Testimony on Alaska House Bill 55 and Buck’s updated fiscal analysis

Prepared for Members of the Alaska Senate Labor and Commerce Committee

Good afternoon, my name is Ryan Frost, and I’m a policy analyst with the Pension Integrity Project at Reason Foundation. I’ve been with Reason since 2019, and prior to that, I spent seven years as the Research and Policy Manager for the Law Enforcement Officers and Firefighters Pension System in Washington state, or LEOFF 2 for short. LEOFF 2 has been one of the top-three best-funded pension plans since its inception in the mid-1970s, and that’s primarily been accomplished by keeping up to date with best practices in pension plans and funding design.
Reason Foundation’s pension team has played a key pro-bono technical assistance role on 55 pension reforms over the past six years, including the two largest being the public safety plan in Arizona and the Texas employee’s retirement system. Those reforms include new defined benefit pension tiers, new hybrid design tiers, new cash balance pension tiers, and new defined contribution plan tiers.

Accordingly, our team is agnostic on plan design. We are not ‘defined contribution or bust’ zealots; in fact, most of the reforms we’ve worked on have included a defined-benefit pension component. But each of those reforms has at its very foundation a way of paying for the pension system that ensures costs do not eat into state and local budgets and that these important benefits earned by employees are fully funded for their retirement.

House Bill 55, which would open a new defined benefit tier for public safety officers and firefighters hired since 2006, does far too little to prevent growing unfunded pension liabilities. Supporters of HB 55 claim that certain tweaks to the “new” pension would eliminate the financial risk to the state. However, these tweaks have faced minimal actuarial scrutiny to support proponents’ claims, and there is no publicly available long-term actuarial forecasting or stress testing to justify such a financially profound policy decision.

While it does have a few modest improvements relative to the legacy pension tier, HB 55 is a weak pension design and still lacks sufficient controls to justify the assertion that there is no risk to state/local budgets, as there is with the current defined-contribution plan.

Recognizing the need for a long-term perspective on funding and costs, the Pension Integrity Project has prepared preliminary modeling of the proposed new House Bill 55 tier. It is important to note that these results do not include potential impacts on the legacy Alaska Public Employees’ Retirement System tier and its current $4.6 billion in unfunded liabilities.

The results of our analysis indicate that HB 55 would very likely expose the state to new and growing unfunded liabilities. Instead of the $743 million paid in employer contributions over the observed 30-year window if all current actuarial and demographic assumptions are met, Alaska would be responsible for paying $887 million instead due to the need to service growing unfunded liabilities. That analysis is available on our website.

Public pension systems operate over generations, but state legislators have only been presented with minimal five-year cost projections based on an assumption that the proposed pension tier would do the impossible: hit all its actuarial assumptions. In short, HB 55 only works as intended if Alaska PERS does something it has never once accomplished in its entire history—get 100% of its assumptions 100% right, 100% of the time.

Despite practically flying blind to the risks and long-term trajectory of the new pension tier, the Alaska House passed it out of its chamber.

The updated fiscal note from Buck is basically a carbon copy of its previous fiscal note, only being updated for results of the latest actuarial valuation. A proper actuarial analysis of the proposal still has not been conducted.
Take, for example, a bill recently passed in Washington State for its police and fire pension system. That system has been overfunded since its inception in the mid-1970s through prudent risk analysis and funding design. Due to decades of hard work, the plan managed to create a side account to pay (in full) for any future benefit increases. The first benefit increase was passed this year: a multiplier increase.

The Washington state actuary’s fiscal note for that multiplier increase—a relatively minuscule task when compared to a complete pension design overhaul—is a shining example of the level of detail that must be considered when adjusting a pension benefit. A few things that Washington state’s fiscal note includes that the Buck fiscal note does not:

• 25-year costs to the state of Alaska;
• Stochastic modeling to assess the effect of unexpected experiences under thousands of potential outcomes;
• The expected chance that contribution rates will exceed 10%;
• Sensitivity analysis on retirement rates due to the effects of bill passage;
• And, funded status trajectory going out 30 years.

