U.S. Public Pension Plans are Different (and Not in a Good Way!)

Filed Under (Finance, Retirement Policy, U.S. Fiscal Policy) by Jeffrey Brown on Jun 11, 2012

I have written numerous blogs about the frustration that the financial economics community has with the Government Accounting Standards Board (GASB) rules that govern the way we account for public pension liabilities in the U.S.  The basic problem is that GASB standards do not account for risk in an appropriate way (in fact, they do not really account for it at all!)  Instead, they allow public plans to under-state the size of their liabilities by acting as if they have a risk-free approach to investing money at approximately 8 percent per year forever.

On occasion, someone will ask me if this is really just an accounting issue, or whether it actually has real effects on real-world behavior.  Although I can give countless anecdotes for why it affects real behavior, it is always better when a highly respected and disinterested party can provide rigorous empirical evidence to support the claim.

Well, now we have such evidence.   Just last month, three financial economists (Andonov, Bauer and Cremers) publicly released a rigorous new research paper entitled “Pension Fund Asset Allocation and Liability Discount Rates: Camouflage and Reckless Risk Taking by U.S. Public Plans?”  In this paper, the authors use an international database to look at the asset allocation decisions and discount rate assumptions of both public pension funds and non-public pension funds in the U.S., Canada and Europe.  What is particularly nice about this paper is that it is able to show what outliers U.S. public plans really are.  Not only do they look quite different from corporate DB plans in the U.S., but they also look different from both public and non-public plans in other countries.

Specifically, the authors state that “U.S. public funds seem distinct in that they can decide their strategic asset allocations and liability discount rates largely without much regulatory interference, due to wide latitudes allowed in the currently applicable Government Accounting Board (GASB) guidelines. In particular, these guidelines link the liability discount rates of U.S. public funds to the (assumed) expected rate of returns of the assets, rather than to the riskiness of the liabilities as suggested by economic theory.”  As I have written before, this is an intellectually vacuous approach to discounting.  What I had not fully realized is how unique this mistake is to U.S. public plans.  The authors go on to state that in Canadian and European funds – both public and private – liability discount rates are “typically … a function of current interest rates,” an approach which (assuming the interest rate is chosen appropriately) is much more in line with basic economic theory.

The most striking finding is the impact that this difference in accounting has on real behavior.  The authors find that “in the past two decades, U.S. public funds uniquely increased their allocation to riskier investment strategies in order to maintain high discount rates and present lower liabilities.”  This really is a case of the tail wagging the dog – by allowing an intellectually flawed approach to discounting to be codified in GASB standards, we have provided incentives for public pension fund managers and their boards to over-invest in risky assets.

There are many losers from GASB-induced deception.  Public workers end up with less-well-funded pensions.  Taxpayers end up bearing financial risk without realizing it.  Investors in public debt are given inaccurate information about the size of the pension liabilities.  Isn’t it time that we fix this?

The Choice Between Two Unconstitutional Options is Not Constitutional

Filed Under (Other Topics, Retirement Policy, U.S. Fiscal Policy) by Nolan Miller on May 29, 2012

As I’ve said before, I’m not a lawyer.  But, since the Illinois House Democrats have decided to move into incentives, why not?  The details of the pension reform proposal that passed an Illinois House committee today are still vague, but here is a write up about it.

Simply put: the proposals currently under consideration in which members are offered a “choice” between options, as currently constructed, are not constitutional.  Here’s why.

The Illinois Constitution says that membership in a state pension program is a contractual relationship the benefits of which shall not be diminished or impaired.

Any contractual relationship has to have, well, a contract.  In this case, the terms of the contract are spelled out in the Illinois Pension Code.

The Illinois Pension Code specifies the way in which pension benefits will be calculated.  The details are slightly different for different pension funds, but I’ll talk about the part that pertains to Tier I participants in the State Universities Retirement System (SURS).  In particular, the amount of the retirement annuity is specified in Section 15-136 of the Pension Code.  Here it is:

Rule 1: The retirement annuity shall be … for persons who retire on or after January 1, 1998, 2.2% of the final rate of earnings for each year of service.

That seems pretty clear.  The “final rate of earnings” is defined in Section 15-112.  For a person who first becomes a participant before Jan. 1, 2011 (i.e., Tier I participants), the final rate of earnings is defined as:

For an employee who is paid on an hourly basis or who receives an annual salary in installments during 12 months of each academic year, the average annual earnings during the 48 consecutive calendar month period ending with the last day of final termination of employment or the 4 consecutive academic years of service in which the employee’s earnings were the highest, whichever is greater. For any other employee, the average annual earnings during the 4 consecutive academic years of service in which his or her earnings were the highest. For an employee with less than 48 months or 4 consecutive academic years of service, the average earnings during his or her entire period of service.

