To calculate the contribution rates necessary to pre-fund the plan’s benefits, an actuary uses an actuarial cost method, asset valuation method, economic assumptions, and demographic assumptions. This section describes the actuarial assumptions we use in our valuation process. Please see Actuarial Methods for descriptions of our actuarial cost and asset valuation methods. Please see the Actuarial Methods and Assumptions section of the latest Actuarial Valuation Report for a list of frequently changing or new actuarial methods and assumptions.

Economic Assumptions

In the valuation process, assumptions are required for four economic variables:

  • Expected investment rate of return.
  • Inflation.
  • General Salary growth.
  • Membership growth.

Economic assumptions affect expectations regarding the accumulation of assets and the growth of projected pension benefits.

The Pension Funding Council (PFC) adopts economic assumptions for all plans/systems except LEOFF 2. The LEOFF 2 Board adopts economic assumptions for LEOFF 2. All economic assumptions are then subject to revision by the Legislature. The PFC and LEOFF 2 Board adopted lower economic assumptions in 2017. For all plans except LEOFF 2, the investment rate of return assumption decreased from 7.70 percent to 7.50 percent. For LEOFF 2, the investment rate of return assumption decreased from 7.50 percent to 7.40 percent. These lower economic assumptions are reflected in the 2017 AVR which is used to set rates for the 2019-21 biennium.

Table 2017 Assumed Investment Rate of Return Schedule

The economic assumptions used in the Actuarial Valuation Report are shown in the following table.

Table Economic Assumptions

Demographic Assumptions

These include rates of retirement, rates at which members become disabled, turnover rates, mortality rates, and several other demographic assumptions as disclosed later on this page.

Step Salary Increases

Probability of Service Retirement

Select a table from the drop-down below:


Our mortality rates include an assumption for future mortality improvements. We took two steps to build our mortality assumptions.

First, we developed the base mortality table by starting with RP-2000, published by the Society of Actuaries, and applying age offsets for each system and plan. When age offsets are negative, it means we think people of a given age in the retirement plan are generally healthier than others their age from the general population. In other words, we expect their mortality experience will be similar to younger people. Conversely, a positive age offset means we expect mortality experience for a given age to match that of a higher age in the general population. For instance, we expect a 50-year-old PERS male to have the same mortality rate as 49-year-old males in the general population because we assume a negative one-year age offset.

Next, we applied mortality improvements to the RP-2000 mortality table using Scale BB, also published by the Society of Actuaries. We use “generational” mortality instead of projecting to a given year. Under generational mortality, a member is assumed to receive additional mortality improvements in each future year, throughout their lifetime.

As an example of generational mortality, consider a healthy PERS Plan 1 male, age 50. To project the RP-2000 mortality rates to a given year (2013 in this example), we use the following equation.

RP-2000 rate x (1 –Scale BB)^13

For a 50-year-old male, this is 0.001995 x (1 –0.003)^13 = 0.001919.

The next tables show the age offsets we used as well as the mortality rates projected to the year 2013 for each plan. Please note that this table is meant to be an example only. Under generational mortality, the mortality rate for each age will improve in each future year by the rates in the mortality improvement table.

Following these tables, the next table shows Scale BB. Please see the 2007-2012 Experience Study Report for more details regarding the development of these rates.

Select a table from the drop-down below:

Probability of Disablement

Select a table from the drop-down below:

Probability of Termination

Other Demographic Assumptions

Select a table from the drop-down below:

Medical Premium Reimbursement

  • RCW 41.26.510(5) and RCW 43.43.285(2)(b) state that qualified survivors and children of line-of-duty deaths (Survivors) in LEOFF 2 and WSPRS shall have medical premiums reimbursed from the retirement fund. The laws also provide that all survivors will be covered by the Public Employees Benefits Board (PEBB).
  • RCW 41.26.470(10) states that LEOFF 2 members with total disabilities (Disabilities) and qualified family members shall have medical premiums reimbursed from the retirement fund.

