Teacher Pensions Blog

  • Late last month, the U.S. Census Bureau released an update of spending in American elementary and secondary schools. I've already dug into the national numbers, but it's also worth taking a look at how spending patterns have changed across the states over time. 

    Overall, schools spent* about 5.4 percent more per pupil ($628) in 2017 versus 2008, in inflation-adjusted dollars. But these national figures hide significant variation at the state level both in terms of total spending and in how states are choosing to spend their money. As with the national figures, states are spending more and more of their education budgets on things like employee benefits and other support services while spending less on teacher salaries. 

    Other reports, like this one from the Center on Budget and Policy Priorities (CBPP), have found that some states are still spending less than they did prior to the 2007-9 recession. I'm looking at a slightly different calculation than CBPP did--they appear to be looking at revenues as opposed to expenditures, and it's not clear if they were looking at total spending or only current spending, my preferred metric. Still, I also find that K-12 spending is down in real terms over the last 10 years in 15 states: 

    K-12 Spending Is Below Pre-Recession Levels in 15 States 
    Florida............................................................................-11.0%
    Oklahoma.......................................................................-8.4%
    Georgia..........................................................................-7.6%
    Alabama........................................................................-7.4%
    Arizona...........................................................................-6.7%
    Idaho..............................................................................-4.3%
    Colorado.........................................................................-4.2%
    New Mexico...................................................................-3.4%
    Mississippi.....................................................................-1.6%
    Indiana...........................................................................-1.5%
    Virginia...........................................................................-1.2%
    Wisconsin.......................................................................-0.7%
    Nevada...........................................................................-0.3%
    Louisiana........................................................................-0.3%
    Texas.............................................................................-0.1%

    The Census Bureau breaks down their spending figures into a few different categories, and I'll look at them separately. First up is total salaries and wages. Nationally, that figure is up but only by 0.3 percent, in real terms. 23 states are spending less on salaries and wages than they did pre-recession: 

    Spending on Salaries and Wages Is Below Pre-Recession Levels in 23 States 
    Florida............................................................................-16.2%
    Oklahoma.......................................................................-13.3%
    Georgia..........................................................................-10.0%
    Arizona...........................................................................-9.9%
    Louisiana........................................................................-9.5%
    Alabama........................................................................-8.0%
    Idaho..............................................................................-7.7%
    New Mexico...................................................................-7.7%
    North Carolina.................................................................-7.0%
    Michigan........................................................................-6.6%
    Indiana...........................................................................-5.8%
    Kentucky........................................................................-4.7%
    South Carolina.................................................................-4.6%
    Virginia...........................................................................-4.4%
    Mississippi.....................................................................-3.4%
    Arkansas........................................................................-2.8%
    Nevada...........................................................................-2.6%
    Kansas..........................................................................-1.6%
    Texas.............................................................................-0.9%
    New Jersey.....................................................................-0.8%
    Wyoming.........................................................................-0.5%
    Missouri........................................................................-0.3%
    Colorado.........................................................................-0.2%

    The trends look even worse when you focus solely on what the Census Bureau calls salaries and wages for instruction. This category separates out things like school and district supports and administration and is the closest we can get to identifying how much money schools paid directly to teachers. Unfortunately, of all the Census Bureau categories, this is the only one that actually declined over the last 10 years. Nationally, salaries and wages for instructional employees fell by 0.3 percent over the last 10 years in real, per pupil terms. In fact, as of 2017, 28 states were spending less on instructional salaries than they were 10 years prior: 

