Tag Archives: Andrew Biggs

GASB’s new guidance and the well-funded plan

As of June 2012, GASB has put forth two new accounting guidelines to help value public sector pension plans. These are GASB 67 and GASB 68. These rules help government actuaries to calculate the value of plan assets and plan liabilities. The new rules are a replacement of GASB 25 and GASB 27. The former guidance – GASB 25 –  has been roundly critiqued by economists for conflating assets and liabilities for the purposes of valuation – a violation of several established principles of economics and finance. The main critique of GASB 25 has been covered many times. The old guidance allowed public sector pension plans to chose a discount rate to value pension plans liabilities based on the expected returns of plan assets – roughly 8 percent annually. The critique of economists is basically this. The value of the liability is independent from the value of the assets. How the liability is financed is independent from how it is valued. The discount rate that should be used to value the liability should be based on the characteristics of the liability. Public plans should be valued according to their relative safety (or risk) as government-guranteed payments to workers. Economists suggest the rate on Treasury bonds is a good choice. Using the expected return on assets is logically misguided and leads to all kinds of trouble – plan underfunding, diminished contributions, more risk taking on the investment side. Will GASB 67 and GASB 68 fix this? No. According to the new standards (which are only for reporting purposes), plans will apply two different discount rates to calculate plan liabilities. To the funded portion (the portion backed by assets)  the assumed rate of return on plan assets will be used. For the unfunded portion plans will use the yield on municipal bonds. Andrew Biggs notes in a recent paper, “the logic is precisely backwards.”And further, the new standards,

…. cement in place the flawed notion that boosting investment risk makes a pension better funded, before a dime of higher returns have been realized. Under the current rules, a pension that shifts to riskier investments can discount its liabilities using a higher interest rate. Under the new rules, a plan that takes greater investment risk can assume its trust funds will last longer and therefore fewer years of benefits would be discounted using lower municipal bond rates. The incentives to take greater investment risk, particularly at a time when state and local governments would be hard-­‐pressed to increase pension funding, are obvious.

How will the new GASB standards affect plans individually? Alicia Munnell and her co-authors at the Center for Retirement Research at Boston College have calculated that. Well-funded plans look pretty good. Consider Delaware. Under GASB 25 Delaware’s main pension plan is 94 percent funded with an unfunded liability of $456 million. Using GASB’s new guidance – the blended rate of 8% – the state employees’ plan is 83 percent funded. And here is my rough estimate of the same plan using market valuation. Using a discount rate of 3.6 percent (the yield on 10 and 20 year Treasury bonds in 2011 when the valuation was performed) Delaware’s State Employees’ Plan is 51 percent funded and it has an unfunded liability of $6.9 billion. That also means the normal cost for the employer to fund employee benefits rises from 9.74 percent of payroll or $125 million a year to 12 percent of payroll or $216 million per year On the asset side Delaware is a leader in shifting its investment portfolio to riskier investments. Between 2002 and 2011 Delaware increased its exposure to alternatives from 9 percent to 24 percent. This puts Delaware in fifth place (in 2009) for the percentage of pension assets invested in alternatives. But with higher returns comes more risk, and that is something the new accounting guidance still does not adequately account for.


Hearing tomorrow on state and municipal debt

The House Committee on Oversight and Government Reform will hold a hearing tomorrow on state and muncipal debt. Witnesses include Governor Scott Walker of Wisconsin, Robert Novy-Marx of the University of Rochester, Andrew Biggs of AEI, Mark Mix of the Right to Work Legal Defense Foundation, and Desmond Lachman of AEI. The hearing is sceduled for 9:30 AM and will be broadcast on C-SPAN.

Yes, the Government is Long-Lived

In the debate over how to value public sector pension liabilities an argument that seems to impress some people is this. Public sector pensions can guarantee their pensions with risky assets because the government never goes out of business. Andrew Biggs has written several responses to this notion. The latest piece to make this claim appeared in The Weekly Standard. Andrew and I offer a refutation.

The recurring theme that the government can value pension liabilities as though they were risky because it is long-lived is an oxymoron.  It is precisely because the government is unlikely to go out of business that public sector pensions are considered guaranteed. That is why economists suggest using a risk-free discount rate to value public pension obligations. It’s because of the government guarantee – the government is more likely, not less likely, to pay it. Current public pension accounting implies otherwise. The subtext in defense of this flawed accounting is, “If all else fails, raise taxes.”

“The risk is borne by the taxpayer” is the missing subtitle to the headline, “The government is long-lived.”

