The military’s retirement system is a mess. But the current proposals to fix it have a hidden agenda. No, I’m not talking about cutting benefits to save money. That’s the stated agenda, which is sure to get attention in this cash-strapped era.

But cutting a benefit paid out over decades, throughout a beneficiary’s lifetime, won’t actually save much money in the next five years or even the next 10. Certainly, military pensions are generous: service members become eligible after 20 years of service, at which point they can retire immediately and receive half-pay for the rest of their lives.

Certainly, that generosity is expensive, costing $24 billion of the Defense Department budget in fiscal year 2011 alone, $46 billion if you add in costs to the Treasury. On paper, a retirement cut would start saving the Pentagon money immediately, since the defense budget sets aside money today to cover the projected costs of retirement in the future. But this “accrual fund” is just a transfer from Defense to the Treasury.

Since no reform proposal would cut benefits being currently paid to troops who’ve already retired, the federal government as a whole wouldn’t save money for years. The real short-term effect would be a sleight-of-hand to move costs from the Pentagon to the Treasury. So if cutting retirement wouldn’t directly help the government through the current budget crunch, what would it do?

Making downsizing easier

The real near-term impact of the proposed reforms would be to make the coming drawdown cheaper and easier to manage. Budget cuts mean all the services will have to shrink, because personnel are expensive. And even under the current retirement system, it still saves the services money to turn an active-duty servicemember drawing full pay into a retiree drawing half-pay. In fact, since retirees don’t get allowances for housing and so on, the cost drops by more than 50 percent.

The problem in past drawdowns has been getting the most senior, most expensive people to leave. The current benefit is not only generous, it’s also all-or-nothing: Leave the service after 19 years and 364 days, you get nothing; stay one more day and you get half-pay for life. So senior service members hang on bitterly until they hit 20. Among servicemembers with 15 to 19 years of service, just seven percent leave the military; among those with 20 to 25, it’s seventy-six percent. During the downsizing of the 1990s, the only way to get that 15- to 19-year cohort to quit was to offer buy-outs.

That’s an expense budgeteers don’t want to repeat. Make retirement at 20 less generous, and there’ll be less incentive to hang around, so you’ll need to pay people less to leave earlier. This objective is buried in PowerPoint slides and hundred-page-plus reports, but it’s there. (Most explicitly on slide 7 of the Defense Business Board’s proposal from this summer).

And the Pentagon’s new power to get rid of people wouldn’t stop there. Today’s lifetime retirement benefits are set by law and can only be changed by Congress, which historically has shot down every reform. (One partial exception was “Redux,” a cut enacted in 1986 but abolished before it ever took effect). The reform proposals would not only cut the retirement benefit but convert much of what was left into payments set at the services’ discretion.

Under the current system, pay-outs start immediately on retirement, typically around age forty. The reform plans would push that back to age 57 or older, excising twenty-plus years of payments. In compensation, the reformers offer an array of benefits that includes a lump-sum payment on leaving the military (a “transition payment” or “separation pay”), bonus payments for staying through various milestones (“gate pay”), and government contributions to servicemembers’ “Thrift Savings Plans” accounts, a kind of military 401(k). But exactly what these payments should be is left up to the Defense Department and the services – who will be able to adjust them without going through Congress.

Managers could dial retirement benefits up and down at will to encourage people to stay or go, depending on current personnel needs. “Flexibility” for the bureaucrats means uncertainty for the troops. The result would be low morale, less trust in authority, less control over one’s financial future, and more of our best troops quitting the service.

Getting reform right

So how should we fix military retirement because there is no doubt that it does need to be fixed. The current all-or-nothing system means that 83 percent of service members don’t get a dime of retirement benefits. For anyone with less than 10 years in uniform — which means everyone who joined after 9/11 — the prospect of retirement at 20 years is too distant to be much of an incentive to stay in. Conversely, those who do make it to 20 years – less than one in five – have a big financial incentive to leave at once. That’s not just inequitable, it’s counterproductive.

This two-tiered all-or-nothing system made some sense during the draft era, when the military had masses of short-term conscripts and a handful of long-serving professionals, who were physically worn out by age 40. Today, people stay healthy longer, troops are volunteers, and both advanced technology and complex conflicts put a premium on experience over raw youthful vigor. But today’s retirement benefits do little to retain the most crucial cadre in the modern military: the battle hardened, highly experienced veterans of multiple tours in Afghanistan and Iraq.

Now sergeants and mid-grade officers, these vets are the backbone of our armed forces. Having reached eight to 12 years in uniform, they are now considering whether to make the military a career. The last decade has been hard on them and on their families. But if they became eligible for at least some retirement benefits at, say, ten to fifteen years of service instead of twenty, that would be a near-term incentive to stay. A reformed system needs to strike a difficult balance: generous enough to be worth hanging around for, not so generous that it encourages leaving as soon as you’re eligible.

Any proposal would have to be rigorously modeled and pilot-tested, of course, but I suspect the answer is for benefits to increase relatively steeply, not year-by-year but in steps tied to each reenlistment contract. That way there would always be a significant gain if you re-up. On the flipside, the reformed benefit probably should not rise so steeply that it offers half-pay for life at 20 years, which is simply too good an incentive to get out. (Sorry). A 40-year-old master sergeant or lieutenant colonel may not be up to infantry combat anymore, but their decades of experience are invaluable for training troops, educating officers, devising doctrine, and advising on weapons development. We don’t want to pay them to leave. (Keeping them around would also require reforming the “up-or-out” system that kicks you out for not getting promoted fast enough).

Reducing retirement benefits for those who reach twenty years of service would definitely hurt; but if it helped pay for offering some benefits to those with less than 20 years – who currently get nothing – it would do more good for more people than it did harm. The military retirement system definitely needs to change.

High-powered commissions have been saying so for 40 years, but Congress has consistently rejected their ideas, throwing out the good with the bad. What we need is a system that aims at keeping our best people, not getting rid of them, and whose benefits are not at the discretion of the Pentagon, but guaranteed by law.