The impact of late-term pay raises on teacher pension obligations in three states: California, Illinois, New Jersey.

Author:Roza, Marguerite

    Unfunded pension liabilities for U.S. teachers are massive--nearly $325 billion in 2012--and only one-fifth of states' plans can be considered well-funded (Doherty, Jacobs, and Madden, 2012, p. i). Throughout the country, state leaders are scrambling to find politically and constitutionally acceptable ways to gain control of ballooning teacher pension debt. (1) Yet most leaders are overlooking an important lever squarely under their control: the salary raises offered to late-career teachers.

    Throughout a teacher's career, the district awards him or her year-to-year pay increases. Those raises are frequently awarded as a percentage of base salary, producing higher-dollar raises for higher-salaried, veteran teachers. The pension payout on retirement is determined by the teacher's final average salary. Because of this, teacher pensions are highly sensitive to even modest changes in final salary. This model is different from Social Security or other defined benefit plans that use a career average salary as the basis of the pension.

    This paper analyzes the relationship between late-career raises and pension debt in three states--California, Illinois, and New Jersey--and finds that on average, every $1 awarded to a late-term teacher's final average salary triggers $10 to $16 of new obligations in present-day dollars. When a district gives a $3,000 raise in the final years of teaching, for example, that $3,000 in extra salary drives up the pension debt by $30,000 or more. These late-career raises thus have enormous consequences for a state's overall pension debt.

    State policymakers worried about pension debt ought to be scrutinizing the pay raises awarded to nearly retiring teachers. But some legislators seem unaware of the connection and instead focus on other, often politically sensitive or less effective mechanisms to tackle their pension obligations. (2) Since 2008, 40 states have raised employer contribution rates, and 27 have raised teacher contributions (Doherty et al., 2012, p. ii). Other changes include altering benefit formulas, raising the retirement age, modifying cost of living adjustments (COLAs), and expanding the types of retirement plans offered. These changes have been unpopular, partly because they tap workers to pay more or remove an anticipated benefit. In other cases, the changes don't control liabilities that much, often because they don't affect the pension plans in which veteran teachers are enrolled. (3)

    While state policymakers make these politically tough changes with only a modest impact on liabilities, in many states school districts continue to operate at cross-purposes. Districts award larger raises to their most senior teachers, driving up pension costs, yet many states appear unaware of the effects. While some states have put into place anti-spiking measures to address fraudulent or deliberate manipulations of final salaries in order to impact pension payments, most states have not yet addressed districts' ability to award the largest raises to senior teachers that work to create new and unfunded pension obligations. (4) This paper clarifies the relationship between late-career raises and pension obligations for three states and makes suggestions for how policymakers can work together to better manage pension debt.

  2. DATA

    In general, the pension a teacher receives from a defined benefit plan is the product of final average salary, years of service, a pension multiplier (which may be constant or change based on a retiree's years of service or age), and a cost of living adjustment, as depicted in Figure 1. (5) Years of service is defined relatively consistently across plans and is determined by the number of years an educator has worked (and maintained active pension plan membership through annual contributions), credit earned through extra duties, credit from unused sick leave, and purchased credit. The COLA supplements a pension starting in the second year of retirement at an increasing rate for every year that the retiree receives an allowance. The COLA can vary in percentage terms as well as the financial method by which it is computed (simple interest annually compounded).

    Figure 1. Basic Teacher Pension Formula Annual Final Years of Service Pension = ( Average x x Multiplier) + COLA Allowance Salary To analyze the pension costs in California, Illinois, and New Jersey, we used publicly available data from the 2012-2013 Comprehensive Annual Financial Reports (CAFR) for the three states' teacher pension systems. Some additional information was retrieved from state pension websites and state departments of education when not available in the state's CAFR. For simplicity and accuracy, our analysis focuses on only the costs associated with defined benefit pension allowances.

    Because we focus on veteran teachers nearing retirement, the final average salary, average years of service, average pension multiplier, and average beginning pension statistics were isolated for teachers retiring with 25 years of experience or more (see Table 1). (6) California and Illinois CAFRs publish average beginning pension allowances, final average salary, and retirement by demographic (e.g., age or years of service). These values were used to determine each state's pension formula, which allowed us to observe average pension multipliers.

    Because New Jersey's CAFR does not present these statistics by demographic (but does provide overall averages), further data retrieval and modification were necessary to isolate statistics for veteran teachers in the same way. We obtained 2012-2013 teacher salary data to calculate the average salary for all teachers in New Jersey with 25 years of service or more (New Jersey Department of Education, 2013). Then, to determine the average beginning pension for veteran teachers, we modified New Jersey's published average pension for all teachers to represent only those with 25 years of service or more. This was achieved by first determining the ratio of the average pension differential for all retirees vs. those retiring with 25 years of service or more in Illinois and California; we found that on average, the beginning average pension for a veteran teacher is approximately 141% of that of the average retiree. Finally, we applied this ratio to the published beginning average pension for all retired teachers in New Jersey (as listed in the state's CAFR).


    The pension formula in Figure 1 appears simple. However, given the varying definitions for each factor across state systems and even within systems, it can quickly spiral into a collection of complex calculations and nuanced application of terms. The equation can be complicated by longevity bonuses for veteran teachers, career factors that affect the multiplier, compensation caps for final average salary used in the retirement formula, and pension modifications for early retirement. For example, in California longevity bonuses are permanently added to the monthly retirement benefit at a set dollar amount of $200 to $400 if a teacher earned 30 or more years of service credit on or before December 31, 2010, and a career factor of 0.2% is added to the multiplier. In Illinois, the cap on creditable earnings that can be used to determine the final average salary is $110,631.

    Rather than examine the many factors that can complicate the pension formula, our analysis explores the connection between final average salary and pension allowances by simply analyzing the relationship between these two terms themselves. Shown as a ratio in Table 2, we summarize the relationship between the final salary and the pension payment in order to quantify the impact of final average salary on pension allowances, and subsequently on pension debt. (7)

    California's pension plan is one of the more extreme cases in that it considers only the teacher's salary in the last year as the basis for its computation. Illinois uses the average of the last four years, and New Jersey uses the average of the last three....

To continue reading