Examining calls to bring back Alaska’s defined benefit pensions
Photo 109416357 © Gregory Browning | Dreamstime.com


Examining calls to bring back Alaska’s defined benefit pensions

Bringing back Alaska's defined benefit pensions would be unlikely to improve retention or recruitment but could add $9 billion in unfunded liabilities.

Alaskan lawmakers replaced the state’s defined benefit pension plan for public employees with a defined contribution plan in 2005. Since then, there have been discussions and significant political efforts to undo this public pension reform and bring back guaranteed pensions to Alaska’s teachers and public workers.     

A recent report from the National Institute on Retirement Security (NIRS) studied Alaska’s ongoing challenges with recruitment and retention of public employees and concluded that bringing back defined benefit (DB) pensions would help alleviate the problems. Unfortunately, a good part of the analysis uses demographic experiences from as much as 20 years ago, when the job market and employee mobility were much different. This important context could lead policymakers down a very expensive path that, in the end, may not be all that effective.

A more comprehensive, modern perspective suggests that retirement benefits are unlikely to deliver better retention rates among the teachers and public workers Alaska hires, and different approaches are more likely to succeed. To the extent that benefits to Alaska’s current retirement plan are insufficient, lawmakers should consider alternative solutions that take into account workforce mobility.

The evolving environment of public employment

In its study delivered to the Alaska Department of Education, the National Institute on Retirement Security correctly identifies a growing challenge for Alaska’s government employers. Consistent with the trend developing at the national level, retention among public employees hired in the last five to 14 years—the range of years the NIRS study uses to single out Alaska’s defined contribution plan members—is significantly lower than governments experienced decades ago. The modern workforce in Alaska, particularly millennials (ages 25-40) that represent most hires after 2006, are increasingly transient and the most likely generation to switch jobs. They are frequently referred to as a job-hopping generation, as are the working-age members of Gen Z (typically defined as people born between 1996 and 2010). This trend of public workers changing jobs and employers—nationally, not just in Alaska—was pronounced throughout the 2010s and is likely to stay

This shift presents several problems for policymakers and those making hiring decisions in the public sector. A high turnover rate among the public workforce increases onboarding and training expenses. This churn also impacts the quality of service provided by public employees as fewer workers stick around to build their skills. There is no doubt that Alaska governments are finding it harder to deliver high-quality, cost-effective services to their constituents.

Where the NIRS report needs expanding, however, is in its analysis of the cause of this growing retention challenge. When examining differences in worker counts between those hired in 2005 (before the switch to a DC plan) and those hired in 2021 (the latest year for which employment data is available), the report finds, as expected, that newly hired teachers are making up less of the total population than they used to. The NIRS analysis also compares differences in retention assumptions between teachers in the defined benefit plan and the more recently hired teachers in the defined contribution plan. These retention assumptions—essentially predictions based on experience—are used by retirement system actuaries to generate standard cost and funding projections. Since actuaries have had to adjust their expectations due to the national shift toward shorter tenures, these results also, as expected, show that DC teachers are expected to remain at lower rates than those who were hired before 2006.

While the NIRS analysis gives some context on this retention challenge, its findings are too narrow to conclude that reverting to a defined benefit retirement plan would reverse a decades-long, national trend. The NIRS study merely identifies the developing retention tendencies among the latest generation of public workers but makes no effort to isolate the specific impact of switching new public workers to defined contribution plans.

For robust and reliable conclusions in social science research, the gold standard approach is to compare a group undergoing a policy intervention with a control group. This control group should mirror the intervention group in all aspects except for the exposure to the intervention itself. A Reason Foundation working paper takes this approach to compare teacher retention rates before and after Alaska’s pension reform and finds that the developing challenges in keeping educators cannot be attributed to the switch to a defined contribution retirement plan for new workers. It is safe to conclude that Alaska’s challenges are related to larger generational trends in employee behavior.

In Alaska, the population under the previous DB pension plan predominantly consists of Generation X (people ages 41-56) and older cohorts. In contrast, the defined contribution plan is primarily made up of millennials and some members of Generation Z. To ensure a more precise comparison between the DB and DC plans, it’s crucial to have comprehensive demographic information on individuals. This allows for the control of variables that could influence the decision to remain in the workforce, such as years of experience, education, and salary.

The NIRS study, however, is a forecast that relies on limited aggregate data on turnover rates. While it does account for age and gender, it lacks other essential characteristics and does not use individual-level data. This limitation impacts the ability to make definitive policy inferences from the report’s findings.

The NIRS paper does acknowledge larger trends among government employers, mentioning that “all states seem to struggle with retention of newly hired teachers.” It also hints at its own weaknesses, observing that the “data does not show that pensions are the only cause for lower retention rates among those in the DC plan,” while nonetheless concluding that the state’s DC plans are the primary driver of retention losses. As a result, the stated policy applications focus entirely on reversing these developments by bringing back a traditional DB pension and do not consider any other causes.

This failure to fully explore the source of the problems could be troublesome for Alaska because adopting these recommendations would be unlikely to effectively address the state’s recruitment and retention challenges and could bring very high public pension costs to future budgets. The costs associated with the defined benefit pension plan that closed in 2005 are much higher today than experts expected they would be, and the legacy pension plan is still billions of dollars in debt. Diving back into guaranteed pension benefits for public workers would likely expose Alaska’s budgets to more of the same runaway costs. A Reason Foundation analysis of a proposal to bring back defined benefit pensions last year revealed that this approach could generate additional costs upward of $9 billion.

