Greece is not Germany, and California is not Vermont

February 28, 2012

Last week Gillian Tett of the Financial Times picked up Meredith Whitney’s municipal bond doomsday flag and started waving it for an international audience. Her article, entitled “Pension gap spells trouble for muni bonds,” broadly painted the entire municipal bond market as having unacknowledged, long-term issues. Her closing line seemed to be a call for investors to shift their concerns from European sovereign debt to the debt of muniland:

Fiscal woes, in other words, are not just a matter for the eurozone; investors had better keep watching that American periphery too.

I agree with Ms. Tett that it is important for investors to dig down into the affairs of municipal bond issuers. Like the nations of the European Union, the quality of fiscal management varies by state. The U.S. has a number of well-run Germanys and we also have a handful of Greeces.

The problem with Tett’s piece is that it both used some less-than-credible sources to assert that unfunded public pension liabilities are an enormous problem and failed to distinguish California from Vermont. For example, this graph shows the combined state and local pension fund expenditures of the country overall and the nine states that devote the highest share of their annual budgets to their pension plans (source: U.S. Census Bureau via the National Association of State Retirement Administrators, page 3):

According to the National Association of State Retirement Administrators (NASRA), the national average for state and local government pension contributions is 3 percent:

Based on the most recent information provided by the U.S. Census Bureau, approximately three percent of all state and local government spending is used to fund pension benefits for employees of state and local government… pension costs since 1980 have been reliably stable, declining from around four percent to nearly three percent in 2009.

That said, there are states and local governments that have severe pension problems. Unlike Greece, which was forced to make pension changes to receive a bailout, almost all state and local governments are taking voluntary steps to reduce their pension liabilities. The case law for reducing and removing pension benefits is becoming well established. A lot of adjustments to pensions and state budgets are being made already. This was a great comment on the FT piece:

Pensions can be renegotiated, as most debt is often renegotiated. A couple of years ago several analysts were predicting multiple state and municipal bond bankruptcies with massive dislocation in the financial markets, but there have only been a few bankruptcies. It will be painful, but states are making big cuts in their expenditures in order to balance their budgets. Many of the states will end up telling their pensioners that they can only pay them 50 to 75% or another percentage of what they are owed.

At this point the pensioners would have the choice of taking what is offered or suing in federal court. The barriers for an individual to sue a state are massive, but it is possible that a class action lawsuit could be brought against one of the states. But there would be strong motivation for both sides to settle, as the risk of leaving a decision in the hands of the court would be too risky for either party. A lot of this debt will jut [sic] be erased, to the detriment of the citizens but probably not the bond holders.

My biggest beef with Tett’s piece, though, is the use of the research of Joshua Rauh of Northwestern University. Rauh insists that when projecting pension fund returns, the interest rate for 10-year Treasuries must be used. Pensions do not allocate their assets 100 percent into Treasuries, though. The Public Fund Survey found that 99 of the largest pension systems allocated only 28 percent of their assets to the entire fixed income market. Because pension portfolios have diverse assets, their returns can fluctuate quite a bit. NASRA reports that for fiscal year ending June 30, 2011, state and local government retirement systems had a median investment return of 21.6 percent.

Here is the core problem with Rauh’s work, according to NASRA, page 1 (emphasis mine):

Over time, a majority of public pension fund revenues come from investment earnings. [Rauh’s] paper assumes state and local pension trusts, which currently have $3 trillion in assets, will generate investment returns roughly commensurate with bond investments rather than the diversified portfolios actually in use.

Public pensions typically assume that such diversified portfolios will earn a real (after inflation) return of 4.0 to 4.5 percent annually, depending on their asset mix. Long-term investment returns actually exceed this assumption, even after incorporating losses from the 2008 market decline. Yet, the Rauh-Novy-Marx paper assumes these portfolios will generate a real return of only 1.71 percent, well below not just historic norms, but also below projections made by investment experts, as shown in Figure 1. The result of this bearish assumption for pension plans is lower investment earnings and higher required contributions.