    28 states are spending less on instructional salaries than they did pre-recession 
    Florida............................................................................-16.8%
    Oklahoma.......................................................................-15.9%
    Arizona...........................................................................-14.1%
    Louisiana........................................................................-13.5%
    Georgia..........................................................................-11.7%
    Alabama........................................................................-8.5%
    Indiana...........................................................................-8.4%
    New Mexico...................................................................-8.3%
    Idaho..............................................................................-8.2%
    North Carolina.................................................................-7.8%
    South Carolina.................................................................-7.4%
    Mississippi.....................................................................-7.1%
    Kentucky........................................................................-6.5%
    West Virginia...................................................................-5.0%
    Arkansas........................................................................-4.6%
    Virginia...........................................................................-4.3%
    Michigan........................................................................-4.0%
    Colorado.........................................................................-4.0%
    Maryland........................................................................-2.8%
    Texas.............................................................................-2.8%
    Missouri........................................................................-2.4%
    Wisconsin.......................................................................-2.3%
    New Jersey.....................................................................-1.3%
    Nevada...........................................................................-1.3%
    Tennessee.......................................................................-1.2%
    Wyoming.........................................................................-1.1%
    California........................................................................-0.4%
    Kansas..........................................................................-0.3%

    It's probably no coincidence that states with notable recent teacher strikes--including Oklahoma, Arizona, North Carolina, Kentucky, West Virginia, Virginia, and Colorado--all made this list. 

    In addition to salaries and wages, the Census Bureau also reports spending on employee benefits, including pensions and health care. In contrast with salaries and wages, which were flat or declining, benefit spending rose 23.5 percent, in real terms, over this same time period. Only five states managed to keep their benefit spending in check, whereas 31 states plus the District of Columbia experienced double-digit (!) increases in benefit spending: 

    31 states and the District of Columbia experienced double-digit increases in benefit spending 
    Ohio...............................................................................12.3%
    Montana........................................................................12.3%
    Nevada...........................................................................13.3%
    South Dakota..................................................................13.7%
    Kansas..........................................................................14.1%
    Oregon...........................................................................14.8%
    Tennessee.......................................................................14.9%
    Missouri........................................................................16.2%
    New Jersey.....................................................................17.6%
    Maryland........................................................................17.6%
    Iowa..............................................................................17.9%
    South Carolina.................................................................19.7%
    Minnesota......................................................................20.3%
    Utah................................................................................21.1%
    Kentucky........................................................................24.3%
    Wyoming.........................................................................26.2%
    Nebraska.......................................................................26.8%
    Louisiana........................................................................29.9%
    Michigan........................................................................30.3%
    Delaware........................................................................30.5%
    Hawaii...........................................................................30.8%
    North Carolina.................................................................30.9%
    Washington.....................................................................32.7%
    New Hampshire...............................................................37.0%
    California........................................................................37.3%
    Vermont..........................................................................39.0%
    New York.......................................................................41.3%
    Connecticut....................................................................49.4%
    North Dakota...................................................................62.0%
    Pennsylvania..................................................................68.9%
    District of Columbia..........................................................105.6%
    Illinois............................................................................152.3%


    To be clear, increased benefit spending has not led to improvements in employee benefits. Most of these cost increases are due to paying down pension debts or changes in accounting rules on retiree health benefits. Teachers should be concerned that rising educational expenditures are not ending up in their pockets but are instead being diverted toward other purposes.

    While benefit costs were the fastest-rising category of spending, schools are also spending more on student supports (up 14.6 percent in real terms), general administration (up 6.5 percent), and school administration (up 5.7 percent). 

    There are a couple main lessons here. One is the importance of digging beneath the national numbers. There's quite a bit of variation playing out in the states. But there are also varying trends across categories. And while it's not completely universal, it's worrying that benefit spending, as well as administration and other support services, are eating up larger and larger shares of education spending. If teachers are upset about flat or stagnant salaries in their states, they should start by looking to see whether total education funding has gone up in their community, and then look to see where the money is going. 

    *Throughout this post, I'm going to be looking at current spending in per pupil, inflation-adjusted dollars. Current spending excludes things like capital and debt costs. 

    Taxonomy: 
  • When most people think about how teachers enter the profession, they might think of what could be called a traditional route--student teaching during college, followed by a full-time teaching job beginning at 22 or 23 years old. 