Before taking on such a high level of costs and risks, Alaska policymakers should closely examine how likely public workers are to respond to a switch from the current DC structure back to the legacy pension approach.

Defined benefit pensions won’t solve Alaska’s recruitment and retention problems

A closer examination of how teachers and other public workers react to their retirement plans casts some major doubt on the effectiveness of this approach. A few recent academic studies found that pension reform has a negligible effect on worker retention.

One study by scholars from the University of Missouri and Beijing University of Technology found that pension benefit enhancements did not improve teacher recruitment or retention. Another study by the same authors found that enhancements to traditional pension plans actually accelerated the retirement of experienced teachers and reduced average teacher quality, whereas defined contribution plans had the opposite effect, the study found.

Another research paper by Andrew Johnson of the University of California at Merced found that pension reform had a negligible impact on the labor supply of public sector employees in Texas, and worker retention increased by 1-2 percentage points following the state’s pension reforms.

Several surveys reveal that retirement plans may not be a high priority for today’s workers, particularly teachers. A nationally representative survey of over 2,000 teachers in the Economics of Education Review found that early-career teachers were indifferent between traditional and alternative retirement plans. Moreover, teachers’ willingness to pay for traditional pension plans is less than their willingness to negotiate for other compensation elements, including salary growth, health insurance coverage, and Social Security enrollment, the survey found.

Another survey of 102 national service fellowship candidates found that among eight different benefits of public sector employment, retirement benefits ranked the lowest, far behind personal satisfaction, salary and health care benefits. These findings suggest that retirement benefits may not be a strong draw for the recruitment and retention of workers.

Additionally, according to a National Bureau of Economic Research working paper, teachers responded more positively to changes in salaries than to equally sized changes in pensions. The same study also found that higher salaries attracted higher-quality teachers, increased overall teacher quality, and improved student achievement.

A recent study by scholars from the University of Alaska Anchorage found that while salaries have a positive effect on Alaska’s teacher supply, non-wage benefits have a negative effect. Wage adjustments may be more effective for employers and policymakers to influence worker retention.

Potential retirement-focused solutions that would work for Alaska

Bringing back defined benefit pensions to Alaska would be very expensive and unlikely to improve recruitment and retention amidst the trend of a more mobile and transitory workforce. There are, however, some upgrades to the state’s current public retirement system that could be pursued not necessarily to achieve retention goals but to address the retirement security goals that are the core purpose of these programs.

For example, most members of Alaska’s Public Employee Retirement System (PERS) and Teachers Retirement System (TRS) do not participate in the federal Social Security program. The majority of PERS employers fill this void with the Alaska Supplemental Annuity Plan (SBS-AP), which automatically enrolls workers into a 6.13% contribution plan with an equal 6.13% employer match. While the contributions are similar in size to those of Social Security, retired members of this supplementary DC plan get to enjoy the investment returns that these funds generate during the years leading up to retirement, making it a vastly more appealing option compared to its federal counterpart.

While public safety and other state workers get to take part in SBS-AP, Alaskan teachers enrolled in the TRS system do not currently have this option, creating an uneven benefit design across different cohorts of public employees in Alaska. Consequently, TRS participants rely on either their primary DB or DC plan—depending on when they were hired—and whatever personal savings they have to supplement their own retirement, which puts a great deal of pressure on teachers.

Alaskan lawmakers should consider expanding the SBS-AP so that it is also available to the state’s educators. This would fill in the missing space that exists due to the lack of a Social Security benefit, and it would do so with a benefit that would be more valuable to its members and more affordable to its sponsor than the beleaguered federal program.


It makes sense for policymakers to seek out ways to improve the recruitment and retention of quality teachers and public workers, but the effectiveness and costs of these policies should be thoroughly evaluated. An analysis that does not account for recent shifts in workers’ behavior or employment trends is not a reliable template for reshaping Alaska’s public workforce, especially when the proposed policy changes are packed with major costs for future generations. It is unlikely that a switch back to defined benefit pensions would alleviate the state’s ongoing retention and recruitment challenges. Bringing back Alaska’s defined benefit pension could come with a hefty price tag exceeding $9 billion in new unfunded liabilities over the next 30 years.

Instead of seeing the state’s retirement systems as a way to attract and keep public workers, policymakers should make retirement security for all employees—not just those that stick around for decades—the goal. Rather than a tool to promote retention by dangling crucial retirement benefits behind tenure requirements, these plans should serve to provide adequate retirement savings for all members, no matter how long they remain employed by the state or political subdivision. That means any year spent as a public servant, be it year one or year 33, should be one in which ample retirement progress is made. This approach values individual flexibility and workforce competitiveness over tenure requirements, and it would ensure that the benefits being accrued are enough even for the large group of hires that are not likely to stay their entire career, both of which happen to be one of the advantages of the state’s current DC plan design.

Stay in Touch with Our Pension Experts

Reason Foundation’s Pension Integrity Project has helped policymakers in states like Arizona, Colorado, Michigan, and Montana implement substantive pension reforms. Our monthly newsletter highlights the latest actuarial analysis and policy insights from our team.

This field is for validation purposes and should be left unchanged.