Reprinted here is David Crane’s testimony yesterday before the Senate Public Employees and Retirement Committee on pension reform, SB 919, (Hollingsworth, R- Murrieta)
Good afternoon, Mr. Chairman, and thank you for allowing me to appear before the committee today on this important matter.
I will divide my remarks into five categories:
- The nature of pension promises
- The rising costs of meeting those promises
- The consequences of those costs
- The causes behind those costs
- What the state can do about it
Please note that I will focus just on the state’s pension costs associated with CalPERS, but in substance those comments also apply to state pension costs associated with CalSTRS, which this year will cost the state $1.2 billion and has announced it will be seeking more from the state. Also please note I will not be addressing the other principal retirement cost paid by the state, retiree healthcare, which this year will cost the state $1.3 billion and also is rising.
The Nature of Pension Promises
Pensions are a simple concept often made unduly complex, so to make sure we’re all on the same page, I’d like to start with a few basics.
When the state makes a pension promise to a state employee, it is simply promising to pay money to the employee at points in the future. Thus, state pension obligations are no different than state debt obligations, which also are promises to pay amounts in the future. But they differ in two important respects: (1) voter approval is not required for pension obligations — governors and legislators have all the responsibility in that regard, and (2) pension costs, unlike debt service costs, are neither capped nor precisely quantifiable in advance.
Pension promises translate into costs for the state budget at two times:
First, when the promise is made, at which time the state and the employee jointly make payments known as "contributions" to CalPERS with the expectation that the combination of those contributions and investment earnings on those contributions will meet the cost of the future pension payments promised to the employee.
Second, if and when additional payments are required in the event that combination is not sufficient to meet the promised pension payments. In that case, only the state is required to make up the deficiency.
It’s the second of those two items that makes our pension system a "defined benefit" system, because the state employee is assured of a defined pension benefit regardless of how well or poorly investments perform. And it’s also the second of those two items that explains in part how the state can end up with extra costs such as we’re seeing now.
The Rising Costs of Pension Promises
In 1999, CalPERS projected that pensions would cost the state budget an average of $450 million per year during the next eleven fiscal years from 2000 through 2011 , or a total of $5 billion.
But CalPERS will cost the state an average of $2.3 billion per year for that period , or a total of $25 billion. As you can tell, CalPERS was off in its projections by a factor of five. This resulted from CalPERS earning less than it assumed it would earn, forcing the state to make up the difference.
For the next twenty-nine fiscal years, CalPERS projects it will draw $9 billion per year on average, or $265 billion . That projection also is based on assumptions, a subject that we will come back to in a moment.
I have not seen CalPERS’s projections beyond the next 29 years. Because state pension promises extend well beyond 29 years, such projections are needed.
The Consequences of Rising Pension Costs
Pension payments are senior obligations of the state to its employees and accordingly have priority over every other expenditure except Proposition 98 (i.e., K-14) expenditures and arguably even before debt service. This senior priority of pension promises means two things:
- State employees face no risk that their pensions won’t be paid because the state’s promise to them not only is unconditional but also, as I just noted, senior in priority and, in addition, secured by CalPERS’s assets. We will come back to this item later when we discuss pension fund governance.
- All of the consequences of rising pension costs fall on the budgets for programs such as higher education, health & human services, parks & recreation and environmental protection that are junior in priority and therefore have their funding reduced whenever more money is needed to pay for pension costs. Thus those who should be most concerned about pension costs are families and businesses concerned about California’s colleges and universities, recipients of the state’s health and human services, users of state parks, citizens interested in environmental protection, and current and future state employees concerned about the potential for pay increases.
The Causes of Rising Pension Costs
Pension costs are rising for three reasons:
- Inadequate initial contributions
- Large retroactive pension increase enacted in 1999
- Debt Masking; i.e., the issuance of more pension debt than reported
Inadequate Initial Contributions
CalPERS has not been requiring adequate contributions when pension promises are made, virtually assuring deficiencies for which only the state is on the hook.
You will recall from my earlier remarks that both the state and its employees fund initial contributions but only the state picks up deficiencies. This means the state takes all the risk if initial contributions are inadequate.
