor to Governor Schwarzenegger

Good Morning, and thank you for allowing me to deliver some thoughts today about GASB’s Preliminary Views on Pension Accounting and Financial Reporting by Employers.

I wish to address just one portion of those views.  In short, I do not agree with your preliminary view of the rule for determining discount rates.

To begin, let’s review the way discount rates are now determined under your current standard.

Your current standard permits a state government like California’s, which has set up a pension fund as an intermediary for delivering pension payments to the employees to whom the state has promised those pensions, to discount its pension obligations to those employees using a rate equal to the return our state pension fund says it expects to earn on assets in that pension fund. Now, if our state had not set up a pension fund, we would have to use a lower discount rate for reporting those exact same obligations, producing a higher valuation.  In other words, your current standard allows for two different present values relating to two identical obligations, ostensibly a violation of the law of one price.

Now, that might be a legitimate outcome if our government is no longer on the hook for the pension promises once money is deposited into the pension fund.  That is, that might be a legitimate outcome if, by depositing money into that pension fund, our government legally defeases its obligation to make the pension payment so that the state has no further liability for that pension payment. In that case, the nature of the liability would have changed.  Before the deposit, it’s a hell-or-high-water undertaking of the State of California.  After the deposit, it’s an undertaking of the state pension fund. No longer are the two obligations identical.

But that’s not the way it works.  Whether or not our state contributes to a pension fund, or whether or not there’s enough money in that pension fund, our state is on the hook for the pension payment.  Our liability doesn’t change.  In fact, should our state pension fund lose every penny, the only people in the state not adversely affected by that loss are pensioners, because they have that hell-or-high-water undertaking from the state regardless of the pension fund’s status. Indeed, a pensioner due a pension payment is not only due an unconditional payment from the government but also is secured by whatever assets reside in the pension fund.  Thus, if anything, an even lower discount rate should be used to discount such a senior secured obligation. Yet, perversely, current GASB accounting allows governments to discount those senior secured obligations at a higher rate.  Go figure.

Accordingly, GASB’s current accounting standard is based upon a fiction, that fiction being that the state is allowed to book an obligation as if it isn’t on the hook even though it is on the hook. And according to the Center for Retirement Research at Boston College, state and local governments are understating pension liabilities by $2.5 trillion as a result of this fictitious accounting treatment.

To make matters worse, GASB’s current accounting standard perversely incentivizes our state pension fund to assume a high rate of return in order to minimize reported liabilities, and then to “swing for the fences” in investing the capital of those funds in the hopes of actually achieving those returns, adding risk for the taxpayers who are already on the hook for the pension payments. Indeed, the Chief Investment Officer of our state pension fund recently was quoted in Bloomberg as saying “Do I think it’s unrealistic to search for returns in [our expected return] range? No, I don’t. It’s really hard to imagine a worldwide market system not providing a return to riskier sources of capital.” (Bloomberg Business Week; italics added.)  Put another way, that official is effectively characterizing pension fund capital, which protects taxpayers from having to reach deeper into their pockets to meet hell-or-high-water pension obligations, as being in search of “riskier” investments.

I think you understand all this.  That is, it seems as if GASB understands that the state is on the hook for pension payments to employees and that whether or not it sets up a pension fund or if there’s a lot or a little money in that pension fund, the state remains on the hook to the employees for the pension payments.  But now, let’s look at what your Preliminary View would do to change this fictional accounting.

You now propose a single discount rate comprised of two rates, one using the expected rate of return on plan assets and the other using a municipal bond index.  The first would be applied to the extent current and future plan assets are projected to be sufficient to meet pension payments and the second would be applied beyond the point at which plan assets are projected to be depleted.

It’s hard to know where to start in addressing this Rube-Goldberg like contraption.  First, what is the justification for using the expected rate of return on assets in any capacity when determining the discount rate for purposes of reporting the size of unconditional liabilities?  One has nothing to do with the other in the case of unconditional undertakings such as pension obligations here in the State of California.  As I have just discussed, the state owes its pension payments whether or not it establishes pension funds or those pension funds have any money.  On what basis is it logical to suggest that it owes less?  Either we owe the money or we don’t.