That also seems pretty clear.

One more excerpt from the Pension Code.  This one has to do with annual cost of living adjustments (COLAs).  From Section 15-136

The annuitant shall receive an increase in his or her monthly retirement annuity on each January 1 thereafter during the annuitant’s life of 3% of the monthly annuity provided under Rule 1, Rule 2, Rule 3, Rule 4, or Rule 5 contained in this Section. The change made under this subsection by P.A. 81-970 is effective January 1, 1980 and applies to each annuitant whose status as an employee terminates before or after that date.

Beginning January 1, 1990, all automatic annual increases payable under this Section shall be calculated as a percentage of the total annuity payable at the time of the increase, including all increases previously granted under this Article.

This part of the Pension Code also seems clear: COLAs are to “include all increases previously granted under this Article.”  In other words, COLAs compound rather than being based on the original amount of the annuity.  And, COLAs start the January after retirement.

So, let’s review.  The Illinois Constitution says that membership in a pension system is a contractual relationship. The terms of that contract are given by the Pension Code, and the Pension Code specifies the way in which final pension benefits should be computed.  In particular, it specifies that the final rate of earnings is average earnings over the final 4 years of service, or the 4 consecutive years in which earnings were the highest.  Thus, the Pension Code states that future pay raises will be included in future pension benefits.  The Pension Code also states that COLAs are to begin immediately after retirement and be computed on a compound basis.

So, let’s return to the “choice” that would be offered to members of the pension system under the proposal.  Details are sparse, but the basic choice to be offered to members will be:

(A)  Keep the current pension plan, but give up the state subsidy for retiree health benefits and having future raises be included in pension benefits, and

(B) Keep the state subsidy for retiree health benefits, but receive a less generous cost of living adjustment (COLA) where annual increases are based on the pension payment at the time of retirement rather than the most recent year’s pension.  That is, the COLA is not compounded over time.  Further, the COLA will not kick in until 5 years after retirement or age 67, whichever comes first.  There is also language in at least the governor’s proposal that will limit the COLA to a simple 3% or ½ the increase in the consumer price index, whichever is lower.

Now, supporters of this approach claim that is constitutional because it offers participants a choice.  This claim is invalid.  While a choice might be constitutional, in order for this to be the case, it must be that one of the options does not impair or diminish the benefits of the current pension system.  This is not true here.  Option (A) denies members their contractual right to have the final annual rate of earnings be based on their highest 4 years of earnings, which would include future raises.  Option (B) denies members their contractual right to have COLAs be 3% compounded each year.  Since both options impair and diminish the benefits of the pension, if members are forced to make a choice between A and B, their pension benefits will necessarily be reduced.

Constitutionally speaking, two wrongs don’t make a right.

Consequently, to me it seems clear that the proposals are not constitutional.  Given that so many of our legislators are backing these proposals, there must be an argument for why the proposal is constitutional.  I can’t see it, though.

ADDENDUM (5/30/12):  This isn’t a post about whether it is right or fair to reduce retiree health benefits (it isn’t), but rather whether it is constitutional (it probably is).  Retirees who began working for the State of Illinois before April 1986 (at least in the case of SURS) may not be eligible for Medicare Part A.  In this case, removing health insurance benefits would leave workers exposed to significant financial and health risk even after the age of 65.  The state also does not contribute to Social Security, so state workers who retire are also not eligible for Social Security (unless it is by virtue of having worked for another employer).  Obviously, removing employer-sponsored health benefits and reducing the COLA is going to expose retirees to substantial new risks, and the proposal becomes much more complicated and controversial in this case.

Fiscal Sustainability AND Retirement Security: A Reform Proposal for the Illinois State Universities Retirement System (SURS)

Filed Under (Retirement Policy, U.S. Fiscal Policy) by Jeffrey Brown on Feb 9, 2012

I have released a paper today that proposes a new plan for the State Universities Retirement System.  Co-authored with Robert Rich, the Director of IGPA, the paper proposes a hybrid system that would be partially funded by both workers and universities. It contains several components that reflect some of the ideas that have been publicly discussed by state leaders in recent weeks.

