Working overtime to bankrupt PSPRS

The latest piece on the upcoming PSPRS reform plan/referendum, "Voters to decide on pension plan," by Pete Aleshire appeared in the December 31, 2015 Payson Roundup.  It does not give much new information, but it does include this:

The details of the proposed reform measure have not been released. However, they reportedly include limits on “spiking” salary in the final three years by counting unused sick leave and vacation time as well as piling on overtime.

I found this interesting as this is the first mention I have seen that pension reform will deal with overtime, specifically with how the accumulation of overtime pay in an employee's high years' average (a high-three or high-five years' average, depending on when the employee was hired) artificially inflates the employee's pension benefit.  In the past, discussions about pension spiking usually focused on the use of sick and vacation leave sellbacks to raise an employee's high years' average in order to spike the employee's pension benefit.  While this was illegal, some employers allowed it until lawsuits forced them to adhere to state law, though the City of Tempe and its firefighters are determined to continue the practice via sleight of hand.  However, the disproportionate loading of overtime into the high years' average is an equally bad or even worse form of pension spiking.

If there is a provision to eliminate or limit overtime pay amounts in employees' high years' averages in the PSPRS reform plan, it may be the result of a recommendation made in the Arizona Auditor General's "Performance Audit and Sunset Review of the Public Safety Personnel Retirement System." ("the Performance Audit")  The Performance Audit listed among its "additional actions necessary to improve system plans’ financial condition and long-term sustainability" was to ". . . develop materials for PSPRS plan employers on overtime pay and implement formal policies and procedures to ensure benefit calculations are correct."  The Auditor General writes:

. . .  overtime pay is not included in the definition of compensation for CORP or EORP. Including overtime pay in the benefit calculation may increase the risk that the member’s final average salary is higher than would be expected from normal salary increases, which may generate unfunded liabilities for employers. If a PSPRS plan member worked a large number of overtime hours during the period that determined that member’s final average salary, these hours would increase the member’s reported compensation and resulting pension benefit.  Although the PSPRS plan would collect more contributions from the member’s increased compensation, this increase could generate unfunded liabilities for the member’s employer.  Specifically, one of the assumptions that the System’s actuary uses to estimate pension obligations forecasts the levels of compensation upon which members’ pension benefits will be based. According to the System, overtime pay is reflected in this assumption; however, there is a risk that the overtime pay in the period used to determine a members’ final average salary may surpass this assumption and generate unfunded liabilities for their employers.  Further, according to the National Institute on Retirement Security, even though pension spiking is not common, a few isolated instances can create the impression of widespread abuse.

The December 2012 Final Report of the Defined Contribution and Retirement Study Committee gave us some numbers when it comes to pension spiking in PSPRS.  It defined spiking as a:

. . . compensation increase by more than 25 percent during the last three years of employment. This is more than twice the average compensation increase that the plans use in determining wage inflation in the final years of employment.

The Committee found that between 2008 and 2011 the percentage of retirees who increased their compensation by more than 25% in their last 36 months ranged from 22.6% to 29.77%.  This shows that it is not an "isolated" practice among only a few PSPRS members, but rather done by one in four PSPRS members.

If a PSPRS member earning $60,000 per year spiked his salary by just 10% over his last three years, he would increase his annual pension benefit by $3,750 per year (62.5% of $6,000) if he retired after 25 years of service.  He would pay an extra $2,097 in member contributions (11.65% of $18,000).  If his employer's contribution rate is 50%, the employer would pay $9,000.  The PSPRS member is obviously happy since he makes up his extra contributions on the 10% spike in the first seven months of his retirement.

In order to figure the financial damage to PSPRS, we need calculations that get more complicated.  Using Bankrate.com's compound savings calculator, we can see that the combined $11,097 in employee and employer contributions will actually be $12,403 when that employee finally retires, if the contributions are compounded monthly at PSPRS' current expected rate of return (ERR) of 7.50%.  That $12,403 will now have to fund the extra $3,750 per year that the now-retired PSPRS member will get for the rest of his life.  Now we see the problem that comes with even a small amount pension spiking.