Second, consider this perverse outcome generated by your Preliminary View. In a recent paper, Alicia Munnell, former member of President Clinton’s Council of Economic Advisors and now head of the Boston College Center for Retirement Research, wrote:

“In 1999, the California Public Employees’ Retirement System (CalPERS) reported that assets equaled 128 percent of liabilities, [after which] the California legislature enhanced the [pension] benefits of both current and future employees. If CalPERS liabilities had been valued at the riskless rate, the plan would have been only 88 percent funded. An accurate reporting of benefits to liabilities would avoid this type of expansion . . ..” (http://crr.bc.edu/briefs/valuing_liabilities_in_state_and_local_plans.html.)

Encouraged by that fictitious accounting, our state legislature enacted a law that year dramatically boosting pension promises, just as Munnell described. For your information, the cost to my state from that boost has already hit $15 billion and is expected to reach at least another $150 billion.

Now, how would your Preliminary View change things? From what I can tell, it would not materially change that outcome. This is because the municipal bond rate only kicks in when both the present and future assets, including future expected contributions, are projected to be insufficient to cover the liabilities. Thus, so far as I can tell, CalPERS would still be eligible to use its expected rate of return for discounting substantially if not all of its liabilities.  As a result, had your Preliminary View been in place in 1999, CalPERS would still have been able to report that it was over-funded.  What kind of accounting reform could be so at odds with financial and economic reality?

And third, your Preliminary View of the discount rate is at odds with what, by all appearances, you yourselves know to be true, which is that, whether or not there’s a pension fund and regardless of its funding status, the state is obligated to the employees for these pension promises.  In other words, it seems as if you’ve chosen a method that is at odds with the very concept you embrace.

When it comes to valuing fully-recourse obligations owed by a state government, I should think your job would be easy, because the only question you have to ask is “what’s the risk associated with the payment?”.  In the case of public pensions, the answer is clear.  As Donald Kohn, who recently retired after 40 years at the Federal Reserve, put it in a 2008 speech, “public pension benefits are essentially bullet-proof promises to pay.

“While economists are famous for disagreeing with each other on virtually every other conceivable issue, when it comes to this one there is no professional disagreement: The only appropriate way to calculate the present value of a very-low-risk liability is to use a very-low-risk discount rate. However, most public pension funds calculate the present value of their liabilities using the projected rate of return on the portfolio of assets as the discount rate. This practice makes little sense from an economic perspective. If they shift their portfolio into even riskier assets, does the value of the liabilities backed by their taxpayers go down? Financial economists would say no, but the conventional approach to pension accounting says yes.”

Likewise, Alicia Munnell writes that “[t]he argument is compelling that the liabilities of pub lic pension plans, which are guaranteed under state law, should be discounted by a rate that reflects their riskless nature.” In this sentiment Kohn and Munnell are also joined by academics from Wharton, Stanford, Northwestern and the University of Chicago.

It comes down to this: Either the state owes the money or it doesn’t. Until and unless its obligations are legally defeased via a contribution to a pension fund or some other technique, the answer is clear: it owes it.  We should account for it accordingly using a discount rate that reflects their riskless nature.

Let me close by making a plea. The value of a fully recourse obligation owed by our state should not be a complicated issue, yet somehow GASB has found a way to make it complicated.  I don’t know why that is, but my plea is simply that you adopt plain-language and common sense reforms that tell the truth about the size of these obligations.  These are the largest obligations owed by state and local governments and I can guarantee you that few legislators have any idea about them.  You must adopt clear and plain language accounting treatment so that laypeople understand them.

Finally, I would think current and future pensioners should be concerned about the under-accounting of pension promises.  It’s risky enough for them that, in California, voters do not approve pension promises at the ballot, unlike the manner in which they approve other forms of debt.  But the fact that pension promises are also not being honestly reported to them seems to me to be a substantial risk to pensioners when it comes to future generations honoring these obligations.  You must be aware that the cost of servicing all these unreported pension promises is growing sharply, cutting deeply into other programs.  There is an element of odiousness to all this given that citizens neither approved all these pension promises nor were notified of their size yet are suffering their consequences.  Were I a pensioner, I’d want to ensure there was full disclosure in order to give myself comfort that these promises will be honored.  I think it’s in everyone’s interest to require full disclosure.