The Evolution of Teacher Retirement Benefits in California
By Jonathan Moody
How have teacher pension benefits managed by the largest teacher retirement system in the country changed over the past century?
Our team at Equable recently analyzed how growing pension debt costs for the California State Teachers Retirement System — commonly known just as CalSTRS—have been influencing the equitable distribution of K-12 education resources. While the focus on the project was on the intersection of retirement plan costs and school budgets, we were also curious about the relative effects CalSTRS pension debt might have had on teacher benefits in California.
In other research projects we’ve found that benefit enhancements, while common in many states around the turn of the century, have not been a primary driver of unfunded liabilities for public pension plans today. Separately, we’ve found that there has been a decline in the value of pension benefits offered to new teachers nationwide. The expected value of a pension at retirement for a full career teacher had steadily improved from the 1960s into the early 2000s. But since 2005, the average value of pension benefits being offered to new teachers has precipitously declined, as Figure 1 below shows.
We were curious whether there has been a similar, decline in benefits phenomenon in California. So we dug into the historic reports provided by CalSTRS (a model of transparency in this respect), to explore how the benefits of the nation’s largest public educator retirement system have shifted over time.
1. Over 100 Years of History and Growing
It is no exaggeration that the CalSTRS is among the largest retirement systems in the country. As of June 30, 2022, the plan held $300.1 billion in assets and was responsible for $369.5 billion in promised benefits to cover nearly 450,000 active K–12 teachers with more than 325,000 retirees currently receiving benefits. To put this into perspective, if the state were to pay off the $69.4 billion in unfunded liabilities in one year, that would consume more than a third of California’s 2021 general fund expenditures. CalSTRS is a pretty big deal.
To understand how the benefits authorized by the California legislature and provided through CalSTRS have changed over-time we have to go all the way back to 1913 when the retirement system was established. The first benefit was offered — an initial retirement pension of $500 per year paid in quarterly installments of $125. To receive this pension teachers were required to have recorded 30 years of service, at least half of which, including the 10 years immediately prior to retirement, needed to have been spent teaching in California schools.
That $500 per year isn’t a lot of money in 2023. But it was a much larger sum in 1913. If you adjust for inflation into 2022 dollars, the benefit works out to $15,118 a year.
We can estimate the total lifetime value of the first promised benefit for CalSTRS if we make a few assumptions. Assuming that a teacher started their career at age 25, retired at their earliest opportunity to receive full benefits (i.e., not under any early retirement provision), and lived until age 86, that $500 annual benefit in 1913 (or $15,118 in 2022 dollars), works out to an estimated total benefit value of $468,655 (again in 2022 dollars).[1]
But that is just the original 1913 plan. In the more than a century of serving K–12 educators, CalSTRS has offered a variety of benefits that were modified at different points to keep with the times. In Table 1 below, we have identified each of the changes to CalSTRS that would affect the value of the retirement benefits and have described both the change that was made and the implication that would have for estimating the value of the benefits.
There are various changes that the California legislature adopted to enhance benefits over time. Some of these have an influence on the expected value of benefits to an individual; others have an influence on the total value of benefits promised.
For example, the 1971 change that increased the multiplier used in determining CalSTRS benefits was an enhancement to individual benefit values — monthly pension checks would be higher, and the lifetime value of benefits would be higher too. These larger checks would also mean that the total size of benefits promised through CalSTRS (formally called accrued liabilities) would be larger.
By contrast, the 2001 and 2008 increases in the purchasing power of benefits was not a direct increase to the size of benefits. These changes were designed to protect retired members from future inflation if it was higher than estimated. But, the very nature of modeling benefits means making assumptions. Since the change to purchasing power minimums didn’t change the assumed rate of inflation, the adjustments didn’t necessarily mean increasing the estimated value of an individual’s retirement benefits.
We ran the list of benefit changes in Table 1 through Equable’s generalized benefit model (the brains behind supporting our Retirement Security Report project) and this allowed us to estimate both the annual value and lifetime value of benefits offered to California teachers from 1913 through today — shown in Figures 2 and 3 below.
The lines in each figure represent the average expected value of benefits (either lifetime value or annual value, depending on the image) for an average new teacher, earning an average salary, which increases at the current CalSTRS salary growth assumption.
By “new teacher” we mean someone who is starting their teaching employment for the first time in California. Because benefits can’t be reduced for people already in CalSTRS, we measure the value of benefits in each year as those being offered to a new teacher that is starting out, and the data reports what that teacher would expect to earn if they spent their whole career working as a teacher in California, earned the average salary, retired at the normal retirement age, and lived to the expected mortality age (86 years).
