Congress has increased active duty pay in the last few years to address retention problems caused by years of earlier pay raise caps. Consequently, we've received many, "What about me?" letters from members who retired while pay was depressed. Having retired well before the current round of pay plus-ups began, I understand the sentiment. But we need to get over it. Active duty and retired pay adjustments are based on different standards.

Active duty raises are supposed to be based on private sector pay growth to ensure military pay stays comparable to private sector pay, as measured by the Bureau of Labor Statistics' Employment Cost Index (ECI). But the government historically has capped them below that standard more often than not. Repeated retention problems have shown that to be a stupid practice, but federal budgeteers don't seem too interested in learning from history.

Retired pay increases, on the other hand, are cost-of-living adjustments (COLAs) that track with inflation, as measured by the Consumer Price Index (CPI). Their purpose is to ensure retired pay purchasing power isn't eroded by inflation over time. The ECI, which tracks wages, and the CPI, which tracks prices, are rarely the same. In recent years, the ECI has been higher, but there have been long stretches in the past when the CPI came out on top.

Before 1958, all retired members' pay was tied to the active duty pay table. When the active troops got a raise, retirees got the same increase. But there were no active duty raises from 1958 to 1963, and Congress shifted retirees to the COLA concept to guard against erosion of their real retired pay value.

Since then, the ups and downs of the CPI versus the ECI have generated differences between the retired pay of people who retired at different times with equal grades and years of service. Someone who retired in 1972 or 1973, just before the high-inflation decade, makes significantly more retired pay than one who retired in 1958 or 1980, just before the two double-digit pay raises of 1981-82.

But those are the extremes. For most officers, the variance is 10 percent or less. Most newly retiring servicemembers also face the relative disadvantage of having their retired pay based on their highest 36-months' average basic pay rather than final basic pay, as experienced by retirees who entered service before 1980.

Will Congress ever agree to base every retiree's pay on the current active duty pay table? No, in part because that would give some retirees a pay cut, and it would cost lots of money to raise others to that level. There's been no support for this idea in Congress for almost 30 years, while we have 40 years of philosophical support for tying military retired pay COLAs (with all other federal annuitant programs) to the CPI.

In the end, active duty pay raises are a retention issue. If the pay at the time was sufficient to keep you and me for a career, it's hard to make the retention case that Congress owes us a retroactive pay raise after retirement.

Also, the time will come when Congress and others will want to cut retired pay COLAs, as was proposed many times in the early 1990s. We'll have plenty of challenges defending the CPI-based COLA system in the years ahead.

The CPI won't lag the ECI forever. Back in 1980, we thought inflation would never again go below 6 percent a year. Anyone today who thinks it will never again go above 3 percent -- or above active duty raises -- needs to learn from history.

Col. Steve Strobridge, USAF-Ret., director of MOAA government relations