Initial contributions are determined by investment return assumptions. For example, in 1999 CalPERS based pension contributions on an 8.25% annual investment return, which implicitly forecast that the Dow Jones Industrial Average would reach roughly 25,000 by 2009 and 28,000,000 by 2099 . As a result, in respect of an employee expected to retire in (say) 40 years with (say) a $75,000 pension, this meant CalPERS required an upfront contribution (jointly from the employer and the employee) of roughly $3,150. If CalPERS had assumed (say) a 6% return, the required upfront contribution (again, jointly from the employer and employee) would have been roughly $7,300.
Now, if the actual return ends up equaling the assumed return, all is well because no deficiency arises. But if CalPERS requires only $3,150 and then earns 6%, there will be a deficiency in the amount of $43,000, all of which the state must pay. Of course, the deficiency is even greater if PERS earns less than half of its investment return assumption, which has been the case since 1999. This delta between investment return assumption and actual return explains a portion of the state’s rising pension costs.
Since 1999 CalPERS has lowered its investment return assumption to 7.75% per annum, but that assumption remains high compared to others. For example, Warren Buffett assumes a much lower investment return for his defined benefit plans, even though his investment record is much better than CalPERS. In the preceding example, the upfront contribution required by Buffett would be nearly 40% higher than CalPERS requires.
CalPERS’s investment-return assumption also infects CalPERS’s projections of future costs for the state. For example, CalPERS’s projection that it will on average draw $9 billion per year from the state budget over the next 29 years implicitly requires the stock market to roughly double every ten years and to keep doubling at that rate. Anything less and the costs to the state will be larger. Even just a little less makes a huge difference in those costs.
For example, the state’s pension costs over the 11-year period discussed earlier were five times the amounts projected by CalPERS as a result of an actual investment return that was 43% of its assumed investment return. That disproportionate outcome – i.e., a 500% change in costs resulting from a 57% decline in investment return –demonstrates the immense power of cost compounding when it comes to long-term obligations. All of that risk falls on the state.
Even though the state has all of this risk, CalPERS refuses to disclose the information necessary for the state to be aware of and plan for those risks, such as sensitivity analyses. For now we know that, should the same thing happen over the next 29 years that happened over the past 11 years, the $9 billion per year currently projected could turn into $45 billion of cost per year for the state.
Note that this issue of inflated investment return assumptions is not a new matter that arose just because of the recent market decline. Warren Buffett wrote at length about inflated return assumptions in 1999 and early 2008 (see attached), and I regularly raised it when I was on CalSTRS’s board in 2005-06 in an effort to get districts to be more realistic about contribution levels.
1999 Pension Increase
In 1999, the legislature passed and sent to Governor Davis SB 400, a bill that embodied "The CalPERS Proposal," which I have brought with me. In that proposal, CalPERS advocated large and retroactive pension increases for active and new employees without requiring an increase in upfront contributions. Promising that "no increase over current employer contributions is needed for these benefit improvements," and that CalPERS would "remain fully funded" despite the increases, the CalPERS Proposal claimed that enhanced pensions would not cost taxpayers "a dime" because investment returns would cover the expense. Presumably in reliance upon those representations, the legislature passed SB 400 and Governor Davis signed the legislation shortly thereafter.
After all those representations, the 17-page document failed to disclose that:
- The state was on the hook for deficiencies should actual investment returns fall short of assumed investment returns;
- Those assumed investment returns implicitly required the Dow Jones to reach roughly 25,000 by 2009 and 28,000,000 by 2099 for no such deficiencies to arise;
- Potential costs to the state were uncapped;
- CalPERS’s own employees would benefit from the pension increases; and
- Members of CalPERS’s board had received campaign contributions from beneficiaries of the legislation.
The sales pitch even included testimonials from the state’s highest financial officer, State Treasurer Phil Angelides, as well as others who had received contributions from beneficiaries of the legislation.
It’s nothing short of astonishing that the CalPERS Proposal, which promoted the largest non-voter approved debt issuance in California history, was not accompanied by disclosures of risks or conflicts of interest. Frankly I’ve never seen anything like the CalPERS sales document, which makes even Goldman Sachs’s alleged non-disclosure look like child’s play.
This episode illustrates the perverse nature of CalPERS’s governance. The party with 100% of the risk and 0% of the upside – i.e., the citizenry – has no control over CalPERS. Indeed, what little influence the citizens have, which comes through the officials they elect and who sit on CalPERS’s board, is diluted because those officials can receive contributions from CalPERS’s beneficiaries (and others, such as money managers who earn fees from CalPERS).