Veronique deRugy offers a fantastic analysis of this debate here. Followed by Yuval Levin here.

What Does An Artificially High Discount Rate Mean?

The appropriate discount rate to be used in valuing public pension liabilities might seem like an arcane and technical question. I suppose it is. But thanks in large part to the indefatigable work of Eileen, her coauthor Andrew Biggs, and professors such as Novy-Marx and Rauh, it is a technical debate that has actually been covered by mainstream media outlets. In short, the official actuaries who value these liabilities use what is known as a high discount rate while most economists—Eileen included—believe that it is more prudent to use a low discount rate. The choice makes a big difference: with the high rate, state public pension systems appear to be underfunded to the tune of about $1 trillion; while with a low rate, they are underfunded by about $3 trillion.

What, exactly, does a low discount rate mean? More importantly: what does a low rate say about the assumptions of those who prefer to use it?

To get at this question, we first need to talk about why anyone uses a discount rate in the first place. The rationale has to do with the fact that the big costs of public pension systems will not be felt for a number of years (interestingly, some have looked at the modest current costs of pensions and concluded that state finances are perfectly healthy; but that is like looking at my child’s current college tuition expenses—which are $0—and pretending that her education isn’t going to be expensive). In any case, the big costs will not be felt for a number of years, and so we need some way to make sense of future costs in today’s context.

The way economists (and accountants, and financial analysts) do this is with a discount rate. This is a number that helps us figure out what we need to put aside today in order to pay for a liability down the road. We call it a “discount” rate because it discounts future costs in order to make an apples-to-apples comparison with current costs. In other words, paying $50,000 in tuition 18 years from now is not as difficult as paying $50,000 in tuition today (because I have 18 years to save for it). So to make sense of that future cost, I discount it. And the larger the discount factor I use, the smaller that liability looks.

Now according to economic theory, the discount rate should take account of the likelihood that a future cost will actually occur. If it is unlikely that my daughter will go to college, then I should use a larger discount rate in valuing the expected costs. In other words, I should discount those costs more than I would if I thought it a certainty that she’ll go.

So, what does it mean when people argue for a high discount rate in valuing pension systems? It means that they don’t think it is very likely that we will actually honor our promises to public employees. Thus, they believe we should discount those costs accordingly. They won’t say this, but this is exactly what such an assumption means. Conversely, people like Eileen who feel it is more prudent to use a low discount rate are saying that we should count on actually having to pay these pensions; that we should plan to honor our commitments.

As if the mathematics weren’t confusing enough, the politics are enough to drive one to distraction. For some reason, those who favor a low discount rate (i.e., those who think we should actually plan to live up to our promises) are often painted as anti-worker. This is because these cold-hearted economists’ calculations come to the conclusion that public employees’ pensions are expensive. Meanwhile, some groups defend a high discount rate. They do this so that they can claim that public employee pension systems aren’t a big factor in states’ fiscal woes. And for this, they are often characterized as pro-worker. In reality, though, their accounting assumes that we should leave workers out to dry.

I’d be grateful to anyone who could help me make sense of that!

Hearing on Public Sector Compensation

Andrew Biggs of AEI presented testimony on his recent research with Jason Richwine on public sector compensation. They estimate federal workers receive a 14 percent salary premium and a 23 percent benefits premium compared with workers in the private sector, for an overall premium of 39 percent, or $60 billion annually. Their analysis controls for skills and experience by comparing federal workers with their counterparts in the private sector. As they authors state, federal compensation is neither “obscenely generous nor does it leave federal workers substantially underpaid.” By paying federal workers the same as their private sector counterparts, there are are some savings to be found. But as Andrew notes, finding the premium is easier than fixing it. Market flexibility would permit adjustments – raising salaries when demand for the position is low, and lowering salaries when the demand for the job is high.



Me on CNBC

I was on CNBC yesterday morning debating Professor Harley Shaiken on the Wisconsin situation. Here is the video:

Here is a link to Professor Shaiken’s website.

Here is a link to the GAO report I referenced. If you think the budget gaps of the last few years have been bad, you ain’t seen nothing yet: States face a $9.9 trillion shortfall over the next several decades.

In order to close these long-term gaps, the GAO estimates that states need to immediately cut 12.3 percent (or increase taxes by the same amount) and maintain these changes each and every year for the next 50 years. To put that in perspective, last year states cut 5.9 percent out of their General Funds (total spending, which includes borrowed funds, other state funds and federal funds actually increased!). So, as painful as the last few years have been, states are nowhere close to doing what they need to do in order to address their long-term problems. 