It is appropriate to point out individual states and municipalities with pension problems. But given that over 50 percent of municipal bonds are held by retail investors, a broad-brush criticism of muniland can easily spook the market and create a sell-off. Dig down and do the analysis. Greece is not Germany, and California is not Vermont.


Public Fund Survey: Scorecard

Government Accounting Standards Board: Fair Value Measurement

Morningstar: States Positioned to Avoid Default

Storms and Nation: More Pension Math

Sacramento Bee: Dan Walters: Pension battlefield shifts to San Jose and San Diego

Bloomberg: Cuomo Urges New York State Mayors to Take Fight on Pension Overhaul Home


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

You may in fact be right that Greece is not Germany and that California is not Vermont. But you unfortunately displayed a lack of understanding of basic financial economics principles with your statement that “Rauh insists that when projecting pension fund returns, the interest rate for 10-year Treasuries must be used. Pensions do not allocate their assets 100 percent into Treasuries, though.” Your second sentence is irrelevant when assessing the first sentence. Finance theory is crystal clear that any stream of cash flows – whether associated with pensions or any other source – should be discounted based on a rate that is adjusted for the risk of those cash flows. For pensions, it is the risk of the pension payments – and NOT the portfolio that is used to back them – that is the relevant consideration. The expected rate of return on the asset portfolio is totally irrelevant for determining the present value of the liabilities. For purposes of valuing the liability, it does not matter whether these public plans invest in long-term bonds, hedge funds, or macaroni. Portfolio allocation DOES matter for many other purposes, but not this one. If you do want to use expected portfolio returns to value the liabilities, then it is critical that one adjusts for risk, especially given the large implicit put option that the portfolio imposes on taxpayers. If you do this, you will get the same answer as Rauh and Novy-Marx. There is very little disagreement in the academic economics and finance community over this basic principle. This degree of consensus is rare in economics, but it exists here. Unfortunately, your post is propagating faulty reasoning and contributing to the continued widespread misunderstanding of this financial principle. Accepting Rauh’s estimates does not necessarily turn California into Vermont, but it does give a more accurate picture of the true size of the pension shortfalls that exist today.
Jeffrey R. Brown
Karnes Professor of Finance
University of Illinois at Urbana-Champaign

Posted by Jeffrey.Brown | Report as abusive

Thanks for your comment Mr. Brown.

The National Association of State Retirement Administrators has a much better rebuttal to Rauh’s work here: -MarxCritiquePART2.pdf

Posted by Cate_Long | Report as abusive

This post misses the point point about pensions and spreads gross misunderstanding about the cost of pension promises to taxpayers. Here is a response:  /02/29/pensions-in-muniland/

Posted by JoshuaRauh | Report as abusive

Thanks for the comment Mr. Rauh.

Can you please direct me to any data/research that shows that states are spending “14% of total revenue excluding federal transfers and 23% of tax revenues.” as you claim in your rebuttal. This is an astounding figure and one that I have not heard before.

Many thanks.

Posted by Cate Long | Report as abusive


You misquoted me. I did not write that they ARE spending that amount. I wrote: “The right question is not how much is being paid now, but how much WOULD HAVE TO BE PAID in order to claim that states are balancing their budgets. That figure is substantially higher, and amounts to 14% of total revenue excluding federal transfers and 23% of tax revenues.”

The research is here: rauh/research/RDPEPP.pdf

Before you take the NASRA rebuttals verbatim, please review my responses here  /06/23/responses-on-revenue-demands-of- public-employee-pension-promises/

and  /06/21/the-revenue-demands-of-public-em ployee-pension-promises/#more-1086

Posted by JoshuaRauh | Report as abusive

I seem to be on a different page Mr. Rauh. Sorry for the misquote. You write:

“The right question is not how much is being paid now, but how much WOULD HAVE TO BE PAID in order to claim that states are balancing their budgets. That figure is substantially higher, and amounts to 14% of total revenue excluding federal transfers and 23% of tax revenues.”

Does the GASB agree with your approach?

