The False Obstacles to Pension Reform
State and local government pension plans have become a lot more expensive for taxpayers over the last few years. This is because public pensions are largely "defined benefit" plans: workers are guaranteed a specific benefit amount regardless of the performance of the assets in which pension plans are invested. When the market underperforms expectations (as it has recently, in spades) it is taxpayers rather than pensioners who bear the loss.
Across the country, pension actuaries are coming to state legislatures and calling for an increase in pension contribution rates, which is needed to shore up the funds since they have lost so much value. For example, New York City's pension costs have risen over the last decade from approximately $1 billion per year to $8 billion, and will soon reach $10 billion. In response, many states are acting to reduce costs, by trimming future benefits and requiring employees to pay more. 16 states enacted pension reforms in 2008 and 2009, and at least five more have followed in 2010.
Yet, with the key exception of Utah, states have shied away from following private firms' lead and abandoning defined benefit. (84% of state and local employees retain defined benefit coverage, compared to 21% of private sector workers.) Even in New Jersey, where Governor Chris Christie is proposing reforms that are sweeping relative to what is being done in most other states, the proposed reform package maintains the essential defined-benefit nature of employee retirement benefits.
This is a concern, because the problems with defined-benefit pensions aren't a one-time fluke caused by a bad recession and a stock market crash. Their flaw, especially in the public sector, is structural: they involve lawmakers making promises today about payments the state will make decades from now. There is a strong incentive to offer extra retirement benefits instead of extra cash compensation, because the cost can be pushed off into the future. And because the actuarial math is so complicated and subjective, analysts and legislators have trouble figuring out how much pension benefits really cost.
These structural issues have led to a pattern in pension benefits. When times get tough, states cut back pension benefits, but usually only for new hires. The existing employees get their full benefits, and taxpayers kick in extra cash to offset poor performance by assets held by pension plans. (A handful of states have touched benefits for active employees in the last two years, but the vast majority of reforms apply only to future employees -- and reforms affecting active employees or retirees may be blocked by courts in some states.)
But when the stock market soars and pension plans become overfunded, states sweeten pension benefits. New York, New Jersey and California are among the states that sweetened pension benefits as a result of the tech bubble, and are now regretting that choice. And unlike the cutbacks, which usually apply only to new hires, sweeteners usually provide retroactive benefits to existing employees and even retirees. For public workers, pension fund investments are a heads-we-win, tails-you-lose proposition.
There is a reason that New York State's most recent pension reform, enacted last year, is called "Tier V" and not "Tier II." It's that New York has been down this road several times before: creating new, less expensive pension "tiers" for new employees while maintaining rich benefits for existing workers. But later, once the stock market rebounds, state lawmakers go back and re-sweeten benefits to about where they used to be.
Consider how odd this would seem in the context of a defined-contribution retirement plan. Companies change the generosity of their 401(k) match terms often. But if a company raises its 401(k) match rate from 50% to 100%, nobody would expect them to go back and retroactively increase the matches paid in past years to active employees, let alone retirees.
Yet this is standard practice when states increase pension benefit generosity; for example, New Jersey awarded an across-the-board pension increase of 9% to all pension beneficiaries in 2001, active and retired. This move cannot be defended as useful for attracting a talented workforce; you raise prospective compensation to do that. It was a pure giveaway to public workers, and one that was not uncommon around the country.
I therefore worry that reforms like Colorado's and New Jersey's, which are supposed to produce large savings over time, will again be undone when the economy looks a bit better and the legislature a bit more favorable to unions. A move to defined-contribution plans would alleviate this concern.
Why aren't states adopting a model that would stop such abuses for good? We're starting to hear one reason around the country -- and it's not a very good one. North Dakota legislators, for example, have been looking at a move to a 401(k) system, but are being warned that such a move will impose transition costs, and therefore be more costly for taxpayers than staying the course.
You may be familiar with the concept of transition costs from the debate over Social Security reform. Moving to private accounts would require a shift from a pay-as-you-go benefit to a pre-funded benefit. During the transition period, you have to tax workers both to pay current retirees and to fund their own accounts, which would make the program temporarily more expensive. This is a serious concern with Social Security reform.
But the same issue is not present with pensions, because these benefits are already designed to be pre-funded. Most public employee pensions are not fully pre-funded today because of poor recent stock market performance, but taxpayers are expected to close that gap over time, whether or not you transition to a new system. The need to close this gap is not created by the transition.