 The proposal has four basic components: 

1) Create a new hybrid retirement system for new employees that would combine a scaled-down version of the existing SURS defined benefit plan with a new defined contribution plan that would include contributions from both employee and employer; 

2) Peg the SURS “Effective Rate of Interest” to market rates; 

3) Redistribute the SURS funding burden to include a modest increase in employee contributions and new direct contributions from universities, thereby reducing state government’s burden on state government; and

4) Align pension vesting rules with the private sector, which would decrease the years new employees hired after January 1, 2011 would need to work for their pension benefit to be vested.

The plan is intended to substantially reduce state expenditures on public pensions, while still providing a reasonable source of secure retirement income to university employees. 

Click here to read the full paper.

The Hidden Economic Logic Behind the “Take it to the Courts” View on Illinois Pensions

Filed Under (Retirement Policy, U.S. Fiscal Policy) by Jeffrey Brown on May 23, 2011

Many public employees and retirees in Illinois are (understandably) extremely agitated by the ongoing discussions about public pension reform in Illinois.  Today, I was forwarded yet another email by someone concerned about recent comments made by Illinois State Treasurer Dan Rutherford. 

In a nutshell, some in the legislature are kicking around an idea to get around the Illinois constitution’s prohibition against “impairing” retirement benefits by offering employees a choice:  pay higher premiums in order to keep existing benefits, or switch into a less generous plan. 

To questions about whether this would or would not violate the impairment clause of the Illinois constitution, there is considerable uncertainty.  Rutherford’s reaction view is (my paraphrase, not an actual quote) – “let’s pass it and then let the Courts sort it out.”

It is obvious why this is not satisfying to public employees – after all, it is their benefits that might get cut, or their contributions that may go up. But setting aside all the questions about what we “should” or “should not” do, I think there is tremendous logic to having the legislature pass a law in order to get a more definitive ruling on what the state “can” or “cannot” do.  Yes, I agree that it is really unfortunate that we may have to pass a law to find out exactly where the limits of the impairment clause are, but that appears to be the hand we have been dealt.  But figuring out where Illinois Courts will draw this line is exceedingly important.

Why?  As I have noted before, when employers provide employee benefits, they are not doing so just to be nice.  They are doing so to attract, retain and motivate employees.  In short, it is one component of the compensation package.  In an environment that is disciplined by market forces, employers will only offer employees pensions if the average employee values the pension at more than it costs the employer to provide.  Otherwise, both would be made better off by paying cash.  As I have also written, however, it is not clear how well this market discipline works for public employees.

The major problem we have in Illinois is that we may be in the worst of all worlds, namely, one in which the pension benefits are indeed fully protected by the constitution, but where the perception of political risk means that employees value them far less then they will actually cost to provide.  If this is the case, then nobody wins!  Taxpayers are on the hook for the full cost, but employees do not value the benefits fully.  So the total cost of providing public services goes up!

We would all be better off to have legal clarity.  If the state courts rule that the benefits are protected, then public employees and retirees can go back to valuing their benefits at full value (which will help with recruitment, retention, and general happiness), and the state can move on to figuring out how else to manage its serious fiscal problems.  If the court rules that forcing higher contributions does not violate the contribution, then we can hopefully have a sensible conversation about what the optimal mix of wages and benefits are going forward. 

Either outcome would be far preferable to the current situation.

A Modest Proposal to Reform Illinois Pensions

Filed Under (Finance, Retirement Policy, U.S. Fiscal Policy) by Jeffrey Brown on Jan 24, 2011

Illinois faces a challenging fiscal future.  Even with an enormous 67% increase in marginal tax rates (from 3% to 5%), Illinois does not have a sustainable long-run fiscal plan in place.  Decades of under-funding our pensions is part of the problem (though certainly not all of it).  And while we can argue all we want about who is to blame, it is an undeniable fact that the unfunded pension obligations are a substantial part of the fiscal mess that lies before us.

Most of the rhetoric on this issue has devolved into finger-pointing: “it is those lavish public pensions” versus “no, it is the irresponsible politicians who failed to fund them.”  Rather than adding to the argument about blame, I would like to suggest a way forward.  This is only a very rough conceptual outline – you will see there are no numbers attached.  And I should also be quite clear that what I am about to suggest does not “solve” our pension funding problem.  But it could help, so here goes:

First, three observations:

1.        We must begin by recognizing that there are two highly inter-related issues, both of which must be addressed.  The first issue is that we face a large fiscal challenge that, mathematically, almost seems to require that we find a way to reduce future pension obligations.  The second issues is that we face a human resources challenge – that at least in some parts of the public sector, we need to provide a compensation package that makes it possible to attract and retain the right kind of employees with the right kind of skills.  We cannot simply “cut pensions” without implications for our ability to compete in the labor market.  So the trick is to “fix” our budget problem while maintaining our ability to attract the professors, teachers, and other professionals that we want in the public sector.  Too often, the debate ignores this second part, implicitly (and mistakenly) suggesting that we can just slash pensions without any adverse consequences.