In reviewing the legislative history, I note that some of the legislators on this committee served in the legislature when SB 400 was passed, so presumably they remember this event well. I can only hope that everything would have been different had CalPERS fully disclosed everything to you in 1999.
As we have seen, pension promises are liabilities in substance no different than the state’s debt liabilities. However, CalPERS has been reporting the state’s pension liabilities at a fraction of the size at which the state would have to report equivalent debt. It can do this because GASB, the Governmental Accounting Standards Board, currently permits public pension funds to assume a discount rate for reporting pension liabilities equal to the investment return assumption on pension fund assets. (The higher the discount rate, the lower the present value of reported liabilities.) Thus, public pension funds have yet another incentive to assume the highest possible investment-return assumption and as a result, legislators and other state officials are being misled about the true economic size of pension obligations.
CalPERS defends its practice by citing GASB, but in relying upon that crutch, CalPERS is taking a page from Lehman Brothers, AIG and Citibank, who similarly and permissibly gamed accounting rules to minimize reported debt – until it was too late. However, neither corporations nor CalPERS are limited to disclosing only what accounting rules require. New York City’s pension system, for example, discloses much more than what’s required under GASB (see attached), and many corporations provide much more information than what’s minimally required.
This too is not a new matter. I brought it up when on CalSTRS’s board, Buffett has regularly pointed out that a realistic assessment of pension liabilities would produce "staggering" results, and this issue was the subject of recent studies by the University of Chicago, Northwestern and Stanford that reach the same conclusion.
What Can Be Done
Nothing can be done about promises already made and pensions already underfunded. As a result, the state will be funding enormous deficiencies for decades to come. Indeed, we should all pray that CalPERS really and truly earns at its investment return assumption so that deficiency costs don’t grow beyond what’s now projected.
All we can do is (i) stop adding to our growing pension debt, (ii) require full disclosure to the real party at risk – the citizenry, (iii) prevent pension funds from utilizing excessive investment return assumptions, and (iv) change pension fund governance to reflect the citizens’ interests and to prohibit conflicts-of-interest.
In practice, this means (i) reducing the size of pension promises being made to new employees, (ii) adopting at least the New York City model of disclosure and making public all information necessary for independent analyses, (iii) governing investment-return assumptions, (iv) proposing a constitutional amendment to change CalPERS’s governance and (v) adopting municipal bond-type rules that ban business with anyone who contributes or provides gifts to board-members or staff.
Regarding the first item, SB 919 would appear to be an excellent move in the direction of reducing pension promises to new employees. Recently Illinois has enacted pension reform of a similar nature that extends retirement ages and vesting periods, reduces benefits, changes formulas, and caps pension-eligible compensation.
Regarding the second item, recently a Northwestern University professor projected that California’s pensions will run out of cash in 2026, the consequences of which would of course fall on the citizens. Yet they can’t know if that’s likely until and unless our pension funds provide all the information necessary for an independent analysis. As their representatives, you simply must demand full disclosure from state pension funds.
Regarding the fourth item, I recommend you take the advice of Girard Miller, who recently testified to the Little Hoover Commission about proper governance of public pension funds. In brief, I would encourage board dominance by independent members who are numerate and (as President Obama has characterized one of his criteria for his next Supreme Court nominee) who have "empathy" for those actually affected by pension fund performance, such as parents, students and healthcare recipients. In this regard, I have a special word for my fellow Democrats: One cannot both be a progressive and be opposed to pension reform. The math is irrefutable that the losers from excessive and unfunded pensions are precisely the programs progressive Democrats tend to applaud. Those programs are being driven out of existence by rising pension costs.
Regarding the fifth item, it’s not nearly enough just to make placement agents register as lobbyists, as proposed under some legislation currently in front of you. Instead you must make it clear that people who make contributions or provide gifts are banned from doing business with CalPERS. By the way, one of the greatest benefits from introducing such a rule is that CalPERS investment staff, which is very good, would finally be able to focus on its one-and-only job of importance: achieving the highest risk-adjusted yield at the lowest cost.
Finally, you will note that I have not addressed spiked pensions. While offensive and absolutely in need of reform, ending spiked pensions alone will not move the needle when it comes to addressing the pension problem. Similarly, I have not addressed recent allegations of fraud, which similarly are offensive and effectively are a direct assault on the state budget since the state has to make up all deficiencies, but avoiding every such fraud alone would likewise not move the needle.
Thank you for having me here today.