Day: February 28, 2018

TriMet Shows That Public Pension Reform Is Possible

By Scott Shepard and John A. Charles, Jr.

The Oregon Legislature is currently meeting, and the conventional wisdom is that reform of Oregon’s overly generous Public Employee Retirement System (PERS) is impossible. According to Governor Kate Brown, we signed contracts with public employee unions, a deal is a deal, and we should just quietly accept our fate that the massive cost of PERS will lead to layoffs and service cuts at schools and other service providers.

There is another way.

The Portland regional transit district, TriMet, is not part of PERS and has been slowly reforming its pension program since 2002. As a result, 100% of all new employees are now in 401(k)-style pensions that have no long-term liabilities for employers. These are referred to as “defined-contribution” (DC) pensions in which monthly payments are made by management into personal accounts owned by employees. Once those payments are made, the employer has no further financial obligations. The eventual pension payouts will be a function of the market performance of whatever investments are chosen by individual employees.

This stands in contrast to “defined benefit” (DB) programs like PERS in which employees are promised various levels of retirement payments calculated through arcane formulas that leave management mostly clueless about the level of funding obligation they’ve agreed to. In many cases, those liabilities turn out to be much larger than expected.

The advantages for taxpayers of moving public employees into DC pensions is now evident in the actuarial valuations done for TriMet. According to the most recent valuation, projected annual benefit payments for TriMet DB pensions will peak in 2034 at $74.6 million, and then steadily decline to $6 million in 2072. They will hit zero by the turn of the century.

This was not something that TriMet did casually. Management was forced into it because of decisions made a decade earlier that caused long-term retiree obligations to explode. TriMet Board members are appointed by the governor. In the early 1990s, Governor Barbara Roberts and TriMet General Manager Tom Walsh wanted public approval of a massive expansion of TriMet’s light rail empire and the tax funding to pay for it. They feared that controversy about a union contract could endanger public support.

In their efforts to avoid strife, in 1994 they granted expensive concessions to the Amalgamated Transit Union Local 757 (“the ATU”) on behalf of its represented employees. Loren L. Wyss, the long-serving president of TriMet, objected and his battle with Walsh became public. In back-channel communications with Gov. Roberts, Walsh made it clear that either he or Wyss needed to go. In August 1994, Wyss met with Gov. Roberts, where he submitted his resignation.

As later explained in The Oregonian,

“…the contract just approved by Tri-Met union employees will protect all its members from additional contributions to their pensions for 10 years. It will also guarantee 3 percent minimum wage increases in the future…every single dollar of health, welfare, dental and vision plans will be paid for by the public employer; [and] the retirement age will decline to 58 within 10 years….”

The die was set for cost escalation. In the decade from 1994 to 2004, salaries and wages increased 72 percent; annual pension costs went up 160 percent; and the cost of health care benefits rose 116 percent. These increases plus stagnant revenues in the latter half of the period resulted in a tripling of unfunded pension liabilities, from $38 million in 1993 to $112.4 million in 2002.

Fred Hansen followed Tom Walsh as General Manger; and he moved new, non-union hires into DC pensions after 2002. This was a first step towards fiscal sanity. Resistance from the ATU kept TriMet from moving its new unionized workers to DC plans for another decade, by which time a citizens’ committee of Portlanders had issued a report declaring TriMet “on the brink” of disaster.

During a protracted negotiation with the union in 2012, TriMet CFO Beth deHamel testified at a binding arbitration hearing,

“TriMet’s union defined benefit plan would be placed on critical status and under federal oversight if it were a private pension plan subject to ERISA.” She also stated that unless something was done to shore up the plan, “TriMet could be forced to default on its pension obligations or its other financial obligations in the future.”

Union leadership eventually agreed to move all new members to DC pensions by 2013, while protecting existing members from reform. As a result of this delay, the union workers’ DB fund remained only 59 percent funded in 2013.

Nevertheless, the trends were now moving in the right direction. The number of active employees still accruing DB pension benefits fell from 1,580 to 1,460 from 2016 to 2017 alone. In 2017 the unionized workers’ DB account reached nearly 80 percent funding, with unfunded liability falling by nearly $50 million in a single year.

Neil McFarlane was TriMet General Manager during that era. He commented recently, “The shift [to DC pensions] has been a success. TriMet is paying more than the required annual contribution every year right now” because the system is closed. “We will be fully funded within the next few years: five to ten for the union plan, fewer for the non-union.”

The DC plan to which TriMet moved new workers has been recognized as one of the best in the country. It features low costs, high returns, and a guaranteed employer contribution that is paid irrespective of employee matching contributions. As a DC plan it does not create open-ended, unpredictable public liabilities to be paid by generations as yet unborn.

TriMet has not fully banished the ghosts of unsustainable employee-benefit promises past. It still faces a massive and escalating unfunded liability driven by health care costs, known in accounting jargon as “other post-employment benefits,” or OPEB. The health care benefits that TriMet granted away in the 1994 contract debacle have been described as “universal health care into the afterlife.”

The description is only a minor exaggeration, as the plan offered TriMet’s unionized employees health care without premiums and with mere $5 co-pays, and benefits that ran not only throughout retirement, but to the employees’ spouses and dependents for fully 16 years after the employees’ deaths. Total unfunded liability for OPEBs reached an astonishing $769 million dollars in 2016.

Compare: State Paralysis on PERS 

TriMet’s pension reform efforts offer a valuable guide to the Oregon legislature on how to contain and reverse the spiraling PERS disaster. The unfunded liabilities for PERS have grown from $16 billion to more than $25 billion in less than ten years, even with the far-too-optimistic 7.2 percent assumed-savings rate (i.e., discount rate) in place. Were the rate adjusted down to its actuarially appropriate level, PERS’ unfunded liability would explode to $50 billion or more at a stroke.

