Why Your Teacher Pension Estimate Changes From Year to Year

A teacher pension estimate can move by hundreds of dollars between two runs without anyone changing the pension law. The reason is that an estimate is the output of several inputs: service credit, salary history, retirement date, tier, age or benefit factor, and the payment option selected. Change one input and the result can legitimately change.

Retirement date changes more than the calendar

Moving a proposed retirement date can affect at least three items. You may earn additional service, add another salary period to a final-average calculation, and cross an age or eligibility threshold.

CalSTRS makes this visible through its age-factor structure. A member’s age factor can increase with retirement age under the applicable 2% at 60 or 2% at 62 structure. Illinois TRS separates the base formula from early-retirement reductions. Texas TRS applies tier-specific eligibility and reduction rules.

Two estimates for different dates therefore are not apples-to-apples unless you account for the service and age assumptions behind each date.

Salary averaging can move as a new year enters the window

Final-average salary is usually built from a defined period of pensionable compensation, not simply the last paycheck. A new high-salary year may replace an older lower-salary year in the averaging window.

Different systems use different windows. Texas TRS generally uses an average of the five highest annual salaries for the standard calculation, with certain grandfathered rules. Illinois Tier II uses the average of the eight highest salary years within the last ten, subject to plan rules. Other systems use different definitions.

An estimate can therefore rise after a year of work even if the benefit multiplier did not change at all.

Service purchases and refunds alter the service side

A completed redeposit, military-service purchase, out-of-state service purchase or other eligible service transaction can increase credited service. A refund can do the opposite by canceling service.

Do not assume a purchase appears in an estimate the day you send payment. Check whether the retirement system has completed the transaction and posted the service. If you are comparing “before” and “after” estimates, keep a copy of the service-credit total used in each one.

Payment options can lower the monthly amount for a reason

The highest monthly amount shown by an estimator may be a member-only or maximum benefit. Choosing a survivor option can reduce the member’s monthly payment to finance continuing income to another person after death.

That reduction is not an early-retirement penalty and not an error. It is the actuarial effect of a different payment promise. Compare estimates only when the same beneficiary age, survivor percentage and option are selected.

Projection assumptions can create an illusion of precision

Some tools project future salary and service; others calculate only from posted records. One estimate might assume you work continuously until age 65 while another uses service already earned.

Before comparing two results, write down:

• retirement date;
• service credit used;
• final-average or final-compensation amount;
• multiplier or age factor;
• early-retirement reduction, if any;
• survivor/payment option; and
• whether future salary or service was projected.

If the portal does not display those assumptions, request an explanation rather than reverse-engineering a monthly number.

A useful way to test an estimate

Run controlled scenarios, changing one variable at a time. For example, compare June 30 and July 1 only if the system treats those as meaningful dates, then compare member-only and survivor options while keeping the date constant.

This is much more informative than generating ten unrelated estimates. It shows which rule is driving the change and can reveal a data problem. If adding a year of service unexpectedly lowers the base benefit, check whether a different salary window, age factor or reduction was also triggered.

When a changing estimate is actually a warning

Investigate if the underlying service, salary or tier changes without a reason. A previously credited service year disappearing, a membership date shifting, or pensionable salary dropping despite identical payroll records warrants follow-up.

Save dated PDF copies or screenshots of official estimates near retirement. They are not contracts, but they document the inputs and can help the system explain a later difference.

An estimate is best treated as a transparent model of the plan rules. The goal is not to find one reassuring number; it is to understand which verified facts produce that number and which assumptions will change before the retirement application is finalized.

Save the estimate inputs as carefully as the result

Near retirement, a screenshot of “$3,412/month” is not enough. Export or write down the exact retirement date, beneficiary option, service total, salary average and factor used in that run. When the next estimate becomes $3,365, those inputs let you see whether the system changed a record or you changed an assumption. This simple habit turns an estimate history into an audit trail and prevents a teacher from arguing over two numbers that were never based on the same scenario.