Testimony of Steven M. Kosiak, Director of Budget Studies before the House Armed Services Committee Subcommittee on Military Personnel
It would be difficult to identify an issue more critical to U.S. national security than how we treat the men and women who serve in our armed forces. The effectiveness of the U.S. military depends, more than anything else, on the quality of our military personnel. Today, we are fortunate to have a military that is filled with men and women who possess outstanding abilities and skills, and are dedicated to serving our country, under what are often extremely trying circumstances.

Unfortunately, over the past year or so, some troubling signs have emerged. Specifically, the Services have had difficulty meeting their recruitment and retention goals. So far, recruitment problems have been concentrated primarily within the Navy and the Army, while retention problems have been largely limited to the Navy and the Air Force—especially among pilots. Moreover, taken as a whole, in terms of troop quality, today’s military remains far closer to the military that fought in Desert Storm than the so-called “hollow military” of the late 1970s.

Nevertheless, the administration and Congress are right to take the recent negative trends in recruitment and retention seriously. They are also correct that there is a need to act promptly, to guard against the danger that the Services’ very real—but still relatively limited—recruitment and retention problems might quickly escalate. Finally, the administration and Congress are probably right that part of the solution to these problems is to provide a boost in military compensation.

My main criticism of the administration’s proposal and, especially, the bill passed by the Senate Armed Services Committee (SASC) is that these measures might lock us into policies with substantial long-term costs and, in some instances, potentially serious negative consequences. According to the Congressional Budget Office (CBO), the proposals in the SASC bill would cost some $55 billion over the next 10 years. By comparison, the administration’s proposals would cost some $37 billion over this period.

A number of these proposed changes should be made, if at all, only after they have been given much more thoughtful consideration than they have been given to date, or can be given if they are to be enacted this year. In other words, Congress should wait until next year before considering some of the more far-reaching proposed changes. It should use the intervening months to thoroughly review these proposals, along with other, alternative options for improving recruitment and retention. Since some of the most important factors affecting recruitment and retention, such as the military’s high operational tempo, are at best only indirectly related to compensation levels, it is also critical that these alternative options include the impact of possible non-pay related changes. An ideal mechanism for considering these proposed changes would seem to be the next Quadrennial Review of Military Compensation, which is currently scheduled to report its findings in the spring of 2000.

That being said, let me make clear that I think there are also some very positive elements in the administration’s proposed compensation package, and some of these proposed changes should be implemented by Congress this year.

First, I think the larger, 4.4 percent across-the-board pay raise proposed by the administration is probably appropriate. Secondly, I think the targeted pay raises proposed by the administration represent an important improvement. The U.S. military’s pay tables have long rewarded performance too little relative to longevity. The administration’s proposal would go a long way toward fixing that problem by adjusting the pay table to more heavily reward promotions. Finally, I support the administration’s proposal to provide an additional $2 billion for other initiatives to improve recruitment and retention, including higher compensation for service members possessing special skills.

Taken together, these measure would add some $10 billion to the Defense Department’s costs over the next six years. They would also send a powerful signal to the men and women of the armed forces about the commitment of the administration and Congress to ensuring adequate compensation, not only next year, but over the long-term as well.

The most critical shortcoming of the administration’s plan is its proposed repeal of REDUX, the Military Retirement Reform Act of 1986. Under current law, personnel who joined the military on or before July 31, 1986 can retire after 20 years of service at 50 percent of their basic pay, while, as a result of REDUX, those who joined after that date will be able to retire after 20 years at only 40 percent of basic pay. The administration’s proposal would raise retirement benefits back to 50 percent of pay after 20 years for all military personnel. This change would cost about $6 billion over the next six years and some $1.5 billion a year thereafter.

I am not necessarily arguing that this change should not be made. What I am arguing is that any change in the military retirement system of this magnitude should be made only after the likely effect of the change and other possible options for changing military retirement have been thoughtfully, and deliberately, considered. To my mind, this means that the administration and Congress should wait until next year to make any major change to the military retirement system.

We need to be particularly careful in considering changes to military retirement for several reasons. First, it is difficult to determine what impact such a change would have on retention, or even precisely what the goal of such a change should be. Second, changes to military retirement, like changes to any government entitlement program, are politically very difficult to make and are made only very infrequently. This means that we will likely have to live with any change for decades to come.

Perhaps most importantly, we should be cautious about repealing REDUX because it so closely tracks the recommendations of previous commissions and study groups in its treatment of retirees with 20 years of service. Between 1947 and 1986 seven major study groups, appointed by the executive branch or Congress, considered possible changes to the military retirement system. The recommendations of those groups differed in many respects. But every one of the seven groups recommended that the immediate retirement benefits received by personnel leaving after 20 years of service be reduced to less than 50 percent of basic pay.