2. A Tale of Inflation, Not Just a 2023 Problem
As both Figures 2 and 3 clearly illustrate, the value of the benefits provided by CalSTRS through the first roughly 40 years of the system’s existence were not particularly generous. The $500, $600, and even $720 fixed annual benefits simply did not provide much in the way toward retirement (at least relative to expectations about retirement income today). In fact, the increase to an annual fixed benefit of $2,040 in 1954 (1 year after the legislation passed in 1953 to make the increase) marked the first time the value of teacher benefits in 2022 dollars exceeded the initial offering in 1913.
But the real game changer was the introduction of a service-based multiplier pension benefit in 1957 that, while still not offering any inflation protection for another almost 15-years, provided a benefit that would not decrease in value for those that retired in the years that followed. Then, the introduction of the cost-of-living adjustment (COLA) in 1972 made a marked improvement in benefit offerings that helped to ensure teachers would receive a quality benefit. This trend continued through the Classic 2% at 60 tier and into the most recent 2% at 62 (PEPRA) tier introduced in 2013, which marked the first decrease in the value of teacher pension benefits, both in total and annual benefit values, since the 1950s.
One of the key stories that come out of these data is the vital role inflation plays in determining the value of teachers’ benefits. While inflation has dominated economic headlines for the past year or so, the ebbs and flows over the last 100+ years have had varying effects. In almost all cases it has served to diminish the benefit being provided to California educators, but it was most profound for the first roughly 40 years of CalSTRS, where teachers were provided with set, fixed dollar annual retirement benefits. These benefits, because they remained fixed, actually saw a continued reduction in value with each given year. This outsized effect of inflation on the value of benefits continued until the introduction of the service-multiplier pension model — even though the plan offerings did not offer any COLA or other purchasing power protection until the 1970s.
To explain the reason why inflation was so much more impactful in the early years of CalSTRS, let us look at the value of benefits in 1913 that we calculated above. For a teacher that retired in 1913, their benefit would be $500 a year. When adjusted to 2022 dollars, that means they were receiving the equivalent of $15,118 a year in pension benefits. But what if a teacher retired in 1923? Despite it being 10 years later, the newly retired teacher would still receive a benefit of $500 a year as it was a set, fixed amount. This is a problem because once we account for the effects of inflation, simply retiring ten years later results in a smaller benefit, this time only worth $8,715 a year in 2022 dollars.
By contrast, if we look at the service-multiplier pension model, we see how the benefits do not decline in value with inflation, despite the lack of a COLA until the 1970s. This is because the value of the benefit is based on a teacher’s salary and years of service, meaning that the initial benefit for a newly retired teacher won’t decrease significantly from year-to-year, instead following the trends in salary growth over time. So in this case, a teacher hired in 1967 could expect to receive the same value of benefits as one hired in 1957, even though inflation reduced the value of the dollar over those ten years.
To be clear, without a COLA or other purchasing power protection the value of an individual teacher’s benefits would decay over time with inflation once they have retired. However, as this exercise is looking at the expected lifetime value of benefits at the point of retirement, we see various plateaus in value in Figures 2and 3.
The introduction of COLAs in 1971 and the respective changes to the COLA then result in increases in the estimated value of benefits for each new tier. Moreover, the value of benefits for teachers who had already left the classroom were notably improved in 1989 when the legislature established a state-funded policy that protects the purchasing power of retirees. Initially this was set at 68.2%, meaning that even if inflation outpaced any COLA, a teacher’s benefits were guaranteed to retain at least 68.2% of their original purchasing power. Moreover, the cost of any benefit enhancements were shouldered by the state, meaning that school districts were not left to figure out how to pay for these bumps to the value of retiree benefits.
In the years since, this level of protection was increased to 75% in 1997, 80% in 2001, and then up to 85% in 2008. To this end, retired teachers in California are afforded a secure benefit both through their plan’s COLA, but also through this supplemental protection against inflation.
3. Comparing the Accrual of Benefits
Whereas Figures 2 and 3 clearly display the value of retirement benefits afforded to teachers who reach normal retirement age over time, it is also a useful exercise to compare benefits over the course of a teacher’s career so that we can understand how much their benefits are worth in the years leading up to retirement as well. Figure 4 does exactly this, displaying the various tiers of benefits from 1957 onward. We note, however, that Figure 4 does not depict the value of any of the fixed dollar amount benefit tiers (those offered from 1913 though 1957) because those plans were, in every sense, all or nothing. Teachers who didn’t reach the full criteria would not receive any benefits, while those that did reach the required years of service and/or age would receive the full amount (or more in some cases).
What jumps out most clearly from Figure 4 is the difference in the speed of benefit accrual.
The two earliest plans lacked any COLA, but more importantly, they simply had a lower benefit multiplier of 1.667% whereas all of the other plans enjoy a 2.0% base multiplier. For the middle three tiers spanning from 1972 through 2013, we see a much faster accrual path, as they enjoy the same base multiplier, roughly the same COLA (the Pre-1999 tier used CCPI rather than a fixed rate), and a higher employee contribution rate (we assumed all legacy tiers to use the same rate as the Classic 2% at 60 tier). These policy similarities resulted in very comparable benefits across the three tiers.