2.       The public defined benefit (DB) model, for all its strengths (e.g., inter-generational risk-sharing, retirement income security, cost efficiencies from pooled investments, etc.) also has some fundamental flaws, the most important of which is that it is far too easy to play financial games at the expense of future generations of taxpayers.  The list of budget gimmicks is far too long to enumerate here, but the gist of it is that we often end up growing our future liabilities in return for small short-term gains and then use accounting gimmickry (much of it blessed by the Government Accounting Standards Board) to hide the real costs to taxpayers.  Even if we did not face yawning budget chasms, citizens ought to be alarmed by the poor governance engendered by this system.

3.       The private sector 401(k) model – which is sometimes suggested as a replacement for public DB plans – is also deeply flawed.  As underscored by the recent recession and financial crisis, the existing 401(k) system is woefully inadequate when it comes to providing good tools for financial risk management.  There are mountains of empirical studies documenting the lack of financial literacy in the population and the resulting biases and mistakes that people make when forced into a “do-it-yourself” retirement system.  A few examples – inadequate savings rates, too little diversification, chasing past returns, failing to insure against the risk of outliving one’s resources, and many more.

So, what do we do?  Here is a modest proposal to get the conversation started:

1.        For future workers, we have already scaled back the public DB.  But I say we go even further.  For starters, let’s think about cutting the DB pensions in half.  But in recognition of the fact that we need to remain competitive in the labor market, please read on …

2.       Then, let’s supplement the DB pension with a fully-funded, income-oriented, Defined Contribution (DC) system.  I am not talking about a private sector 401(k).  I am talking about a sensibly designed, mandatory savings program that automatically diversifies people into low cost funds at appropriate savings rates, and that automatically convert into guaranteed retirement income as one approaches retirement.  For example, TIAA-CREF annuities, (DISCLOSURE:  since 2009, I have served as a Trustee of TIAA), which provide low cost investment options and opportunities to convert into lifetime income annuities.  This is akin to what Orange County California did last year – which you can read more about in Roger Ferguson’s op-ed in the WSJ last week.  They reduced their DB plan and supplemented with an income-oriented DC plan.  Importantly, this plan was supported by government officials and the unions.  Unlike the typical 401(k), this approach would manage risk more effectively, and thus would still provide valuable retirement benefits to public sector workers.

3.       For existing workers covered by the constitutional non-impairment clause:  since we probably cannot force them to take a lower benefit, we give them the voluntary option to switch to this same hybrid DB/DC pension.  But with two tweaks:  (i) We announce that the decision is voluntary, but that the default option is that they will be switched to the new system.  Those who wish to remain in the old system would be given a fixed period of time (say, 3-6 months) to fill out the paperwork (or the online form) to state their desire to stay in the old system.  A huge literature in psychology and economics suggests that many people will go with the default.  (ii) When determining the “price” at which we will convert future DB promises into DC contribution, we use a high discount rate that reflects the subjective political risk associated with the benefits.  This is key to this proposal saving the state any money, since the conversion of DB to DC, by itself, does not create any savings.  So let me say a bit more — in essence, we know that a lot of workers are concerned that they will never get their full DB benefit.  So this provides an opportunity to make a fair, voluntary trade – you get a guaranteed contribution to your DC account – which you will then own and which will be protected from the State – and in return, you accept a small “haircut” on the size of this contribution.  Yes, it would be a benefit “cut” relative to the currently promised but under-funded benefits, but numerous studies from the U.S. and abroad suggest that people would be willing to accept such a haircut in return for no longer having their future benefits reliant upon the ability of the State of Illinois to finance their DB pension.

4.       How do we deal with the “transition cost?”  To be clear, this approach would reduce the long-run obligations of the state.  But it would also require that – in the short-run – the state come up with more cash in order to fund the DC plans while still making good on DB promises to existing retirees.  How do we handle that?  A few ideas:

a.       Spread the contributions to the DC tier (that are payment for the reduced DB) over several years

b.      Don’t be afraid to issue debt to finance this.  Keep in mind that this would not represent a net increase in indebtedness – we would simply be exchanging implicit debt (to pensioners) with explicit debt (to bondholders).  Indeed, with the higher discount rate, the *total* debt (implicit plus explicit) would go down.