    While this is still the typical path into teaching, only about 55 percent of incoming teachers in American public schools start out this way. Twenty percent enter in their late 20s, 16 percent enter in their 30s, and the remaining 9 percent enter after age 40. These data come from a representative sample of American public school teachers surveyed by the National Center for Education Statistics in the 2011-12 school year. 

    Still, these national data hide quite a bit of variation by state. At one end, Kansas and Iowa teachers are much more likely to take the "traditional" path. In both those states, more than 70 percent of teachers begin their careers by age 25, and more than 85 percent enter the profession at some point in their 20s. 

    In contrast, states like California and New Mexico have very different patterns. Less than 40 percent of teachers in these states enter the profession by age 25. In fact, New Mexico has the highest percentage of teachers who begin their careers after the age of 40, at 17.6 percent of their teachers. 

     Age when first started teaching
     20-2526-3031-3536-4041+
    National Average55.0%19.9%8.8%7.0%9.3%
      New Mexico33.7%23.4%10.9%14.5%17.6%
      California39.4%25.9%13.8%8.1%12.8%
      Alaska41.8%24.3%13.7%9.0%11.1%
      Rhode Island43.1%24.7%12.0%7.9%*12.2%
      Nevada44.0%21.0%11.8%*7.1%*16.1%
      Arizona45.4%21.9%10.0%7.5%15.1%
      Idaho46.1%22.1%14.0%8.8%8.9%
      Florida46.2%23.2%11.2%5.8%13.7%
      Texas48.5%22.6%7.8%8.2%12.8%
      Hawaii49.0%25.7%*10.3%*6.8%*8.2%*
      Oregon49.2%25.1%8.8%7.4%9.5%
      Vermont50.2%22.0%9.2%8.9%9.7%
      Georgia51.2%23.6%7.4%8.5%9.4%
      Utah52.0%23.5%6.8%7.3%10.4%
      Wyoming52.1%26.3%9.7%*6.2%*5.7%
      Washington52.2%21.7%9.2%7.3%9.6%
      Oklahoma52.4%22.4%8.2%7.8%9.1%
      New Hampshire52.5%16.5%11.5%8.3%11.3%
      Colorado53.4%21.4%9.5%6.8%8.8%
      Maine54.3%15.6%12.2%10.2%7.7%
      North Carolina55.2%16.6%9.5%9.3%9.4%
      Mississippi55.4%17.7%9.5%7.3%*10.1%
      Massachusetts55.7%18.4%7.6%7.8%10.6%
      West Virginia55.7%17.2%9.4%8.4%9.4%
      New York56.3%19.2%9.0%6.4%9.1%
      Kentucky56.5%20.7%9.5%5.8%7.5%
      Louisiana56.6%19.3%11.1%7.8%5.3%
      Tennessee56.8%15.1%11.0%7.5%9.6%
      District of Columbia56.9%19.7%8.5%*4.9%*10.0%*
      Connecticut57.7%21.3%7.9%5.0%8.1%
      Alabama58.1%18.8%9.1%7.7%6.3%
      Virginia58.3%15.2%6.9%9.0%10.6%
      New Jersey58.4%19.8%6.1%6.9%8.8%
      Michigan58.6%19.9%7.1%8.2%6.2%
      Montana60.7%17.9%8.9%5.7%6.8%
      Arkansas61.0%17.6%9.0%5.9%6.6%
      Delaware61.1%15.3%9.4%*5.7%*8.5%*
      Illinois61.6%17.0%8.3%6.2%6.9%
      Minnesota62.4%19.5%6.8%5.3%6.0%
      South Carolina63.4%16.8%8.7%5.0%6.2%
      Missouri63.5%18.6%6.7%5.8%5.5%
      Pennsylvania63.8%16.3%8.0%5.7%6.2%
      Ohio64.2%16.0%7.4%6.0%6.4%
      Wisconsin65.2%15.5%7.4%5.4%6.5%
      Indiana68.2%15.7%6.7%4.9%4.5%
      South Dakota69.0%17.5%5.9%*3.8%*3.8%*
      Maryland69.7%13.6%4.7%*4.9%*7.1%*
      North Dakota69.7%15.4%7.2%*3.5%4.2%*
      Nebraska69.9%18.4%4.7%2.6%*4.4%
      Kansas70.5%15.2%7.0%3.8%3.5%*
      Iowa71.3%15.0%5.7%4.2%3.8%*

    Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS), Public Teachers Data File 2011-12

    *Represents small sample size, interpret with caution 

    This question may be interesting in its own right way, but the answers have financial implications for teachers. Because teacher pension formulas are based on a teacher's salary in the last year  they taught, regardless of when that happened to be, those formulas offer greater rewards for late-career service than than they do for the same years performed earlier in the teacher's career. So, for example, the same pension formula would be more beneficial for the typical teacher in California or New Mexico than in Iowa or Kansas, simply because California and New Mexico teachers are already closer to retiring. (See here for a longer explanation with examples for teachers of various ages.) 

    Teachers who start at younger ages do have the potential to eventually earn larger pensions by the time they retire, but, because they have more years to go, they're also more likely to leave before then. Younger teachers also face different and more difficult calculations about what to do with their pension. For a 35-year-old with 10 years of service who decides to leave the profession, it may make sense to cash out their contributions and roll over that money into an interest-bearing investment account. For someone closer to retirement, the pension may be a better deal. 

    From a public policy standpoint, it doesn't make a whole lot of sense to prioritize some teachers over others. But that's exactly what teacher pensions do. 

  • The Census Bureau’s Annual Survey of School System Finances compiles total education spending and revenue across the entire country. The latest data, released earlier this week, shows teacher benefits continue to eat away at school budgets. 

    In total, public school expenditures have nearly tripled since 1992 (including inflation). The table below shows how those dollars are being spent across broad categories. Figures for all years reflect current expenditures and do not include capital costs or debt. 

    As the table suggests, school districts are spending more and more of their budgets on employee benefits, at the expense of base salaries and wages. Over the long term, the percentage of education budgets going toward salaries fell from 65.0 to 56.2 percent, while the percentage spent on employee benefits increased from 15.0 to 23.7 percent.

    Public School Expenditures, 1992-2017

    As a share of total expenditures, benefits now eat up almost eight percentage points more than they did in 1992. 

    These are steady, long-term trends, but they're no less troubling. Increased spending on benefits is one reason teacher salaries have been flat, in real terms, over the last few decades. Benefits are also a less efficient use of money than salaries, since teachers value $1 in take-home pay more than they do $1 in benefit spending. Worse, much of these rising benefit costs reflect the growing burden of debt costs, not actual benefit enhancements for teachers and retirees. 

    Updated on May 23, 2019

    Taxonomy: 
  • As I wrote about previously, Arizona’s teacher pension system is complex, expensive, and fails to produce an adequate retirement benefit for the majority of its teacher members.

    Over the past decade, the problem has been getting worse. The Arizona State Retirement System (ASRS) has simultaneously cut benefits and raised costs for teachers.

    An important way to look at the impact of these changes on teacher retirement benefits is to analyze teachers’ retirement benefits net of their own contributions. As we’ve written about extensively, pension systems are back-loaded, which means that for years, often decades, teachers can actually have a net negative retirement benefit (that is, the pension they would qualify for is worth less than what they themselves contributed).

    To show how the various changes to the ASRS system affect retirement wealth for Arizona teachers, I modeled how the benefit structure has changed over the past decade. Each curve in the graph below illustrates how benefits would accumulate for a new teacher hired in that year.

    As shown in the graph, Arizona teachers’ net retirement benefits have decreased with each change to Arizona’s system. Part of this is due to rising teacher contribution rates, which will go from 9.45 percent in 2010 to 12.51 percent in 2020, according to the state’s projections. Teachers may feel that as a decrease in their take-home pay, but it also means a cut in their net retirement benefits. All told, a new hire in 2020 will have a much less valuable retirement plan than a teacher who was hired in 2010.