Posted by Cate Long | Report as abusive

The confusion here is caused by the use of different measures of pension liabilities. Professors Brown and Rauh, who are financial economists, seek to determine a price, or value, for pension liabilities based on current interest rates.

By contrast, public pensions calculate their liabilities not to price them, per academic theory, but to fund them, per the practical realities of public budgeting and stability of costs. Public pensions seek to answer the question, How much money will the pension plan require, if actuarial assumptions (including projected investment returns) hold true, to fund the cost of promised benefits?

Messrs. Brown and Rauh believe government accounting standards should divorce funding from valuation, and value their liabilities using a “risk-free” interest rate, i.e., the current Treasury rate. This price or value calculation comports with financial and economic theory. It fluctuates as interest rates rise and fall. Because interest rates are very low now, the price (value) of those liabilities looks very high. Because corporations, by and large, are required to calculate the funding level and cost of their pension plan on this basis, their funding condition has declined sharply and their costs have risen sharply.

This approach is defensible for corporations, which could, at least in theory, be acquired, declare bankruptcy, or dissolve, potentially leaving shareholders or taxpayers with unfunded pension obligations. The pension liabilities are part of the price, or value, of the corporation.

Public pension liabilities are not for sale; there is not a market for them. They don’t need to be priced. The sponsors of public plans, i.e., states, cities, school districts, etc., are “going-concerns,” and are not going out of business or selling their liabilities. Using current interest rates to “price” liabilities, as Messrs. Brown and Rauh call for, would introduce untenable volatility in pension funding levels and costs, as interest rates fluctuate.

Using the financial economist approach, all that would be necessary for public pension plans to become fully funded would be for interest rates to rise to eight percent A generation ago, interest rates reached double-digits. Using academic theory, public pension plans at that time were vastly overfunded. Now that interest rates have declined, per this theory, they are vastly underfunded. All of which highlights a major problem with applying this theory to the real world: it reflects the underlying condition not so much of the pension plan or its sponsor, but of bond markets.

Current GASB standards permit the use of a projected long-term investment return to determine the funding level and cost of public pensions. This approach promotes predictability and stability of cost, and reflects the long-term nature of the entities that sponsor public pension plans, i.e. states, cities, school districts, etc.

When Professor Rauh says that public pension costs should “amount to 14% of total revenue excluding federal transfers and 23% of tax revenues,” he is using financial theory to arrive at those numbers. He is ignoring the reasonable expected investment returns that public pension funds expect generate through their diversified portfolios. There also is more to those numbers than meets the eye, and more than this space or time will allow. For example, his use of “tax revenues” occludes the fact that about one-half of state and local government revenues come from non-tax sources, such as fees, tuition, etc., but much of which is used to fund pension benefits.

Posted by KeithBrainard | Report as abusive

The comment by Mr Brainard highlights the main points on which financial economists disagree with GASB and NASRA.

Municipal bankruptcies do happen. Even sovereign countries become insolvent and require bailouts. The thinking that financial analysis does not apply because governments can roll over their debts indefinitely contributes to these events. Even if a state is not going out of business, the debts of today’s taxpayers will still at some point have to be paid back by tomorrow’s taxpayers if the cost of servicing the debts are not going to swallow the whole budget.

The financial logic that increases in interest rates improve pension funding is certainly recognized in financial accounting for corporations, and it should be recognized in government accounting as well. A non-defaultable promise of, say, $1000 in 10 years time is much less costly to meet by setting aside resources today when interest rates are 5% than when they are 0%. If interest rates on low-risk securities were to rise, then public pension funding would improve – unless of course the rise was due to a spike in inflation in which case COLA increases would likely undo much of this effect.

Finally, I am not sure how much citizens will allow themselves to be overcharged for the provision of specific public services in order to pay for legacy pension liabilities. That is why we present the figures both as a share of total revenue and as a share of tax revenue. Regardless, each year that the contributions continue to be far below these levels, the needed contributions to catch up become larger.

Posted by JoshuaRauh | Report as abusive