This obviously leaves a LOT Of details to be worked out.  But the system would have several advantages over the status quo. A few of the big ones are:

1.        It imposes funding discipline.  The state would be legally required to make its DC contributions going forward.

2.       It maintains a focus on the retirement income security of participants.  It is true that, with the “haircut”, the DC would not cover 100% of the reduction in the DB benefit.  But it is also true that the participant would no longer have to worry about leaving the entirely of their retirement income security in the hands of Springfield.

3.       If I am correct that many participants would be willing to accept a reduction in expected benefits (relative to the full DB benefits) in return for no longer being subject to the political risk that Illinois defaults on its DB obligations, then it has the potential to save money.  (But let’s be clear – there is no free lunch here!  The cost savings only come from people accepting a smaller expected benefit in exchange for reducing the political risk to their benefit).

Any takers?

Misguided Reform Rhetoric Around Illinois Pensions

Filed Under (Retirement Policy) by Jeffrey Brown on Mar 31, 2010

Illinois pensions are in the news yet again.  Last month, the Pew Center on the States reported that Illinois was once again the poster child for everything wrong with the funding of state pensions, noting that we had the worst funding ratio of any state in the country.


Last week, Illinois House Speaker Michael Madigan decided – finally – to take some action.  He secured a House vote to change pension benefits for future Illinois state workers.  Specifically, this proposal would raise the full benefit age to 67, cap the maximum pension income at a bit over $100,000, limit cost-of-living increases, and so on.  In short, the package amounts to benefit reductions for not-yet-hired future state workers.  


Why this option?  To put it simply, there are only two options for fixing the funding problem. 


Option one is increase revenue to the system.  In other words, make additional contributions.  But this would require that Illinois lawmakers raise taxes or cut other state spending, neither of which is politically popular.  


Option two is to reduce the liabilities.  But as I have written before, the impairment clause in the state constitution prohibits benefit reductions to existing retirees and existing employees.  So the only way to reduce liabilities is to cut benefits for future workers – those that have not yet joined the system.  And that is precisely what Madigan pushed through the House.


[By the way, the only “option three” is to, in the words of Alan Greenspan when discussing Social Security, is to “repeal the laws of arithmetic.”  I am pretty sure that most state governments would choose this option if they could!]    


As a fiscal conservative, I have no real objection to the decision to reduce future liabilities in the way that the House has chosen to do.  But two issues that have come up in the debate that I think are worth a bit of analytical clarity.  


First, estimates of future savings are almost surely inflated.  There are two reasons for this.  One is that some of the estimates appear to have simply looked at undiscounted dollar flows, which implicitly assumes a dollar saved in 2050 is the same as a dollar saved in 2020.  This is obviously not the case, since a dollar saved earlier has a much higher present value.  A second reasons is that – as I have written before – pensions are part of the overall compensation package.  If we reduce future retirement benefits, our ability to attract top faculty members, for example, will be reduced unless we increase compensation in some other way.  None of the cost savings estimates account for this.    


Second, there is clear confusion about the source of the funding problem.  Much of the rhetoric around this legislation focused on the level of benefits.  The Champaign News-Gazette is a typical example, stating:

“A big part of Illinois’ horrendous budget problems can be traced to the high costs for the lavish pensions many public employees enjoy. They are far more generous than those available to workers in the private sector, and that’s a big reason why state public pensions are underfunded to the tune of an estimated $80 billion.”

This is wrong for several reasons.

First, the real source of the funding problem is not level of benefits.  It is the fact the Illinois legislature has consistently failed to make the annual contributions that are called for under standard funding formulas.  My colleague Fred Giertz has done some calculations suggesting that if the legislature had made its required contributions every year, the Illinois system would be slightly over-funded, not under-funded.  In short, don’t blame the pensioners for the lack of fiscal discipline on the part of our politicians.

Second, the comparison of public pensions to private pensions is misleading.  One reason is that the public pension replaces both Social Security and a private pension.  Social Security costs roughly 12% of payroll today.  Private employers who offer pensions typically contribute several percent more.  On that basis, Illinois public pensions are not “lavish.”  A second reason is that – yes, I am repeating myself – this is part of an overall compensation package.  So any comparison needs to account for the value of all salary and benefits, not just a single piece of it.