     

    Source: Author’s calculations based on ASRS financial reports.

    Each line represents a new entrant in a given year. The blue line shows a teacher’s net benefits in 2010. In a set of reforms enacted in 2011, ASRS raised teacher retirement ages and despite a 50-50 split of the total benefit cost, teachers assumed a greater share of the normal cost of benefitis since the employer contribution is mostly going to pay unfunded liabilities. The orange line shows the result of those reforms. The reduction in benefits shifted the curve to the right, while subsequent increases to teacher contribution rates have continued to depress the curve, producing lower net benefits. 

    Altogether, pension design issues affect how much teachers get out of the plans. Due to benefit cuts and cost increases, teachers are getting less bang for their buck as their net retirement benefit continue to fall. 

    This problem is not unique to Arizona. States across the country are asking teachers to pay more for lower benefits. But just as teachers in Arizona earned a much-deserved raise last year, they also deserve higher-quality retirement benefits.

  • Merely vesting in a pension plan is not sufficient to guarantee a decent benefit, and many vested teachers would be better off withdrawing their contributions than waiting to collect a pension. 

    That reality has taken me a while to grasp, because it's counter-intuitive, and because vesting in other types of plans matters much more. So let me start by explaining how vesting works in defined contribution, 401k-style plans. In a defined contribution plan, the employer makes a contribution as a percentage of the employee’s salary. Employees qualify for those employer contributions after they reach a certain number of years of service, called the vesting period. Employees who leave before vesting forfeit their employer's contributions.

    Say an employee starts out with a salary of $40,000, and the employer contributes 5 percent into the employee’s 401k. In the worker's first year of employment, the employer contributes $2,000 (5 percent of $40,000) to the employee's 401k, but that money is not vested yet. The employee is entitled to that money only if he or she reaches five years of service; short of that, and the employer keeps it.

    This is part of the reason why, in a new report out this week, we found that immediate vesting would significantly improve the percentage of workers who qualified for adequate retirement benefits in defined contribution and cash balance plans. In contrast, immediate vesting would do virtually nothing for the vast majority of teachers in defined benefit plans. That's because vesting works differently in these types of plans. 

    An example may help here. Consider the case of a 25-year-old teacher who started her career last fall in Los Angeles. She’s automatically enrolled in the California State Teachers Retirement System (CalSTRS). CalSTRS has a five-year vesting period, which means she’ll first qualify for retirement benefits if she leaves after her fifth year of service, at age 30. But she won’t be eligible to begin collecting her pension benefit until she reaches age 62, in the year 2055, and it will be based on her salary in her final years of service, in the year 2023. By the time she's able to begin collecting, inflation will have significantly worn away her benefit.

    Critically, this distinction affects young teachers much more than older ones. That’s because older teachers are already close to retirement, and they won’t have to wait as much time for inflation to wear away their benefits.

    The graph below, from an Urban Institute report by Richard Johnson and Benjamin Southgate, helps show what this looks like. Each line represents the value of a teacher’s pension benefit, net of her own contributions, depending on when she started teaching and how long she serves. I’ve added an arrow pointing to the five-year mark when teachers in California qualify for a vested pension from CalSTRS. Note that, while all teachers technically qualify for a pension after five years of service, that benefit is negative for all teachers hired at age 25 or 35. It’s only for teachers hired at age 45 or 55 where vesting produces a positive net benefit. 

    Value of CalSTRS teacher pensions, by age of entry

    To put this in perspective, consider that most teachers begin their careers in their 20s or 30s. According to data from the National Center for Education Statistics, only about 7 percent of California teachers began their teaching career at age 45 or later. Another way to say this is that vesting produces a positive benefit for only 7 percent of California's incoming teachers. Of course if they continue teaching their benefit will continue to grow, but as the graph above makes clear, California teachers may have to stay for 15, 20, or 25 years before realizing a positive net pension. 

