Pension Promises, Real and Imagined

PODCAST

Steven Malanga interviews Josh Barro about his PSI article, "Pension Promises, Real and Imagined."

Across the country, actuaries are coming to state and local officials and telling them they must sharply increase their contributions to pension funds. In New York State, my colleague E.J. McMahon and I calculate that school districts' required contributions to the New York State Teachers' Retirement Fund will more than quadruple over the next five years, which will require an 18 percent rise in school property taxes for pensions alone.

Something has to give--these rising costs must be handled with some combination of tax increases, cuts to pension benefits, and cuts to other government spending. But when lawmakers talk about cutting pensions, public worker unions are naturally objecting.

The core of the objection to pension cuts is, quite reasonably, "you promised!" Pension benefits represent compensation for work previously provided, and so a government that changes a retiree's benefits is effectively taking away duly-earned compensation.

Or at least, that's true of some cuts to pension benefits. One problem with the debate over pension reform is that we have conflated different kinds of cuts--some of which involve breaking promises, and others which do not. Not everything that's being called a "promise" really is one. Making this distinction clear will allow lawmakers to make cuts that don't break any real promises, and will overcome the unions' most valid objections.

Broadly, a government's pension obligations can be broken into three categories. Category 1 is benefits due for work done in the past. This category encompasses all the benefits due to retirees, but only some of the benefits due to current workers--more or less, the benefits they would be entitled to if they stopped working for the government today. Category 2 is benefits that current workers can currently expect to accrue for future work. Category 3 is benefits that will be earned by workers not yet in the system.

Of the three categories, only Category 1 is truly a promise--workers provided labor and accrued Category 1 benefits as part of their compensation. Categories 2 and 3 include benefits that might be provided as compensation for work to be performed in the future. Therefore, changing them should not be thought of as breaking a promise, any more than cutting a worker's pay (or reducing the generosity of his or her health benefits) breaks a promise.

I agree that governments should avoid breaking their pension promises and maintain Category 1 benefits unchanged. (There are a handful of exceptions, mostly when a government is truly insolvent; more on that below). But lawmakers ought to cut benefits in Categories 2 and 3--benefits that will be earned in the future, whether by current or future employees--whenever doing so is in the public interest.

For reference, this is exactly what happens in the private sector when firms choose to close their defined benefit pension plans. If a firm is not bankrupt, it can terminate its defined benefit plan, but it must honor all benefits accrued to date. That means that retirees don't face benefit changes at all, and current workers are entitled to keep whatever benefits they would have gotten if they left the company on the date of plan termination.

Unfortunately, that's not what is happening in state and local governments. While many states and localities have been implementing pension reforms designed to save money, these reforms generally take one of two routes. The most common is to touch only Category 3--cut benefits, but only for workers who aren't on the payroll yet.

Unfortunately, this saves little money in the near term. It also arbitrarily offers protection to a portion of employees' future compensation that (as demonstrated by the viability of a new benefit package for new workers) is not necessary to attract or retain talent. And we have historically seen states make such reforms and then undo them later; in Albany, calls for "tier equity" are familiar and have undone previous efforts to save money with Category 3-only reforms. Essentially, this approach makes meaningful cost savings through pension reform unlikely, and safeguards a cost area that should be open for cuts.

A minority of states are taking a different route--enacting reforms that reduce the generosity of benefits in all categories, including Category 1. Most notably, in the last year three states (Colorado, Minnesota and South Dakota) have reduced cost of living adjustments (COLAs) due to retirees, and to active employees upon retirement. These reforms generally also include more severe cuts for future employees (Category 3), but they treat active employees and retirees (and Category 1 and 2 benefits) in the same way.

This approach can save real money in the short term. But, depending on state law, it may not survive legal challenges. And it involves governments breaking promises that they made to public employees in exchange for their work--not a step to be taken lightly.

Essentially, most state lawmakers are treating Categories 1 and 2 as fundamentally similar, and very different from Category 3. While it is hard to make an ethical or contractual case for this practice, there is an excellent political one: unions represent the interests of current workers foremost, and they do not represent the interests of workers not yet on the payroll. So, there is similarly strong political pushback against cuts to benefits in Categories 1 and 2, even though workers' moral claim on Category 2 benefits is substantively much less strong.

A few reform proposals get the distinction right. The Civic Committee of the Commercial Club of Chicago has floated a pension reform for Illinois that would essentially copy private sector practice--existing pension plans would be terminated, and existing workers would keep their benefits accrued to date while receiving future benefits in a new pension system.

And Chris Christie's reform proposal in New Jersey is a mixed bag. Some components (like elimination of COLAs) follow the Colorado model of cuts for all classes of benefits; others apply only to new workers. But some changes apply to all future accruals by new and existing workers. With some tweaking, his plan could be much closer to private sector practice than what we are seeing elsewhere.

Finally, a few thoughts on when a government should think about abrogating its Category 1 liabilities. One specific case where this should be considered acceptable is the reversal of a previously-awarded retroactive pension increase. Especially at the peak of the dot com bubble, some states awarded benefit increases that applied to benefits accrued in the past--even for retirees. The reason Category 1 deserves special protection is that it represents compensation for work already provided, but this is not the case for retroactive increases, which were pure gifts.

More broadly, no promise is absolute. While states cannot go bankrupt, they can default on their pensions--and if a state's pension obligation load is so heavy that it cannot plausibly make its budget work, it might reasonably consider defaulting. (Some states have statutory or constitutional provisions that prevent default, but these are all creatures of the state government itself--states can change these provisions if they wish to default.)

Of course, the statement that states can default on their pensions applies equally to bonds--and that doesn't mean that states should default on their bonds wily-nily. Indeed, I would say that pension accruals for past work should be thought of roughly as like bond debts, and states should modify them only in the same sort of extreme circumstances in which they would negotiate with their bondholders.

This doesn't mean that bonds and pension debts should be treated as strictly pari-passu. Just as a bankruptcy judge overseeing a restructuring might prioritize some bills over others, states might have good reasons for treating bondholders and pensioners differently. But generally, there should not be situations where states are defaulting on accrued pension liabilities while bondholders are held fully harmless.

I do not believe any state is yet in a sufficiently dire fiscal situation to need to go back and restructure its bonds or its already-accrued pension debts. General budget discipline, and an appropriate revision of pension promises in Categories 2 and 3 (such as by terminating existing pension plans and moving all new accruals to 401(k) accounts) can provide states enough room to fix their budgets without default.

So, when public employee unions say that cuts to already-accrued pensions aren't needed to fix states' fiscal problems, I think they're generally right. What is needed--bold reform of pension benefits in the future, for both current and new workers--isn't going to make the unions a lot happier. But both public sector retirees and taxpayers should like such a program a lot better, and it won't allow anyone to reasonably claim that a promise is being broken. 

Josh Barro is the Walter B. Wriston fellow at the Manhattan Institute focusing on state and local fiscal policy. He is the co-author of the Empire Center for New York State Policy's "Blueprint for a Better Budget." He writes bi-weekly on fiscal issues forRealClearMarkets.com, is a regular contributor on National Review Online and has also written for publications including theNew York Post, Investor's Business Daily, the Washington Examiner, City Journal, and Forbes.com.