If we use an annuity calculator and the same ERR, the amount that needs to be there when the employee retires is actually $42,552, if he lives for another 25 years.  With only $12,403 the retiree should only expect to receive his extra spiked benefit for 3.79 years before that money ran out. This assumes many other things such as no COLA's, decreases in the ERR, a lower employer contribution rate, or a spouse that outlives the retiree.  The deficit on the spiked benefit amount must be made up by either taxpayers or active PSPRS members.

This is how we slowly kill our own pension system, and it is the problem the Auditor General identifies with overtime pay.  It simply makes no financial sense for PSPRS to include overtime pay in the high years' average used to calculate pension benefits, and the other two pension systems PSPRS manages, EORP and CORP, wisely forbid this practice.  Spiking pension benefits with overtime pay is another of the perverse incentives embedded in PSPRS, like the Deferred Retirement Option Plan (DROP), that leads employees to collectively chip away at the financial solvency of their own retirement system.

There are several underlying principles here that are very important for us to remember.  The first is that a pension system relies on contributions to compound over time.  The funds paid early in an employee's career are much more valuable since they should (hopefully) grow over the decades to an amount sufficient to cover the lifetime pension benefits earned.  Overtime spiking throws this formula out of whack by dramatically raising the pension benefit amount during a brief period at the end of the employee's career.  Not allowing those end-of-career contributions to compound over the years leaves the pension with a deficit that needs to made up somewhere else.

The second principle is that employees should not be able to influence their own pension benefit calculations.  Economics tells us that individuals will naturally attempt to maximize their own gain, and as employees, this is not always done in a manner that is mutually beneficial to the organization that employs them.  Every employee who spikes his pension does so at the expense of the organization, taxpayers, and fellow and future employees.  This is not a criticism of any individual taking advantage of a flawed system but simply a hard financial truth.  Every unfunded dollar of pension liability caused by spiking will be a dollar not spent of equipment, facilities, raises, or other services provided by the government.  If overtime were not pensionable, an employer who has a contribution rate of 50% would immediately save 50 cents on every dollar paid in overtime, and the employee would also save the 11.65% contribution paid on his overtime pay, nevermind the long-term savings to the pension. 

The last principle is that public safety unions have failed their members by not addressing this issue.  Like all unions,public safety unions expect loyalty and discipline from their members in order to enhance the collective good.  This can and should work, but as I have mentioned before, unions tend to work not for the collective good but rather as a seniority protection agency.  Pension spiking is a perfect example of this.  It benefits only one group-- those close to retirement, and as mentioned before, it actually harms members collectively.  The justification in the past has always been that someday the junior person will be senior and will be taking advantage of the same benefit.  However, we now have what is essentially a zero-sum game where any benefit taken by one group comes out of someone else's end.  Union leadership, usually made up of senior personnel, has no incentive to change this as it will negatively affect their own retirements, and so they pretend like a problem does not exist until they are forced to deal with it by passing the costs on to someone else.  This it at the heart of the Professional Fire Fighters of Arizona's (PFFA's) pension reform plan, which will transfer costs to another generation of public safety employees (and taxpayers) in order to protect the benefits of those close to retirement.

An organization that was dedicated to the collective good would work to prevent pension spiking, and instead, work at raising pay for everyone.  This would benefit everyone, both in the form of higher current wages and in the future with higher pensions, and it would not be done by shortchanging PSPRS.  Wage increases across the board would boost pay for all ranks and seniority, whether newly hired or getting ready to retire, and increase the amount of funds going into PSPRS.  This would allow for the normal compounding of contributions so necessary to keep PSPRS from slipping further into debt.  As for those who still want to raise their pensions, they will have to do it the old-fashioned way via promotion or other merit-based processes.  Eliminating overtime pension spiking will help to align the interests of all employees, unions, and employers, instead making them work against one other.


I believe that two of the best potential reforms to PSPRS would be the elimination of the DROP and ending all forms pension spiking, whether overtime pay, sick and vacation leave sellback, or anything else that artificially inflates an employee's high years' average.  I am optimistic for the implementation of the former reform though not the latter, but we will have to see what comes out of the Legislature.  Hopefully, we will be seeing something real soon.