Even at the current recognized rate, funding status has fallen below 70 percent, even while mandatory payments to PERS by government employers have passed 26 percent of payroll.

Municipalities are laying off workers, depleting public services, and raising fees in order to fund the present level of recognized PERS unfunded liabilities. Some reduction in pension benefits will have to happen, one way or another. All parties will benefit from an orderly effort to reform benefits while there is still time. 

The Way Forward

The state should follow the tracks laid by TriMet by moving its employees from DB to DC plans as soon as possible. As TriMet has demonstrated, this move will begin to stanch the fiscal wounds that have been inflicted by a generation of recklessly overgenerous pension benefit promises.

Unfortunately for everyone, PERS reform has been hamstrung for more than 20 years by a wayward state Supreme Court, which has thwarted previous attempts at thoughtful change with erroneous interpretations of the federal Contract Clause. The legislature will be obliged to make bigger changes than would have been required years ago. It will have to move all current workers, whenever they were hired, to DC plans for all work performed after the date of the effective legislation.

While this reform will be significant, it also will be deeply equitable. Right now, older workers are receiving higher benefits for each hour worked than ever will be available to younger workers. This isn’t fair, and it may violate civil rights laws: Younger workers are more diverse than their older peers, which means that benefit reductions that affect only new workers have a disparate impact on women and minorities.

The reform will also pass constitutional muster. As the Oregon Supreme Court finally recognized in its Moro decision, correcting its long-held error, the legislature may change any benefits for work not yet performed, even for current employees.

The Oregon Legislature can and must follow TriMet’s example. The sooner this is done, the less drastic any later steps will be. According to TriMet General Manager McFarlane, solving a pension crisis “doesn’t get any easier with passing time.”

John A. Charles, Jr. is President and CEO of Cascade Policy Institute, Oregon’s free-market research center. Scott Shepard is a lawyer and was a visiting law professor at Willamette University during 2016. This essay is a summary of a case study of TriMet’s pension reform written by Mr. Shepard for Cascade Policy Institute. The full report is available here. This essay was originally published in the February 2018 edition of the newsletter “Oregon Transformation: Ideas for Growth and Change,” a project of Third Century Solutions.

Click here for the full report, Following in TriMet’s Tracks: Defined-Contribution Plans a Necessary First Step to Oregon’s Fiscal Health:


Read Blog Detail

Following in TriMet’s Tracks: Defined-Contribution Pensions a Necessary First Step to Oregon’s Fiscal Health

By Scott Shepard

Scott Shepard is a lawyer and was a visiting law professor at Willamette University in Salem, Oregon during 2016. He is the author of “A Lost Generation but Renewed Hope: Oregon’s Pension Crisis and the Road to Reform,” an academic study on Oregon state pensions published August 1, 2017 by the Mercatus Center at George Mason University. He is an Academic Advisor to Cascade Policy Institute, Oregon’s free-market public policy research center.


As recently as 2012, TriMet faced a pension funding disaster. Indefensibly overgenerous pension benefits granted in the early 1990s threatened to bankrupt the public transit system and to cripple the Portland metro area. While TriMet still has difficult reform ahead of it (regarding its other post-employment benefits promises), it has achieved pension fund stability by replacing its unsustainable defined-benefit pension promises with a well-designed, defined-contribution retirement plan.

“Defined-contribution” (DC) pensions are retirement benefit plans in which monthly payments are made by management into personal accounts owned by employees. Once those payments are made, the employer has no further financial obligations. The eventual pension payouts will be a function of the market performance of whatever investments are chosen by individual employees.

This stands in contrast to “defined-benefit” (DB) programs like Oregon’s Public Employees Retirement System (PERS). Under DB programs, employees are promised various levels of retirement payments calculated through arcane formulas that leave management mostly uninformed as to the level of funding obligation to which they have agreed. In many cases, those liabilities turn out to be much larger than expected. 

TriMet has brought its pension funding liabilities under control by moving its employees from defined-benefit plans to defined-contribution plans: first its management employees hired after 2002, then its unionized employees hired after 2012. The shift followed the lead of most private sector businesses, the federal government, and an increasing number of states. As a result of the change, TriMet’s pension obligations are moving steadily and reliably toward full funding within the near to medium term. This glide path to full funding is allowing the organization to focus on other vital personnel issues such as managing the cost of other post-employment benefits (“OPEBs,” which are primarily health care benefits for unionized workers) for current workers and retirees.

Oregon and its municipalities can only envy TriMet in this regard. The defined-benefit PERS funding costs continue to spiral out of control. These unbridled expenses are crushing local governments and school districts, forcing layoffs, hiring and wage freezes, bigger class sizes, reduced government services, and increased taxes. The failure to reform harms younger and more diverse workers at the expense of their older colleagues, and private-industry workers in favor of their government-employee neighbors. Taxpayers have said “enough,” voting 60-40 in 2016 against significant state tax hikes that inevitably would have been dedicated to helping to fund the PERS shortfall.

One necessary step toward addressing this problem is for the state of Oregon to follow in TriMet’s tracks, moving PERS workers from DB to DC plans. TriMet started down this road fully 15 years ago, while the state has dithered. Oregon must play catch-up by moving all PERS-covered workers to DC plans for work to be performed after the changeover.

This move by itself likely will not be enough to solve Oregon’s public pension crisis. The state has already promised more than it can reasonably pay. But moving to DC plans for all work not yet performed is a necessary first step. And the faster the legislature acts, the less severe—and the less upsetting to retirees and current and future employees—will be the other reforms required later.


Read Blog Detail