They did not make this recommendation out of any disdain for the men and women who have served their country in uniform for 20 years. They made this recommendation because they believed that providing 50 percent benefits after only 20 years of service was harming the Services’ ability to retain the quality personnel they needed. Among other things, the members of these seven different commissions were convinced that providing this annuity discouraged individuals from staying for more than 20 years.

It is possible that all seven of these commissions and study groups were wrong. But we should remember that this recommendation was made based on the Services’ experience over many decades. Now we are being asked to change military retirement back to the old system based on, essentially, the Services’ experience over the past year or so. To my mind such a rush to judgment would be ill-advised.

As I noted at the outset of my testimony, I also have concerns about the SASC proposals on pay and retirement. The Senate bill would provide even more generous benefits for military personnel who retired after 20 years of service than would the administration’s proposal. Among other things, it would provide larger cost-of-living adjustments and a thrift savings plan for certain military personnel. In addition, the Senate bill would increase funding for veteran’s readjustment benefits by nearly $5 billion over the next six years and some $2.5 billion a year over the longer term.

As with the administration’s plans for changing military retirement, I think it would similarly be a mistake for Congress to act on the proposed changes in the Senate bill before it has had time to thoughtfully and deliberately review both the likely impact of these changes, and other possible options for improving recruitment and retention.

My other main criticism of the SASC bill is that it contains a provision that would lock the Services into providing what may turn out to be excessively large military pay raises in future years. The Senate bill would provide an across-the-board pay raise of 4.8 percent in 2000, to be followed by pay raises set by statute at half a percentage point above the employment cost index, or ECI, thereafter.

The Employment Cost Index is a measure of how much wages and salaries of the civilian workforce have grown over time. Since 1981, military basic pay has grown more slowly than the Employment Cost Index, and the cumulative gap between the Employment Cost Index and military pay is now nearly 14 percent.

But the Employment Cost Index appears to greatly overstate the military-civilian pay gap. One problem is that the overall private-sector workforce differs substantially from the military’s workforce in terms of education, age, sex and other factors. In 1994, RAND developed an alternative index, called the Defense Employment Cost Index, which compared changes in military pay with changes in pay for the subset of the civilian workforce that most closely resembles the military. When this alternative index was used, RAND found that there was essentially no pay gap. And this was at a time when, according to the Employment Cost Index, the gap had grown to almost 12 percent.

Another problem with the Employment Cost Index is that it measures only changes in pay. It does not even purport to measure differences in the actual level of pay. This may be problematic, since the military received two very large pay raises in 1980 and 1981, the base year from which changes in the Employment Cost Index are measured. Moreover, a recent study that attempted to measure and compare levels of earnings rather than changes in earnings found that there essentially is no pay gap.

But the best reason to doubt the usefulness of the Employment Cost Index’s pay gap estimate is that, measured by that index, a large pay gap has existed for most of the past 15 years. Yet until the last year or so, the Services have been extremely successful at recruiting and retaining quality personnel.

This does not necessarily mean that military pay should not be increased substantially over the next six years. If the Services’ continue to experience serious problems with recruitment and retention, it may well be wise to increase military pay by more than required under current law. This would be true even if the Defense Employment Cost Index developed by RAND, for example, continues to show little or no pay gap.

On the other hand, the combination of an extra-large across-the-board pay raise in 2000, substantial targeted pay raises and various recruitment initiatives might prove to be enough to reverse current problems with recruitment and retention. Alternatively, these problems might be stemmed by the use of improved deployment patterns that reduce personnel tempo, such as the Air Force’s Air Expeditionary Force concept, other non-pay related changes, or a downturn in the civilian economy. In either case, the Services should not be locked into providing overly large pay raises for the indefinite future—especially on the basis of a measure as suspect, in terms of its relevance to military pay, as the Employment Cost Index.

In closing, let me reemphasize that there are parts of the administration’s new military compensation package that I strongly support. These include the proposed 4.4 percent pay raise in 2000, substantial targeted pay raises, and initiatives to improve recruitment. I also think that Senate Armed Service Committee’s bill includes some provisions that may ultimately be worth enacting.

But because of the complexity of the issues involved and the importance of “getting it right” the administration and Congress should wait until next year before considering major changes to the military retirement system, or large increases in veteran’s adjustment benefits. Likewise, the administration and Congress should refrain from locking the Services into future pay raises that may well turn out to be higher than are required to maintain a quality military.

©1999 Center for Strategic & Budgetary Assessments. All Rights Reserved.