Cover: Military Compensation

Military Compensation

Trends and Policy Options

Published 2000

by Beth J. Asch, James Hosek

Research Brief

Since 1998, the armed services and Congress have engaged in a vigorous policy debate on military personnel issues and the adequacy of military compensation. Although this debate is not over, so far it has led to the passage of significant increases in military compensation, a restructuring of basic pay, and a re-reform of military retirement benefits. In addition, certain special pays have been increased, a step has been taken toward the introduction of a 401(k)-type thrift savings plan for service members, and a number of proposals have been put forth to increase the level of educational benefits.

The origins of this policy debate lie in concerns about readiness, retention, and recruiting. The pace of peacetime operations is starkly higher in the post-Cold War world, and the services have found themselves stressed to maintain readiness for major theater war while coping with a steady stream of peacekeeping, disaster relief, and humanitarian missions. The extraordinarily vibrant economy has created superb civilian job opportunities, and the services have had increasing difficulty holding onto pilots and personnel in technical areas such as information technology. In addition, the services have been hard-pressed to meet their goals for high-quality recruits, and the Navy, Army, and Air Force have at times missed their overall recruiting goals—an outcome not seen since the manpower crisis of 1979-1980. Adding to tensions, personnel who entered service after July 1986 and who hit their 10th to 14th year of service came to realize, and resent, that their retirement benefits were slated to be less generous than those of their immediate predecessors. Finally, concerns about military quality of life, especially in health care and housing, have continued to simmer.

By mid-1998, the services, the Office of the Secretary of Defense, and Congress all understood the need to learn more about the factors affecting readiness, retention, and recruiting, and to devise policy options to remedy the situation. Within this broad context, the Under Secretary of Defense for Personnel and Readiness asked RAND to undertake a special study of military/civilian pay comparability, the role of heightened peacetime operations on retention, and the costs and benefits of alternative policy options. RAND's research effort drew upon and extended its long tradition of work on defense manpower and compensation. The results were documented in a report titled Military Compensation: Trends and Policy Options by Beth J. Asch and James R. Hosek. The two authors supplemented their analyses with congressional testimony in 1999, which RAND has reproduced in a document titled Military Compensation: Testimony to the Senate Armed Services Committee, Subcommittee on Personnel.

RAND found that, since 1993, overall military pay has declined relative to civilian pay, by 6.5 percent for enlistees and by 10 percent for officers. These declines in relative pay have occurred as the U.S. economy has boomed, cutting into the services' ability to attract and retain high-quality individuals. RAND analysis suggests that retaining and motivating high-quality personnel in this environment can best be achieved through targeted compensation: By giving higher pay increases to personnel in higher grades and by offering reenlistment bonuses, the services can most effectively tailor their forces to meet changing needs. Further, contrary to conventional wisdom, RAND found that personnel involved in long or hostile duty are often more, not less, inclined to reenlist.

Concern Over Recruiting and Retention

It's no secret that attracting and keeping high-quality personnel have been challenges for the military services during much of the 1990s. The services' difficulty in meeting their recruiting targets has been acute, and each service has reported difficulty retaining individuals in certain skill areas. The Air Force has had low retention of pilots, and in all services reenlistments are low in technical fields, where private-sector opportunities are unusually strong at present. Also, Army first-term attrition has been high.

At the same time, all the services report that mid-career personnel have expressed growing discontent with the military retirement system, arguing that changes put in place in 1986 have left them less well off than peers who joined the military in earlier years. Such a disparity has lowered morale, the services report, and could be a factor undermining mid-career retention.

Is Military Pay Keeping Pace with Pay in the Civilian Sector?

The adequacy of military pay has been the subject of intense policy debate in recent years. According to a comparison based on the Employment Cost Index, military pay has grown 13.5 percent less than has civilian pay since 1982. This has created a widespread belief that a civilian-military "pay gap" of 13.5 percent exists.

However, this comparison may not be appropriate for the military. Compared with the civilian labor force, active-duty personnel are younger, are more likely to have completed high school (enlisted personnel) or college (officers), have different occupations, are more than 80 percent male, and are more likely to be black and less likely to be Hispanic. It is important to account for these differences between the composition of active-duty personnel and the civilian labor force; civilian wage changes are not the same for every group but can, and do, differ by age, education, occupation, gender, and race or ethnicity.

Percentage Difference in Pay Growth for Military Personnel Compared with Civilian Counterparts, 1982–1997

RAND developed a civilian wage index that controls for all these characteristics. This alternative index reveals that, during the 1980s and 1990s, military pay growth kept pace with or outpaced civilian pay growth for some personnel, but lagged for certain other personnel (see the figure). In particular, from 1982 to 1997,

  • basic pay for enlisted personnel grew about 5 percent faster than did pay for civilian workers of comparable age, education, occupation, gender, race or ethnicity:
    • for junior enlisted personnel, basic pay grew about 10 percent more than did pay of civilian counterparts
    • for senior enlisted personnel, basic pay grew 5 percent less than did pay of civilian counterparts.
  • basic pay for officers grew about 20 percent less than did pay of their counterparts in the civilian workforce.

Not surprisingly, civilian pay has grown more rapidly than basic pay during the economic boom that has been under way since 1992. Junior and senior enlisted personnel and officers all have seen civilian pay grow at faster rates than their military pay in recent years.

How has this situation affected retention? Our analysis suggests that this recent slide in relative pay and a robust civilian economy have made enlisted personnel 11 to 14 percent less likely to reenlist after their first tours.

Has the Heightened Pace of Work During Peace Operations had an Adverse Effect on Retention?