The costs for these benefits are included in the results presented in this report. However, the benefits are funded through irrevocable trust funds, known as 401(h) accounts, from contribution rates selected by the Department of Retirement Systems (DRS) and the LEOFF 2 Board. These contribution rates are “carved out” of the total adopted contribution rates. DRS and the LEOFF 2 Board will periodically review the funding requirements for these benefits and adjust the 401(h) contribution rates as necessary.

The information in the next section represents methods and assumptions tied directly to the medical premium reimbursement benefits. Please see the 2013 Other Post-Employment Benefits Actuarial Valuation Report (OPEB Report) for the following referenced assumptions.

Medical Inflation

Current and Future Survivors, and Future Disabilities: Uniform Medical Plan (UMP) Medicare and Pre-Medicare assumptions without excise tax (OPEB report). Current Disabilities: 5 percent per year.

Percent Married

Future Disabilities: 85 percent.
Current Disabilities: 100 percent.

Percent with Children

Disabilities and Survivors: 100 percent, one child each.

Premium Percentages

When the data for members currently qualifying for total disability benefits does not provide information about how many family members are covered, we use the assumptions below to split the total premium into each family member’s share.

Assumed Coverage Type, Future Disabilities

Fifty percent covered by policies provided under the Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA). Fifty percent covered by employer-provided policies.

Assumed Timing/Length of Coverage

Assumed Premiums

Miscellaneous Assumptions/Methods

We include the following miscellaneous assumptions and methods in this valuation.

  • Minimum and maximum allowable ages are set in the data as follows.
  • Default entry salaries, usually increased for past service, are assigned for active members with less than two months’ service during the valuation year.
  • Historical salaries for vested terminated members are not provided in the valuation data. Beginning with the 2008 valuation year, we first look to see if we kept a historical salary for such a member in the prior year’s data. If so, we copy the salary to the current year’s data. If a member was active in the prior year and terminated in the current year, we copy the prior year’s salary to the current year’s salary and keep it as historical.

    Additionally, in 2009 we searched our data for actual salaries up to ten years prior for terminated vested members who did not already have historical salaries listed. To estimate salaries for the remaining terminated vested members, we use the following procedure. First, a salary appropriate for the given system/plan and the member’s total past service is assigned. These salaries are determined as of a given base year. Second, the salary is divided by the general salary increase assumption for each year the member has been inactive as measured from the base year. Please see the 2007-2012 Experience Study Report for a record of the salaries used.
  • While DRS reports salaries earned during the year prior to the valuation date, the salaries used in the first year of the valuation process have received an additional merit salary increase. In other words, the valuation software projects salaries to the coming year, beginning the day after the valuation date.
  • We assume the gender of beneficiaries for active members will be of the opposite gender as the member. Upon retirement, members typically report the actual gender of their declared beneficiary, and we then use this gender.
  • We assume that the mortality rates of beneficiaries will follow the healthy mortality table for the same system which the member belonged. For example, if the member belongs to SERS, his/her beneficiary will use the SERS healthy mortality rates.
  • All systems use a midyear decrement timing assumption.
  • Members who receive a disability benefit are not assumed to return to active duty in the future.
  • Disability and termination rates are discontinued after members are eligible to retire (with the exception of LEOFF and WSPRS disability rates).
  • LEOFF Plan 1 and WSPRS Plan 1 use a slightly increased ratio of survivors selecting annuities over that displayed in the table presented earlier in this section. This assumption estimates the likelihood a qualified survivor will receive an ongoing benefit at the time of the member’s death. The valuation software used also applies mortality assumptions to potential survivors, which results in an understatement of that future liability for those plans. To recognize this liability, we assume a 67 percent survivor annuity.
  • We assume all survivors of currently retired WSPRS Plan 1 members who opted for the standard survivor benefit at retirement will receive an initial survivor benefit of 50 percent of the member’s average final salary.
  • We value the Basic Minimum COLA for PERS 1 and TRS 1 outside of our valuation software for members currently not receiving the benefit.

Last Reviewed: 08/27/2018

Last Updated: 08/27/2018