However, there are some aspects of a move to defined-contribution that look, at first glance, like a transition cost. Here are three, none of which is a true transition cost.
1. If a pension plan is underfunded, that funding gap must be closed over time. Closing the plan to new participants shrinks the plan's "covered payroll" -- wages paid to active employees in the plan -- and therefore closing the funding gap will require that states contribute a greater percentage of covered payroll than if the plan were kept open. This is true. But the flip side is that new employees are participating in a plan with no funding gap to close at all. The government's contribution rate for those employees is lower than if the plan had been kept open.
Subject to points 2 and 3 below, these effects should wash. A pension contribution consists of two parts: the cost of benefits accrued by employees in the current year ("normal cost") and a payment toward closing whatever unfunded liability exists. If the old and new plans have equal normal costs, and the starting unfunded liability is the same, switching to a new plan is no more expensive than keeping the old plan.
2. When you close a pension fund to new participants, the average time from today when plan benefits will be paid starts shrinking. When this happens, plans shift to less risky investments to be assured of having cash to cover payouts due soon. Therefore, their expected returns fall and taxpayers must put in extra cash to make up the difference. This is the key concern in North Dakota.
This is a false concern for two reasons. One is that it assumes the higher average return of risky investments is free; it's not, as it increases the likelihood that taxpayers will have to come up with added dollars to cover below-target returns when the economy is weakest. The added certainty of returns is valuable to taxpayers; as demonstrated by Andrew Biggs of AEI, that value is equal to the price taxpayers would pay for a stock option covering any losses on risky investments.
The other reason is that actuaries offer guidance, but cannot force a plan to adopt a particular investment strategy. If lawmakers prefer to have taxpayers take a flier on the stock market instead of investing in safe bonds (something, I would note, they have already elected to have taxpayers do in pension funds all over the country) that's their prerogative; and it would eliminate (on average) the need for taxpayers to come up with any extra money to pay for transition.
3. The employer contribution to a defined-contribution plan might have to exceed the current "normal cost" of the defined benefit plan. This is a purely political question. Consider a pension plan with a target investment return of 8% per year, which is the most common return target among state and local retirement plans. If you take that plan's "normal cost" and turn it into an employer contribution into a 401(k), you get a retirement account that will on average be equal in value to the old defined benefit, assuming the employee invests his 401(k) in a portfolio with an expected return of 8% per year.
That might sound like a bad deal to you. You would probably prefer to get a guaranteed $100,000 in ten years rather than $46,319 today (which is enough, on average, to generate $100,000 in ten years at 8% annual returns). But just as that guarantee is valuable to public employees, it is costly to taxpayers, as they must provide the insurance that public workers receive. So, an increase in contribution rates above the normal cost to compensate for the loss of that insurance isn't an added cost; it is just the conversion of a hidden cost to an overt one.
Additionally, I would not be quick to assume that it is necessary to compensate public employees for the added risk they will bear in a defined contribution system. Political incentives and opaque costs have driven employee retirement benefits to a level that is likely in excess of what is necessary to attract quality talent, making a real reduction in benefits (at least for new workers) appropriate. One form that reduction can take is a shift of risk from taxpayers to employees.
Some workers will even prefer a 401(k) system. Pension benefits are rich only for workers who spend a full career inside the same retirement system; since half of public school teachers leave the profession in less than five years, some public workers would actually be advantaged by a shift away from DB plans.
The financial issues here are complex and unfortunately are not exactly what many state lawmakers signed up for when they first ran for office. In that light, it's nice to see one state getting its pension reform pretty much exactly right: Utah, which will move all new employees into a new pension system starting July 2011.
Utah employees will have a choice: a 401(k) plan with employer contribution, or a "hybrid" pension plan with many defined-benefit characteristics. The key difference is that the state's contribution to the fund will be capped at 10% of covered payroll -- if asset underperformance requires additional contributions to shore up the fund, employees will be required to come up with the money. Benefit sweeteners will be prohibited until the state has fully closed its pension funding gap.
State Senator Dan Liljenquist, who chairs the Retirement Committee, has been speaking around the country explaining how that state's reform will save money, not impose transition costs. Other states, many of which face pension funding shortfalls much more dire than Utah's, would do well to follow their lead.