In Search of a Fair Pension Formula
Realistic income-replacement ratios in the 'new normal' economy
My previous column on the growing generation gap in pension benefits stirred up a hornets' nest. Readers chimed in on all sides of the debate and started spewing out formulas and "multipliers" to express what they consider a reasonable pension benefit. So, let's take a look at what's happened with pension benefits and where we go from here.
A multiplier is the pension formula factor that determines the size of the lifetime annuity a pensioner receives, usually expressed as a percentage of final compensation times years of service. For example, a building inspector with 25 years of service in a plan with a 2 percent multiplier would collect a life pension of 50 percent times her final average salary.
From bare bones to multiplier madness. Most public pension plans started out with very frugal multipliers. Historically, if you go back to the early formulas of the 1950s and 1960s, many plans used multipliers of 1 percent or a little more. Pensions for retirees of that era were very spartan because their postwar public-service career was abbreviated, public salaries back then were very meager, and their multipliers were just too low. Reforms were needed and gradually made. When today's Baby Boomers joined the work force 25 to 30 years ago, pension multipliers of 1.5 percent were more common.
As stock markets took off in the 1980s and unions became more powerful in the public sector, "pension creep" took hold and multipliers moved up -- even though life expectancy had increased which kicked pension costs even higher actuarially. As the CalPERS chief actuary has explained, public employee salaries rose significantly during this period, coming closer to parity with private pay but further increasing the actuarial liabilities. These upward trends culminated in the magical-thinking years of 1999-2000 when legislatures and public pension boards instituted irrationally exuberant multipliers. Worse, they awarded them retroactively with the expectation that stock prices would forever grow and pay for both future obligations and past-service credits that were never actually funded with anything other than paper profits that quickly vaporized.
This gave many Baby Boomer employees a constitutionally protected, irrevocable benefits formula that is now woefully underfunded and simply can't be supported by the tax and revenue base of most public employers in the new American economy. To make matters worse, many public employers have topped it off with a free retiree medical benefit that is totally unfunded. Nationally, the total retirement-plan deficit is now $2 trillion and growing.
The ratchet effect. Now that reality has set in, public managers are beginning the painful work of reviewing their retirement-benefits formulas. In some states, like California, their hands are tied with respect to incumbent employees. Perversely, some state laws preclude plan-design changes for incumbents who are viewed as having vested rights to the highest multiplier awarded during their entire career. The great economist John Maynard Keynes presciently called this the "ratchet effect:" What goes up can never come down.
As a result, most public managers must look first at new "tiers" for pension benefits. This means that new hires will get a lower benefit with a lower formula and often a higher retirement age than the senior incumbents -- who happen to sit on the negotiating committees. That's what gives rise to the friction between generations of employees that I cited in my previous column.
A sustainable number. So, if we are going to design a new system that can be (1) sustainable for new employees, (2) affordable for both employers and employees, and (3) sufficient to provide a reasonable retirement benefit for the next generation, what are the right numbers?
Interestingly, nobody has been bold enough to put objectively based numbers in print and to explain the rationale for benefits reform. I'll take a crack at it. Having examined a host of long-term state and municipal financial projections to see what's actually affordable, I've performed detailed sustainability analyses. My article on sustainable retirement-plan financial management in the August 2009 Government Finance Review provides an overview of the analytics required. I will e-mail a copy to public managers who request a copy.
The new normal. My formulations take into account the states' and localities' crippled financial condition and hobbled long-term financial outlook, as we slowly slog our way out of the economic malaise of the Great Recession and evolve into what is now called the New Normal economy. This new realism reflects the long-term hangover from the bubble years of 2000 and 2007, when stocks doubled every three years. It takes into account the long-term impact of the housing washout on property taxes; the shift toward more savings and less home-equity borrowing, which stifle consumption-based sales taxes; and the decade of capital loss carry-forwards that will curb income tax receipts. It also considers the overhanging liabilities for vested pensions and retiree medical benefits that will deplete the pension funds and consume the lion's share of any new revenues coming from economic growth -- as well as deferred capital and equipment spending that cannot be postponed indefinitely. In states such as California where constitutional tax caps rule out the potential for higher revenues, the sustainability problem has reached crisis proportions.
