Welcome to the Senate Republican Press Search.

View Article Details

Print

No, Public Pension Reform Experiments Have Not Failed

Forbes Online

Friday, August 16, 2019  |  Commentary  |  Elizabeth Bauer - Contributor

Pensions (70)
Earlier this week, the National Institute on Retirement Security published a new report, "Enduring Challenges: Examining the Experiences of States that Closed Pension Plans," by Tyler Bond and Dan Doonan, which evaluates the developments subsequent to pension reform in four states: Alaska, Kentucky, Michigan, and West Virginia. Their bottom line claim:

"Switching from a defined benefit pension plan to a defined contribution or cash balance plan did not address existing pension underfunding as promised. Instead, costs for these states increased after closing the pension plan."

Are they right?

Let's start with a brief reminder: when one reads reports of pension underfunding in a given state, this is always based on benefits for existing employees for work performed up to this point in time, with relevant incorporation of the impact of projected pay increases for current employees and projected COLA increases for all.

This is a debt that will continue to exist regardless of whether new employees are moved into another type of plan. If a pension reform is implemented in which existing accruals are left untouched, this debt will continue to exist. Only in the case in which past accruals are affected will the existing debt be reduced — a reduction which might take the form of calculating retirement benefits using pay at the time of a benefit freeze rather than at retirement or reducing/freezing COLA adjustments, to take two examples that might be a part of a typical public pension reform rather than a more drastic bankruptcy-driven version.

In a case in which new employees receive a lower-value benefit formula, the savings from reduced contributions might help the state pay down this debt - but the value of a new/young employee's accrual is small enough relative to employees closer to retirement that the impact of this change won't be felt for many years. If a pension reform affects existing employees, then, yes, there is a real immediate savings.

At the same time, in Defined Benefit actuarial valuations, the value of the benefit accruals is small for young employees, high for older, closer-to-retirement workers. In a switch to a Defined Contribution system for new employees, assuming a fair and appropriate employer contribution, in the short term, overall system expenses will increase as the new employees receive percent-of-pay contributions that are higher than the very small accruals the employer would have to reserve for them in a DB plan. And, what's more, employees who are required to contribute to the plan may even subsidize the plan itself, relative to their accruals, in their younger years. (The negative normal cost of Illinois' Tier 2 employees is an extreme case due to the benefit cuts, but this is true of contributory plans generally-speaking.)

But the benefit in moving to a Defined Contribution system, or any similar system (e.g., a risk-sharing system, a level-accrual cash-balance system), is to ensure that the state pays what it promises rather than pushing off contributions into the future for a later generation, and to prevent the state from overpromising and sticking future generations with the bill. (See "Why Public Pension Pre-Funding Matters (An Explainer)" for a more complete explanation.)

Finally, a common argument in favor of open Defined Benefit plans is that in such a plan, employers can take greater risk because of the longer time-horizon of their investments. But all DB plans should take into account the mix of participant groups; the need to invest more conservatively as closed-plan participants age is, in the bigger picture, balanced out by the ability younger participants have, in their Target Date Defined Contribution funds, to invest more aggressively.

That being said, what actually happened in their examples?

In Alaska, the teachers' and public employees' Defined Benefit plans were closed to new entrants, who were moved to Defined Contribution plans starting in 2006.

Separately, in 2005, the state had an unfunded liability of $4.1 billion. In part, the plan's actuary, Mercer, was blamed for valuations which understated funding requirement. But even in the years since proper Actuarially Determined Contributions were calculated, Alaska still underpaid its contributions. And, yes, Alaska lost the subsidy-contributions of the new employees now accruing benefits in their own accounts, and will over time need to shift its investments to reflect conservatism more appropriate for aging participants. None of this is surprising, or suggests that there are any problems inherent in a switch.