    Teachers who leave the system also have the option of withdrawing their own contributions, but even vested teachers in most states do not qualify for employer contributions. That's true in California as well. For many departing teachers, cashing out and withdrawing their own contributions is often a better deal than leaving their contributions with the plan and waiting to collect a pension. In doing so the employee forfeits the guarantee of the pension, but that’s often the right financial decision.

    States know this too. In a piece last year for The 74, I noted that North Carolina dropped its vesting period from 10 years to five after finding the change cost very little. Other states could likely do a similar analysis and find that their vesting periods are largely symbolic. Moreover, when we looked at the data on teacher behavior right around vesting periods, we found that teachers do not stick around longer just to vest in their state's pension plan.

    Given the data I've explained above, teachers are making a rational decision to ignore vesting periods when they're deciding whether to remain in teaching or not, but that means vesting periods are really only accomplishing two things. First, vesting periods do limit the retirement benefits earned by teachers who enter the profession in their 40s and 50s. Second, for younger teachers, vesting periods are forcing risk-averse teachers to choose between withdrawing their contributions or waiting for a "guaranteed" pension, even if that pensions might be worth less than the teacher's own contributions. In both cases, vesting periods help the finances of state pension plans, but they're bad for teachers. 

  • Our news cycle is lightning quick. Proponents and naysayers alike are often loudest just prior to a policy change’s implementation, before moving on to tackle the next one. But what happens after the dust settles? Were the changes as meaningful as supporters hoped? Were ramifications as cutting as detractors warned? We don’t always get the chance to go back and check.

    Alaska’s 2005 teacher pension system reform legislation provides a unique opportunity to do just that. We’ve collected pre and post reform data to examine teacher workforce impact in the wake of retirement plan changes. While it is important to note that these trends should not be interpreted as causal, we feel there are meaningful takeaways all the same.

    But first, some history. In a 2005 special session, the Alaska state legislature passed retirement reform laws, effectively freezing the state’s existing defined benefit pension plan, and enrolling new employees in a 401(k)-styled defined contribution plan. The state faced a $5.7 billion unfunded liability, and while changing the plan structure would not eliminate that debt, it would prevent it from snowballing further. Critics were concerned that the switch would cause teachers to leave the profession in droves, or never enter it at all, citing pensions as a key recruitment and retention incentive.

    So what actually happened? Not much. The table below tracks the average Alaska teacher salary, the number of teachers in the state*, and the percentage of Alaska teachers leaving their district in a given year. The 2005-06 school year is highlighted to denote the policy change. Alaska’s average teacher salaries have risen steadily from 2002 to 2012. The number of teachers has fluctuated over that time, but remained largely consistent.

    And teacher turnover, one of the biggest concerns going into the reform, has remained steady as well. The chart below tracks the percentage of Alaska teachers who left their district each year (including those who leave the profession entirely or moved to a different school), and highlights the new hire cohort first enrolled in the new retirement system. Warnings of dramatic teacher shortages in the wake of retirement plan changes have not come to pass.

     

    This is not to say that Alaska’s reform is a perfect exemplar. The DC plan sets decent contribution rates, with a mandated 8 percent employee contribution and a 7 percent employer contribution, but in passing the 2005 reform legislation, the state missed an opportunity to extend Social Security coverage to its teachers. As a result, Alaska teachers continue to be uncovered, putting their retirement security at risk. The plan isn’t ideal, but it does offer better benefits to a larger group of mobile workers, and it should help keep the state’s finances stable over the long-term.

    Pensions make up an increasing portion of a state’s education spending and any changes to the plans must be balanced to both do right by new hires as well as uphold promises made to existing employees. There are no easy answers, but taking opportunities to examine results in case studies like Alaska allow future policy makers to cut through some of the discourse.

    *Classroom teachers, includes part-time