Seeking a Fix for 'Spiking'

Abuses of overtime in pension calculations
April 23, 2009
Girard Miller
By Girard Miller  |  columnist
Girard Miller is the Public Money columnist for GOVERNING and a senior strategist at the PFM Group.

As I note in my companion column on pension system reforms, one of the most flagrant recurring abuses of the defined benefit system is the "spiking" of pensions by older workers who pile on overtime as they approach retirement. Union seniority systems allow this--sometimes even formally as in Buffalo, New York, where they have embedded an actual formula into some union contracts. In other cases, the practice just somehow magically happens through informal arrangements between fellow union members.

Public officials and the media are beginning to wake up to this issue and demand change. In New York City, the fire department's senior management has actually created an overtime-control officer and has placed a cap on overtime in order to control pension costs. This could be the beginning of a sea-change.

Here's the problem: If every worker earned the same amount of overtime every year, this would not become an issue. The pension system would collect contributions evenly through their careers, and the plan would invest that money to provide the funds necessary for payment of a pension based on average final compensation. There would be no pilfering of public money. But that isn't how it works in the real world. In the real world, employees game the system, and find ways to load up on overtime in the final years before they retire, which is the period in which the pension formula calculates their benefit. As a result, there is insufficient savings during their working careers to pay for this benefit. So guess who now picks up the bill? You guessed it: the taxpayers, in the form of a systematic unfunded liability that is created by these pension-spikers.

Here are five solutions to the spiking problem:

o Establish firm management control on overtime and require even distribution throughout the workforce, to eliminate the seniority game and the spiking issue at its roots, as New York is attempting to do. This requires the complete elimination of seniority-based assignments of overtime, whether informal or formal as in the Buffalo scheme.

o Eliminate overtime from the pension calculation entirely and compensate overtime earnings for retirement through a sidecar deferred compensation plan, as described in one of my previous columns.

o Lengthen the calculation period for the final average compensation (FAC). Instead of using two years, a proposal in Allegheny County, Pennsylvania (see below), is to use four. Some plans use five years. This reduces the opportunity for abuses, by averaging out the peak years of overtime. It also reduces the pension benefits for all employees during periods of high wage inflation. This hurts the non-overtime employees and helps the taxpayers generally because overall pension benefits and liabilities will be reduced by a percent or so for each year the calculation period is increased. Some pension systems will be forced to move to longer FAC periods just to make their books balance, so I expect to see more of this.

o Conduct a "spiking audit" and make employees contribute more annually from their payroll in order to bankroll the overtime abuse. The actuary can actually calculate the cost of this gamesmanship and determine how much extra all employees would need to pay in order to sustain the ongoing practice of spiking. If unions want to continue this practice, then they should fund it actuarially through their members' contributions.

o Charge employees a higher contribution rate on overtime. That will help fund a higher pension for those workers who expect to receive it. The actuary can easily determine how much each employee would need to contribute in order to continue to receive pension payments based on their overtime earnings. I'll bet that for public safety, the actual number works out to be something like 10 to 20 percent of their O-T, if employees pay half the true cost.

Recently, the Allegheny County, Pennsylvania, retirement board went to the state legislature to try to fix this problem. The board suggested the first two of my remedies: a longer final average compensation period and a "sidecar 457 plan." We'll see if they make any headway. It's unfortunate that they have to take matters like this to the state legislature, which is one of the dysfunctional aspects of public pension plans: They are governed by state legislatures that have little if any responsibility for local taxpayer interests.

For other important, and needed, pension reforms, see today's companion column.