Of greater concern is the question of retention. To whatever extent the trouble is that the Defined Contribution amounts are simply too low, is not a reflection on the appropriateness of one type of plan versus the other. But because Defined Benefit pensions so disproportionately benefit full-career employees, often with generous early-retirement subsidies for attaining given levels of service, they serve to handcuff employees to a given state; in particular, the NIRS report identifies an increase in recruitment and retention challenges. And admittedly, there is not necessarily an easy answer here, though it is also worth noting that there are options for Defined Contribution contributions to backload benefits to mimic a Defined Benefit accrual structure, and yet, at the same time, it is generally considered a "plus" for Defined Contribution plans to offer meaningful benefits for younger workers in the first place.

In Kentucky, participants hired in 2014 or later now receive a cash balance pension benefit rather than a traditional final-pay benefit. That this did not fix pension funding issues is not a surprise; it is in the nature of a cash balance plan (assuming that in the long-run, the overall benefit levels are unchanged) that benefit accruals are shifted and accrue evenly rather than being heavily backloaded to the years just prior to retirement). The report rightly pins the blame on the state's historical ongoing failure to fund these plans, though, of course, there comes a point at which this is all a bit circular, as it is generous benefit levels in the first place which bring a state to the point at which it deems its contribution levels "unaffordable."

In Michigan, the state closed its Defined Benefit pension plan for state employees (SERS) 22 years ago in 1997. At the time it was closed, it was 109% funded; as of September 30, 2017, it was only 66.5% funded. What went wrong? The report doesn't say, but an examination of the state's financial reports shows a persistent failure to pay required contributions in the early and mid-2000s, even before the recession hit. Again, this is not proof of the failure of a move to DC benefits. (Why didn't they pay their contributions? It is true that the contributions grew at a fast clip during this time, which is inevitable whenever contributions go unpaid, due to the need to make good for past year's underpayments; whether there are further asset or liability losses for other reasons contributing to the required contribution increases is not apparent in the reports.)

What's more, that a move to a DC plan is not a cause of the underfunding can be seen in a comparison to the teachers' pension plan. This plan was likewise fully funded in 1997 (though, to be sure, 100.9% rather than 109% funded), and similarly declined to even poorer funding levels, currently 61.6%. If the drop in funding for the state employees was somehow precipitated by the reform, one would expect a different fate for the un-reformed plan.

In West Virginia, the Teachers' Retirement System (TRS) was closed in 1991. In the early 2000s, the state concluded that it could "provide equivalent benefits at half the cost of the defined contribution plan" if it reopened the Defined Benefit plan (the report footnotes this with a reference to their 2008 study rather than that state's own research), and in 2005, they reopened the DB plan and allowed teachers to switch in, improving the plan's funding from 25% in 2005 to 70% in 2018.

Is this proof that Defined Benefit plans are the right way to go? According to a report at TeacherPensions.org, the closing of the plan to new hires was not the cause of the 25% funding level but, in fact, it was even worse at the time of the plan's closing, with a ratio of 11.6% in 1994. And how did they boost their funded ratio up to 70%? In part, with a Pension Obligation Bond (POB) in 2007 which boosted its funded status from 31.6% to 51.3% in one fell swoop. In addition, for nearly every year from 2000 onwards (the earliest data available online), the state made contributions of one-third of payroll or more, not including the costs of paying back the POB. This is not magic, and this is not tied to the closing and re-opening of the DB plan.

The bottom line is this:

It is entirely possible for a state or city to properly fund a Defined Benefit pension plan; it just takes discipline, plus either a risk-sharing design or a willingness to suck it up and pay more when the math requires it.

It is likewise possible under either a Defined Benefit or Defined Contribution system for the state to promise a retirement benefit that meets its employees needs for adequacy while not overspending. (Employers are also perfectly free, in a Defined Contribution system, to make contributions to employees unconditionally rather than, as is typical in the private sector, requiring a match.)

But it is only a Defined Contribution program that can wholly remove from legislators the temptation to ask (OK, require) future generations to pay for this year's benefit accruals, sometimes leaving them with bills of such magnitude as to imperil provision of basic human services. Only a Defined Contribution ensures that the retirement benefit is always by definition 100% funded. And only in a Defined Contribution program is the benefit defined up front and transparent to all.

And that, dear readers, is the purpose of pension reform.