Personnel tempo—the rate at which personnel are used in conjunction with military operations—has become another matter of intense policy debate. The nation's peacetime operations clearly have intensified since the end of the Gulf War, and many defense observers have suggested that this development has reduced the propensity of service members to reenlist.

However, a separate RAND analysis of assignment patterns and retention—Does Perstempo Hurt Reenlistment? The Effect of Long or Hostile Perstempo on Reenlistment, by James Hosek and Mark Totten, 1998—found that having some long (30 or more consecutive days away from home base) or hostile duty, but not too much, actually tended to increase reenlistment, relative to personnel who had no such duty. In particular, the pattern of long or hostile duty in the period 1993-1995 tended to increase first-term reenlistment in the Army and Marine Corps and early-career reenlistment across all services. Early career refers to individuals beyond their first term but with less than 10 years of service. This relationship between deployments and retention was the opposite of what many policy observers expected.

This positive effect was particularly strong for Army first-term enlisted personnel. Army first-termers with no prior long or hostile duty were 28 percent more likely to reenlist if given an initial three months of nonhostile duty—and 13 percent more likely given the same length of hostile duty—than were personnel experiencing no such assignments. But adding an additional tour of duty atop the first—such as another three months away from home—reduced the likelihood of first-term reenlistment, especially in the Army and Marine Corps. The negative effect of the extra tour was stronger when it involved hostilities. For Army first-termers, an additional three-month hostile-duty assignment made them 17 percent less likely to reenlist than if they had no long or hostile duty.

The study suggests that the services should spread the burden of peacetime military operations among personnel to the maximum extent compatible with readiness, particularly if personnel-tempo levels have risen above the levels prevailing during the study period. Further, because personnel are being used more intensively today, the net positive effect of perstempo on reenlistment could have declined and in some cases might have become negative.

How Well Do Alternative Compensation Packages Work in Retaining High-Quality Personnel?

As part of this research, the Office of the Secretary of Defense asked RAND to evaluate several alternative compensation packages intended to increase retention and decrease discontent with the two-tier retirement system put in place in 1986 that gave individuals who enlisted up until then a more lucrative retirement payout than those who enlisted thereafter. Although the packages differed, RAND considered several options, including the following:

  • Returning to pre-1986 retirement benefits
  • Offering a thrift savings plan
  • Providing across-the-board pay raises
  • Increasing special bonuses
  • Providing targeted pay raises.

Using a computer model, RAND researchers predicted the effects of the packages on retention, productivity incentives, and costs. Among all the alternatives, targeted pay appeared to be most effective.

A 4-percent pay raise targeted toward personnel in the E5 to E9 grades would increase retention by 7 percent and productivity incentives by 11 percent—increases far greater than any other alternative could create. This targeted pay raise would specifically address the growing divergence of junior enlisted pay and senior enlisted pay relative to their civilian wage counterparts, as discussed above. Moreover, such a targeted pay raise would cost no more than rolling back retirement benefits to their pre-1986 levels.

However, no proposal, including a targeted pay raise, addressed the compensation system's limited flexibility. The current system, which pays full retirement benefits when service members reach 20 years of service, creates what some call "golden handcuffs." Members who have, say, 10 to 19 years of service face a big financial loss if they leave before 20 years. The services recognize this and tend to avoid separating individuals during those years, even when other factors—poor productivity, new requirements, the need to promote junior personnel—would make it desirable to do so.

Thus, to make the system more flexible, targeted pay raises would need to be supplemented by other compensation tactics, such as special bonuses, which would enhance retention in critical areas; separation pay, which would break the golden handcuffs; additional recruiting resources, which would raise budgets for recruiters, advertising, and college benefits; and a 401(k)-type thrift savings plan, which would allow service members to tax-shelter current income and save toward retirement. In combination, these tactics would help enable the DoD to flexibly manage its forces and retain high-quality personnel in key positions.

How Sound Have Recent Legislative Proposals Been?

After completing the analyses described above, RAND in 1998 and 1999 was asked first by the DoD and later by Congress to quickly assess several legislative proposals to revise military compensation then under consideration by policymakers. These assessments began in late 1998 with RAND's examination of the DoD's "Triad" proposal. Under this proposal, all service members in FY2000 would have received a 4.4-percent pay raise; mid-grade personnel would have received targeted, higher raises; and members retiring at 20 years of service would have received more-lucrative benefits, equal to 50 percent of their basic pay, up from the current 40-percent payout.

RAND found that these Triad provisions, which were then part of the Clinton Administration's proposed DoD budget, would have boosted productivity incentives by 7 percent and increased overall retention by 6 percent. Moreover, the measures would have increased by 20 percent the proportion of new enlistees who stay in service for 20 years.

RAND outlined its assessments of Triad and various other proposals in testimony before the Senate Armed Services Committee in March 1999. At that hearing, committee members asked RAND to provide further assessments of a Senate proposal known as S.4, which would have given a more generous pay raise than the Triad proposal and allowed personnel with 15 years of service to choose between taking a bonus to stay on the current retirement system or reverting to the system that was in place before 1986. This Senate proposal also would have included a thrift savings plan akin to what is available to many civilian workers.

RAND found that this proposal would purchase more retention for given increases in cost. However, RAND recommended minor changes—e.g., making bonuses dependent on pay grade rather than giving lump-sum payouts—so that the most productive service members would receive the bulk of the proposal's rewards.

A revised version of S.4 was included in the FY2000 Defense Authorization Bill, which passed both the Senate and the House in September 1999.

This report is part of the RAND research brief series. RAND research briefs present policy-oriented summaries of individual published, peer-reviewed documents or of a body of published work.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.