What Felix Salmon gets wrong about public pensions
Felix Salmon has recently posted a creative critique of my foundation‘s pension reform efforts, and specifically of Mayor Chuck Reed’s ballot initiative in California. As is true with any complex policy debate, simplified criticisms make good headlines but miss precisely the nuance that makes these problems challenging and controversial.
Here is the crux of Salmon’s argument: Current public retirement systems severely backload benefits, as illustrated by LJAF’s own analysis, which Salmon includes in his article in graph form. He argues that any proposal that seeks to renegotiate prospective benefits, such as the Reed proposal in California, is inherently unfair because it fails to recognize the benefits that workers rightly expect under the system, and it provides jurisdictions the power to renege on their promises to employees. Salmon then uses a simplistic but attractive analogy to argue in favor of what is known as the “California Rule.” He argues that entering into a pension system is like being given restricted stock — a concept he calls “restricted pension units” (RPUs) — and that these RPUs, granted at the beginning of a career, mature over the course of a worker’s 25-to-30-year tenure. If a worker walks away from employment, she forfeits the right to earn additional benefits under the system, but as long as she is employed, she is guaranteed to earn benefits under the same RPUs (pension parameters) that were in place when she was hired.
Let’s start with the points on which Salmon and I agree. We both believe that public pension plans should work differently than they do today. Under current plans, workers generally earn meager benefits for much of their careers, and only become eligible for a significant retirement benefit after 25 or more years of work for the same employer. This back-loading of benefits means that many public workers are retirement insecure for many years of their working lives. In fact, most of these workers, who will not stay for the requisite 25-plus years, will never achieve the level of retirement security promised under the current system.
Salmon and I agree that it would be much better to put workers from the outset in a system that places everyone on the path to retirement security regardless of tenure or date of hire. And we agree that we should hold governments accountable for fully funding the benefits that they promise to workers, whatever their form.
Our major point of disagreement is whether governments should have the ability to alter prospective benefits. I agree with Salmon that it would be unfair to use the back-loaded nature of the current system to take away a benefit that current workers can reasonably expect to earn in the future. Contrary to Salmon’s assertion, we would not take the position that, upon enacting prospective reforms, governments can “write off” any accrued amounts to which a public employee is entitled, nor would we want to eliminate future benefit accruals. That is, we do not believe that a public employee’s accrued or future benefits, whatever their size, should be “zeroed out” merely because the employee has not reached eligibility thresholds. Any such argument would contradict the spirit of fairness and responsibility with which we approach this issue. And I strongly suspect that any such effort, if made, would be summarily stuck down by a court of law– an action with which I would wholeheartedly agree.
To be clear, no one — not I, not LJAF and not the Reed proposal — is advocating an extreme approach whereby all accrued and future benefits are taken away upon a government employer’s adoption of reform. But the choice here is not binary, and it is a mistake to couch this debate, as Salmon appears to do, in terms of extremes. Many jurisdictions are struggling to keep pace with rising pension costs, resulting in negative consequences for workers and taxpayers alike. There are fair solutions that would place retirement plans back on sustainable footing, but these solutions require that we grant governments the flexibility to address their specific circumstances.
In practice, many jurisdictions that have this flexibility have used it to adjust prospective benefits — and have done so fairly and responsibly. Since 2009, 48 out of 50 states have enacted some form of pension reform, and many include changes to prospective benefits. None of these jurisdictions has enacted, or even sought, changes of the nature that Salmon suggests. The most common way jurisdictions have changed prospective benefits is by increasing employee contributions, which 23 jurisdictions have done since 2009. For example, in 2012, Florida raised the employee contribution to the pension plan from 0 to 3 percent. When employee contributions are increased, the total benefit remains the same, but the employer-funded portion is reduced. Because workers pay more for the same level of benefits, this represents a benefit reduction.
The graph below shows the effect of Florida’s 2012 reform on a mid-career teacher with 23 years of service at the time of the change. It is clear from this graph that benefits for this mid-career worker were not “zeroed out,” and while her future benefits will be somewhat lower, her employer-funded benefit at age 58 would be reduced by less than 5 percent.
Mayor Reed’s own reforms in San Jose did not summarily cut accrued or future benefits. Instead, the ballot measure, which passed by an overwhelming majority in 2012, gives workers the choice between paying more to receive the same benefits or paying the same amount but earning benefits at a lower rate going forward.
Even in Rhode Island, which in 2012 enacted the most comprehensive reform package to date, great care was taken to protect workers from reductions that would be too onerous. The law specifically ensured that those who were approaching retirement were the least affected by the changes. The recent amended reforms, which were the result of a compromise among all interested parties, maintain the spirit of the 2012 law and do not materially change its objective or anticipated outcomes.
There are numerous advantages to granting governments the ability to responsibly address these issues through the collective bargaining or democratic process. Among other things, workers would benefit from flexibility regarding future retirement compensation as rising pension cost is crowding out salaries, health benefits and even jobs in some jurisdictions. Employees themselves may seek the option of switching to a system in which their benefits are portable, thereby increasing their own career prospects and geographic and professional mobility.
In short, I fully agree with Salmon that it would be unfair to exploit the current system’s back-loading to harm workers. But unlike Salmon, I do not believe that the answer to this is to deny governments flexibility to address their financial condition through a collaborative and democratic process. And that is the essence of the Reed proposal.
Let’s now consider Salmon’s “RPU” argument. Salmon’s analogy likening defined benefit promises to restricted stock is attractive, but it inaccurately represents the way the current system works. Under current plans, workers’ benefits are similar to restricted stock in that the benefits workers earn across their careers are not immediately accessible, but rather only “mature” once they reach the plan’s retirement eligibility thresholds. But that is where the similarities end. Contrary to Salmon’s suggestion, workers are not granted RPUs at the beginning of their careers. On day one of employment, an employee is not granted an RPU valued at a fraction of the anticipated aggregated pension wealth in year 30. Instead, the employee earns RPUs throughout her career. The value of an employee’s RPU holdings can change substantially from year to year because both the maturity of the RPUs and the level of promised benefits can jump dramatically when workers cross the plans’ various eligibility thresholds. We certainly agree that governments should not be allowed to exploit the back-loaded structure of pension plans and harm workers’ retirement security through prospective changes that do not recognize accrued benefits to date or that reduce future benefits too sharply. But it is inaccurate to imply, as Salmon does with his RPU example, that retirement compensation is completely inflexible from and after the date a worker is hired.
Salmon has published three blog posts about LJAF and pensions in the past two weeks. In each of these, he has been quite liberal in making assumptions of my words, positions and thoughts. I invite Salmon to ask us for comment on our beliefs and actions, and to understand our views on technical questions, prior to running his next article. The financial sustainability of the public pension system is an immensely complicated issue that could benefit from Salmon’s superior intellect and journalistic focus. And we would all greatly benefit from fact-based dialogue.