Professor Shaiken mentioned studies that find public-sector employee pay is comparable to private sector pay. Here is one such study. And here is another.

Here is Andrew Biggs and Jason Richwine from yesterday’s Wall Street Journal on why these studies are flawed. To wit: a) they typically don’t account for health benefits, b) they fail to accurately compare the value of guaranteed 8 percent returns in public pensions with 4 percent guaranteed returns in private 401(k)s, and c) they do not take account of greater job security among public sector workers. Here is a link to Biggs and Richwine’s analysis. Here is Veronique de Rugy on the matter. Here is Megan McArdle. Here is a New York Times graphic that focuses just on Wisconsin employees (counting only cash compensation, the median Wisconsin public employee–who is typically more-educated–earns 22 percent more than the median Wisconsin private employee).

Here is a paper that assesses the empirical link between public sector unionism and government spending.

Public Sector Wages: Freezing, Hiking, Measuring

The Obama Administration is receiving a range of reviews for its decision to freeze federal employee pay for the next two years. Is it a mark of good fiscal management, or a symbolic gesture. As expected, public unions don’t like it. Though few are praising Iowa’s departing Governor’s decision to do the exact opposite and approve a six percent wage increase for state workers.

Are public wages too high? Andrew Biggs finds while four recent studies conclude that state-local workers are no better paid than their private sector counterparts, these studies fail to account for health care benefits and pensions and thus understimate the real level of public sector compensation. When factoring these benefits in, the public sector pay premium is as high as 20 percent in California.

The Economist on the U.S. Pension Crisis

This week’s Economist features a report on the crisis in U.S. state pensions, with special reference to New Jersey and the recent study authored by Andrew Biggs of AEI and myself. The piece begins with a view of the municipalities, in this case, San Jose, California which has seen its pension costs triple in the last ten years. San Jose offers workers its own muncipal plan and according to Robert Novy Marx and Joshua Rauh, their unfunded liability is about $4 billion, or 321% of the city’s 2006 revenues.

Underestimating the Pension Bomb’s Impact

Today’s Wall Street Journal discusses why both corporations and governments are sticking to “unrealistic return assumptions” in forecasting their pension liabilities. The majority of pension plans expect an 8 percent return and have clung to this expectation, “through thick and thin.” These estimates the WSJ notes are partly due to the high returns witnessed in the 1990s bull market years. Indeed over a 25 year period pension plans had an annualized median return of 9.3%. Over a 10 year period that fell to 3.9%.However, it’s not the number they’re selecting that matters, it’s the rationale.

Part of the difficulty of lowering the discount rate lies in what happens as a result. Reducing the discount rate increases the size of the  liability and the contribution needed to ensure adequate funds. That is one reason states are moving slowly. New York, New Jersey, and Colorado have all reduced their discount rates from the 8 percent to the 7 percent range.Virginia cut its investment return from 7.5% to 7 percent to avoid an even worse strategy – investing the funds in more risky assets to make up for losses.

The discount rate issue will continue to be a big challenge for government pension systems in part due to GASB’s guidance.

We will be discussing this and other issues facing state pension systems this Friday at Mercatus.  Speakers include myself, Dr. Andrew Biggs of AEI, Scott Pattison, Executive Director of the National Association of State Budget Officers, and Utah State Senator Dan Liljenquist. You can register for the event, or view it online.

Pension Reforms on the Horizon in New Jersey

Governor Chris Christie has announced his plan to tackle New Jersey’s pension shortfall. Officially estimated at $46 billion, Andrew Biggs and I estimate the figure is closer to $174 billion (using the risk-free discount rate to assess the size of the liabilities).

Today’s Philadelphia Inquirer reports Christie will ask for a  rollback the 9 percent benefit enhancement enacted in 2001 for current workers. This is a good step to putting  New Jersey’s pension plan on more stable footing.

In addition to this announcement, Orin Kramer writes in the New York Times about the  role investment assumptions played in the pension crisis. He points to government standards that allow pension systems to measure their asset values looking back over a period of years which ultimately gives the plan the appearance of a higher level of funding.

Both articles emphasize the impact of decades of pension deferrals and also raise the issue of the role of government accounting standards in creating the pension crisis. As these issues are hammered out states will continue to face increasing fiscal pressures as benefit payouts increase making public employee benefits, in Governor Christie’s words, “the public issue of this decade.”