Friday, March 06, 2009

Kinky Sex Makes The World Go Around

Did I get your attention with that title? Hope so. Today's entry is a short explanation. I hope it helps people understand the relationship between employer contribution rates to PERS and the actuarially assumed interest rate. I've been receiving a lot of questions about this and I feel that a short, simple, explanation is in order. If you've been following my comments over at PERS Oregon Discussion, this shouldn't be new information, but it will be the first time I've put it down in one place.

To think of the employer's contribution rates, we need to think about employee salaries, the Full Formula, the assumed interest rate, and the projected growth of salaries over a career.

Imagine a new employee with a $35,000 annual income. When this person became PERS eligible, he/she and his/her salary became part of the employer contribution matrix of PERS. Imagine that employee having a 30 year career and try to forecast what his/her salary will be 30 years down the road. There is no simple way to do that, but actuaries make certain assumptions about salary growth. Suppose, for the sake of argument that salaries grow (on average) about 4% per year. That means that our $35,000 per year employee in year 1 will be making about $131,000 per year at the end of 30 years of work. Surprisingly, this formula works very well in forecasting final salary. It worked for me, for many of my friends, and for many acquaintances who shared their salaries with me.

With me so far. Our hypothetical employee is projected to be making $131,000 about the time he/she comes up for normal retirement in 30 years. The full formula (for Tier 1 and Tier 2) requires that the final retirement benefit be equal to Final Average Salary * 1.67% per year of service. A 30 year employee making a FAS of $131,000 (or close) will require a pension under the formula of roughly $65,000 annually drawn from two sources - employee contributions and employer contributions. Employee contributions (regardless of who pays them) are statutorily fixed at 6%. The employer is required to make up the difference between what the employee contributes and what is required to fully fund that pension.

Both parties have assumptions (especially Tier 1 members) about the growth of funds. Currently, the assumed rate of growth is 8% for both. In the past, the employers contributed just about the same amount as employees, but as Money Match overtook the formula, the employers were required to come up with increasingly larger amounts of money to "true up" the employee's account at retirement. But, the entire funding depends in large part on the Full Formula. So suppose, both employer contributions and employee contributions are assumed to only grow 6% instead of 8%. Money grows more slowly, the money match ceases to be the dominant method of retirement, and the Full Formula retakes its primacy. Well, if the employee's account balance grows more slowly, and the employer has no guarantee on its earnings, the final "true up" to get to the Full Formula takes more employer cash. The actuaries control for this in employer rates. They try to spread out the costs of that Full Formula retirement over the employee's work lifetime. Thus, the employers contribute significantly more than the employees do to the final retirement. If the average earnings rate declines from 8% to 6%, but the formula doesn't change, the employers have to contribute more than they would have to if the average rate were higher.

Over the long haul, there will come a point where the cost to employers will be lower than it is today once Money Match disappears completely, but the Full Formula is *the* method of retirement for all classes of PERS members. The plain fact is that so long as salary growth continues to average a positive amount (which it will), the employers will be paying more under a lower assumed rate than under a higher one. The higher assumed rate lets the market take care of the employers' excess costs, while the lower assumed rate dumps more costs back on the employers up front.

I hope this helps people better understand the relationship between employer contribution rates and the assumed interest rate. It varies inversely in a pure pension (Formula) system. This is the reason why so many private employers want to move employees from a traditional pension (defined benefit) to a defined contribution (401K-like) plan. The costs are determinate in a DC system, while they vary considerably in a defined benefit system.

Apologies in advance to all actuaries. This is a very simplified example. I know that the determination of employer contribution rates is far more complicated than this, but I think that in a traditional pension system, the basic principles I've explained above hold pretty true.

Enjoy your weekend.


5 comments:

TruthSeeker said...

Wish my salary as a Ph.D School Psychologist had grown at 4% for 30 years! There were quite a few years that our salaries didn't grow at all...instead we were promised a fine retirement package...which PERS, soon after my retirement, started to try to reduce (the 1999 crediting issues for Windows Retirees).

mrfearless47 said...

I'd be surprised if it didn't grow by that much. I'd need to know your beginning salary and your ending salary as well as accumulated sick leave and other add-ons to final average salary. My salary pattern matched yours in that many years there were no increases at all. But I was surprised when I actually ran the numbers how close to annualized 4% increase it turned out to be. Of course, when I started, I was barely making more than the food service workers at PSU. Ending up as an administrator helped, but even without the administrative stipend, my salary grew by 4%.

timo said...

Thanks for your explanation, but can you take it one step further and give the perspective from a PERS manager. PERS has income from a)employer contributions and b) earnings on its investments. It has expenses in the form of paying benefits to retirees. For someone retiring on money-match, I believe their benefit is determined by the size of their account, their life expectancy, and an assumed 8% earnings rate. If the PERS income from item b) above is signficantly less than 8% over a long period of time (which many people are predictiing), it appears that the system is not sustainable. If PERS managers decide that this is the case, can they lower the benefits paid to retirees?

mrfearless47 said...

PERS also has income from individuals, whether paid by individuals or paid by employers. It is still the employees' money no matter who pays it.

PERS has NO legal authority to change benefit structures. Only the Legislature has that power. PERS managers can recommend changes, PERS actuaries can recommend changes, the Governor can recommend changes, and the OIC can recommend changes, but only the Legislature is empowered to Legislate changes. All changes are subject to court review and will be rejected if they retroactively change the "PERS contract" as described in the Hughes ruling of 1991.

The only power PERS has is to lower the assumed rate of return on investments (currently 8%) which is the genesis of this series of posts here and elsewhere. It is the likelihood or possibility of the lowering of assumed rates that will make the system more sustainable in the long run. It will immediately lower benefits for future (not current) retirees as the projected rate of earnings on their account in retirement is lower and so the takeout will be lower. But the consequence of this in the immediate term is to raise employer contribution rates.

The system costs will lower once all the Tier 1 members have retired. There are still about 40% of active members in Tier 1 and the last member joined Tier 1 in 1995. This means that for younger Tier 1 members, their benefit must be projected out to 2025, which is 16 more years out.

Dave1310 said...

I calculated my salary rate of growth for the 35 years I spent with the state and it equaled 7.4% a year.

That makes sense because I usually received the 5% step increase and cola. I was promoted enough to be able to avoid getting no increases at the top of the scale.