The table below summarizes my findings. They are from a white paper I've written for California municipal officials on the topic of sustainable and sufficient pension benefits.
This table will stir up another hornets' nest, I'm sure. Taxpayer advocates who place little value on public service will think it's too generous. Unions will say "No Way, Jose" to the roll-back of multipliers. The "20 and Out" crowd will cling to the myth that first responders born after 1985 cannot possibly be expected to serve the public productively until they reach age 57 -- despite life expectancies nearing 90 for those who have already survived childhood mortality. Those who don't understand the math will revert to slogans and ideology. Some may be confused about my new concept of a reduced pension payout formula for first responders once their Social Security benefits kick in.
Several of these topics will occupy entire columns in the future. For now, let me introduce the basic logic behind each feature of a New Normal pension paradigm.
New normal retirement age: Civilian public employees hired tomorrow will have to work until they reach today's normal Social Security retirement ages. Public safety workers can be compensated with a sustainable pension plan at age 57 after working for 25 years. Those who want an earlier retirement can take off five years earlier, but they must accept a lifetime actuarial reduction of about 30 percent, judging from the factor that Social Security already applies to early retirees of that generation. Age 57 allows ten years to acquire Social Security credits, as explained below, and ensures that the pension-able lifespan is not too much longer than the contribution period. Otherwise, the math is simply impossible as we have all learned from current experience. Those who retire earlier can accumulate even greater retirement benefits elsewhere, to complement their public pension.
I would not oppose earlier retirement ages if we could make the math work. But it's just not sustainable as a matter of public policy. Short working careers and long retirement annuity periods are a mathematical inequality. It's really that simple. As longevity increases, the problem worsens.
New normal career span. Today's workers entering the workforce after college at age 22 will live 45 years until they are eligible for Social Security. Compared with private-sector benefits in the labor market today, it's not unreasonable to expect them to work 30 years (2/3 of their potential working lifetime) to receive a lifetime old-age pension at taxpayer expense. Parents who take some time off to raise a family have plenty of time to do so and still earn an old-age pension. For first responders, it's reasonable to expect a 25-year commitment and covenant. "20 and out" is out. Those who work elsewhere during part of their lives can no longer expect taxpayers to pay their entire retirement benefit for periods of time that exceed their public-service period. That math just doesn't work.
The new normal public safety pension paradigm. Public safety employees cannot expect to work 20 years and then go fishing and hunting the rest of their lives. Having worked in local government, I know that safety professionals may not be able to carry unconscious fire victims down ladders from flaming apartments or to wrestle armed thugs to the ground to handcuff them at the age of 60. But everybody who applies for work in the public safety professions in their twenties should be prepared to expect that in later life, they will pursue other interests that produce at least 38 percent (3/8) of their professional salaries after they check out from the police station and the firehouse.
Even those working for employers exempt from Social Security can be reasonably expected to qualify for Social Security benefits (albeit reduced by federal offset formulas) between ages 57 and 67. That's their responsibility, not that of state and local government taxpayers. Remember that there are 45 years between ages 22 and 67, which leaves 20 years to do something else before, during and after working a distinguished 25-year career as a first responder. Qualifying for Social Security requires 40 working quarters, by the way -- one-half of the available "outside" working years for a 25-year public safety employee.
This means that when public safety workers retire from that job at age 57, a reasonable pension plan would provide them 62 percent of their former income until they become eligible for Social Security (or a second pension or retirement benefit) to complement the income from personal savings in 457 accounts and from their personal IRAs. Thus, the formulas above provide a more generous benefit from 57 to 67, and then the same multipliers as civilians receive once they reach Social Security age.
Of course, most public safety workers accumulate Social Security benefits long before they qualify for retirement from their governmental employers. There is a lot of moonlighting, side-work and second-careering going on out there -- as revealed in any confidential public-policy study of state income tax records, Social Security earnings statements and unreported cash income of employees with those job titles. If you think of all the military retirees of similar ages who are engaged in other, productive work, you get the picture.