The Classic 2% at 60 tier enjoys the highest benefit potential, as it also included policies that could increase the benefit multiplier as high as 2.4%. The current 2% at 62 (PEPRA) tier demonstrates a much slower accrual path as it features lower employee contributions and has a higher age requirement for normal retirement than the earlier tiers. In addition, the current tier introduced a much lower cap on the maximum salary eligible to be used for pension benefits. However, we notice the slight jump in benefits beyond age 62 for the current tier as it also features policies that can enhance the multiplier up to 2.4%, but it requires slightly more service than the Classic 2% at 60 tier.
See the notes at the end of this article for a technical commentary on the maximum pensionable salary rules under the PEPRA tier.[2]
4. Conclusion: The Evolution of CalSTRS’s Benefits
So what then have we learned about the benefits offered by CalSTRS over the past 110+ years?
First we can see how in the earliest days, CalSTRS provided what could have been considered a generous benefit, but by using fixed dollar totals, California educators’ retirements were very much subject to the ebbs and flows of inflation for the first roughly 40 years of the retirement system.
Second, as CalSTRS adopted more “modern” service-multiplier based pension design, we see the benefits quickly transform into something much more comparable to what we are familiar with today. The various changes to the benefit design through the 1970s, 1980s, and 1990s led to increasingly more generous benefits.
Last, we can see how the most recently introduced tier, CalSTRS’s 2% at 62 (PEPRA), represents the first reduction in retirement benefits since the 1950s, when the value of a teacher’s benefit was shrinking not because of an active policy choice, but because of inflation’s effect on a fixed-dollar benefit.
Although the 2% at 62 (PEPRA) tier is still very similar to its predecessor, the elevated retirement age requirement, lower employee contribution rate, maximum pensionable salary cap, elimination of the career factor (adjusting the multiplier for teachers that serve 30+ years), and modifications to the age factor (again adjusting the multiplier for service beyond normal retirement age) have all combined to result in less generous benefits for new California teachers being hired today.
We are fortunate to have had this opportunity to compare the value of benefits over time due to the care of the detailed reporting by CalSTRS regarding the tiers of benefits offered over time. This is not common, at least in publicly distributed documentation by public pension plans. California and CalSTRS should be commended for the transparency provided in this instance.
____________________
Notes:
[1] The age 86 mortality assumption is a conservative estimate that is intended to be a general measure that lines up with current actuarial mortality assumptions. Mortality rates in the early 20th century were higher, and age 86 is also higher than the average pre-pandemic life expectancy rate in California generally according to the CDC (link here). Using a lower assumption would mean a relatively lower lifetime value of benefits; using a higher assumption would mean relatively higher estimated lifetime benefit values.
[2] A Quick Aside on Maximum Pensionable Salary: The current 2% at 62 (PEPRA) tier introduced a cap on the maximum salary eligible to be considered for a teacher’s pension benefit. As of the 2021 actuarial valuation report, the most recent one available, that cap was set at $146,230. Notably, the very same report documents that the average starting salary for a teacher is roughly $56,550 and the average salary among active teachers is $78,928, suggesting that it seems unlikely that teachers would hit the cap on their maximum pensionable salary. While this would seem to be the case, our model operates using the average starting salary and then applies the assumed merit and inflation wage increases used by the retirement system to project a teacher’s salary over the course of their career. We also model the maximum pensionable salary cap to increase in accordance with CalSTRS’s inflation assumptions, allowing that to also grow over time. The result of these projections reveal that a new teacher would be expected to have their salary hit the cap after roughly 20 years. Yet, the data from the valuation report again seem to indicate that salaries for California teachers don’t tend to get that high. The answer to this discrepancy is, unfortunately, that we are bound by the data available. We use the plans’ most current assumptions regarding salary and merit increases and the projected salary is what it is. However, we’ve also seen in the actuarial gain/loss data for CalSTRS that shows how between 2001 and 2022 unfunded liabilities were reduced by $22.3 billion due to actual demographic experience relative to assumptions. That is, CalSTRS has been assuming certain salary growth rates, retirement timing patterns, and mortality rates that all added up were so wrong relative to what actually happened over the last two decades that CalSTRS overestimated the size of pension benefits it would have to offer — at least relative to just those assumptions. On the whole, the assumptions used to fund CalSTRS have been off target enough to result in tens of billions in unfunded liabilities. But, because salaries grew slower than expected, among other demographic experiences, the total amount of CalSTRS unfunded liabilities today are lower than if California schools had increased payroll at rates the retirement system assumed. So, while our salary estimates end up hitting the maximum pensionable salary cap and our modeled value of benefits are subject to that, it would seem that CalSTRS is actually benefitting from overestimating how fast teacher salaries grow.