Even with the limited scope and rigor of the updated fiscal note, Buck’s analysis still raises some major concerns for state policymakers, stating, “Adverse plan experience (due to poor asset returns and/or unexpected growth in liabilities) or changes to more conservative assumptions will increase the PERS DB [defined benefit] unfunded liabilities, resulting in higher contribution rates.”

On day one of this new tier being opened, the plan will have up to 15 years of liabilities already on its books because House Bill 55 stipulates that, for any member who chooses to enter the new tier, all their previously earned service in the DCR will be transferred using an unrealistically high discount rate of 7.38%. This number is completely out of step with national trends, where the national median is now dropping under 7%. This sets up a ‘pension obligation bond-like situation where any downturn in market performance or lowering of investment return assumptions would quickly create unfunded liabilities in the pension system.

Previous proposals to put public safety into a defined benefit plan, specifically House Bill 79 (2020), set the plan’s assumed rate of return at 7%. HB 79 is nearly identical to HB 55, apart from a few changes to assumptions and minimum employee contribution rates.

Why is this proposal going the opposite direction on investment risk than previous iterations by using a higher rate, 7.38%, than the previous proposal? Our hypothesis is that the 7% assumption would raise the normal cost too high for a cost-neutral argument to hold any water because even fewer dollars would be going to legacy debt. Additionally, Buck Global notes a critical point—the current defined contribution (DC) benefit offers nearly the same retirement benefit as proposed under HB 55, suggesting that simply increasing the employer contribution in the current DC plan and adding more annuity purchase options in the current plan could yield an equivalent benefit and provide lifetime income options. Buck finds, “On average, approximately 94% of DCR service as of June 30, 2021 was credited to PERS DB.”

This means that the assets in the current DC plan holdings would cover nearly all the equivalent liabilities in the proposed new HB 55 pension tier, suggesting that the current retirement offering could potentially be improved to achieve the same policy goals. It’s also another potential explanation for why the supporters decided to use a 7.38% discount rate. Using a 7% rate would make the service-credit purchase more expensive at the time of transfer, meaning that members entering the pension tier would be effectively taking a benefit cut.

What would it take to push the actuarially determined rate over the 9% employer contribution floor? This is the first question that should have been asked in the first committee hearing on this bill. We have provided a list of other questions that should have been asked in a document accompanying this testimony.

There is a strong likelihood—and you don’t know how much until you get a real actuarial analysis beyond the minimal Buck fiscal note analysis—that the arbitrarily high 7.38% discount rate will need to be lowered within the first five years of the new tier’s life to keep up with the larger pension system trends. When that happens, costs will necessarily increase.

House Bill 55 is being proposed by public safety associations concerned with recruitment and retention challenges. Proponents claim they are having trouble recruiting and retaining members due to the lack of a defined benefit pension for their members. However, this claim does not hold up to the data as 86% of police stations across the country are facing a shortage of members. Every one of those stations, outside of Alaska, has a pension with some defined benefit component. In fact, we have an academic working paper that shows retention rates did not change when Alaska swapped from a DB to DC in 2005.

The national trend since the Great Recession of 2007-2009 has been for states to adopt greater risk controls in their traditional public pension systems and move towards a variety of plan design options with the goal of avoiding re-exposing state and local budgets to the risks of worsening unfunded liabilities over the long-term. Texas, Arizona, Michigan, and Colorado are among the states that have recently adopted new, risk-managed pension-like guaranteed retirement plans that also do not disproportionately burden employers with financial risk.

Unfortunately, HB 55 does not resemble those types of prudent pension reforms but it could be redesigned to do so if given sufficient time and if informed by robust actuarial modeling and stress testing analysis.

In conclusion, there is a way forward that can meet whatever needs this legislative body and stakeholders desire, but House Bill 55, as currently written, would not put this new tier on a successful path. We’d be happy to work with the committee and stakeholders to help draft a design that best meets those needs and follows current best practices in pension design.

Thank you for your time today, and I’d be happy to answer any questions.