Retirement income replacement ratio. Employees don't need 100 percent of their working income to sustain their lifestyle in retirement. After retirement, they don't pay into the retirement plan, their savings have already been accumulated, and they don't pay into Social Security any longer. Financial planners also take into account lower expenses for commuting, paying off the mortgage and other factors. Policymakers can safely begin with the premise that taxpayers have no business providing replacement income which exceeds 85 percent of pre-retirement income (when combined with Social Security and the income stream from personal savings.) Social Security provides replacement income of approximately 20 to 25 percent for most state and local government workers, and the income stream from a 5 percent annual contribution to a 457 plan works out to be 15 percent of final income when distributions begin at age 67. For civilians in Social Security, that requires about 50 percent to come from pensions.
Some professional analysts can quibble and split hairs over these numbers, but most respectable financial planners will agree that I'm in the right ballpark on the income replacement numbers and multipliers they imply. Nobody can defend a multiplier of "3 percent for Life," that's for sure.
Next month I'll explain the assumptions and conditions that must be fulfilled for these estimates to be valid, which include a post-Medicare retiree medical benefit with employee premiums similar to active employees, some reasonable inflation protection, and a labor-market-competitive salary during the working career.
Employee contributions. The bottom row of the table above shows what new employees will need to pay into these New Normal pension plans. As I've suggested in a prior column, the ideal ratio of employee-employer cost sharing should be closer to 50-50 than most of today's public retirement systems. Of course, the actual contribution proportions should also reflect local labor market conditions and total compensation levels, financial sustainability of the employer's revenue base, and other factors. For civilians, a 5 percent employee contribution is right in line with current practices in most states. In systems outside of Social Security, the rates are higher because the employees and their employers don't each pay 6 percent into the federal retirement system while working in public service.
There is no reason that incumbent employees in retirement systems with multipliers more generous than those in the table above should not be immediately required to pay more into their systems. If their benefits are grandfathered and guaranteed, then at least they should pay a fair share toward their over-sized rewards on a going-forward basis. Even that won't touch the massive unfunded liabilities their generation has racked up -- much of it by granting themselves benefits retroactively during their careers. And when state laws and court decisions allow, there is no reason that the multipliers above could not be applied prospectively for incumbents' services provided in the future, preserving the benefits earned already but reforming the pension system faster.
Finally, for those contributing to Social Security, any increase in employee contribution rates above 5 percent of salary should be phased in over time. New hires or those paying zero now might pay 3 percent initially, 5 percent next year, and so forth. For public safety workers whose 50-50 share of the normal cost could be 10 percent or more in addition to their Social Security taxes, it may take four to five years to reach the full employee share. Of course, that delay will increase the actuarial normal cost, which should then be shared by the employer and employees. If those numbers place too much burden on the uniformed employees' take-home pay, a different cost-sharing ratio may be appropriate, such as 60-40 employer/employee, as long as it discourages demand for "free" benefits, and employees all understand that this is a partnership.
The clay is still wet. I am not suggesting that these benchmark plan features should be set in stone this month, nor that one size fits all. Every employer must re-evaluate its HR objectives and decide what's a fair bargain for employees and taxpayers. My research is still in progress. Upon receipt of collegial input, I expect to publish a second, refined version by year-end. For employers that can still afford to be more generous than the formulas suggested here: Be my guest if you have supportive taxpayers. In some cases, a modest, supplemental defined contribution plan with a 457-matching feature should be paired with this baseline defined benefit plan to encourage and reward individual saving.
Just remember that the total package including retiree medical benefits must be sustainable, affordable and sufficient. The structure I've outlined above will bring us far closer to that objective than the public retirement-plan designs that prevail today in many states. Sadly, I know that these realistic numbers will make me about as popular as a rattlesnake in a locker room -- on both sides of the political spectrum. But the math is inescapable and the sooner we face facts, the fewer the casualties to be suffered in this slow-motion train wreck.