Cato Policy Analysis No. 1 May 15, 1981

Policy Analysis

Can Conscription Work?

by Roger Mils Folsom, with the assistance of Ann Arlene Marquiss Folsom

The author wishes to thank Richard S. Elster and George W. Thomas at the Naval Postgraduate School for suggesting data sources; Rudolfo A. Gonzalez and Geoffrey Nunn of San Jose State University for comments and suggestions; Velma Burrs, Education Coordinator for the San Francisco District of the U.S. Army's Recruiting Command, for help in understanding Army educational benefits; and especially Terryl L. Wisener, Office of the Assistant Secretary of Defense for Manpower, Reserve Affairs, and Logistics, for unpublished data tabulations. None of these people are responsible for the accuracy of the facts as presented and interpreted here, and none should be assumed to agree with the opinions given.

Executive Summary

In his sixteen years as a major political figure Ronald Reagan consistently opposed conscription. In his 1980 presidential campaign he made it clear that his opposition extended to President Carter's draft registration program. But today, after three months of the Reagan administration, draft registration is still in existence, and there has been no move by President Reagan to end it, a step he could take by executive order. This paper will review the condition of the volunteer armed forces, particularly in light of the concerns expressed by Sen. Sam Nunn (D-Ga.), a leading congressional critic of the volunteer force.

Recently Senator Nunn reviewed the performance of the all-volunteer armed forces over the past six years. He concluded that because of increasing difficulties in recruiting a representative selection of qualified men and women for military service, particularly in the Army's combat arms, "the All-Volunteer Force is in trouble now and the problems are going to get worse during the 1980s. No facile solutions are at hand." he was skeptical of solutions using fewer eighteen- and nineteen-year-olds, whether by using fewer people in total, more labor-saving capital equipment, more women, more older recruits with "civilian-acquired skills," or somehow reducing attrition. Although he stated that a compulsory national service program (including not only military but also civilian service) "for all youth . . . would ultimately be of great benefit to the nation,'' he noted its "very large" budgetary costs, asserted that the "Federal bureaucracy is not knowledgeable enough or capable of administering such a program," pointed out that worthwhile national service tasks would need to be identified, and concluded that national service "cannot provide the needed, immediate relief for the growing problems of the All-Volunteer Force." He dismissed higher compensation: "Pay and recruiting proposals are very expensive and would dramatically increase manpower costs."

Finally, he commented that although "peacetime registration would make a significant improvement in our mobilization capabilities . . . registration alone will not solve the problems of the All-Volunteer Force." Although elsewhere he has stated "I believe the All-Volunteer Force (AVF) may be a luxury that the United States can no longer afford," [2] he has been careful not to go on record explicitly advocating the restoration of military conscription. But many in Washington regard him as one of the foremost congressional forces pressing for some sort of draft. What other alternative is there? The fundamental alternatives, of course, are only two: individual choice and governmental coercion.

Certainly coercion is morally objectionable. Indeed, we Americans fought a civil war over the issue of compulsory labor. For most Americans that war, and the Thirteenth Amendment to the Constitution which followed it, outlawed involuntary servitude. Yet what is military conscription if not involuntary servitude, a condition totally inimical to the principles of the American Revolution as well as the Civil War? It should be clear that conscription is inconsistent with the principles of a free society.

But however one views the issue of coercion, it should also be clear that conscription is not in any sense necessary to American national security, and our purpose here is to counter Senator Nunn's implicit argument for coercion with additional information. Revisions in the structure of military compensation, together with other management changes, can solve the manpower quantity and quality problems outlined by Senator Nunn without increasing real manpower costs at all.

The Downward Trend of Real Military Compensation

Between 1962 and 1970 Congress raised real military pay substantially. Then, following the recommendations of the Gates Commission on an All-Volunteer Armed Force, between 1970 and 1972 Congress again raised real military pay, this time focusing on junior personnel with less than two years of service, whose financial welfare had been largely ignored since the Korean War. [3]

Congress subsequently failed to maintain real military pay in the face of inflation, however. Since the last draft call in December 1972, entering-level basic military pay (grade E-1) has risen from $307 to $501, a nominal increase of 67%. But concurrently the consumer price index rose from 127.7 (January 1973) to 263.2 (February 1981), or by 106%. In real terms adjusted for inflation, entering-level basic military pay has decreased by more than 20%. [4] Similar calculations hold for total military compensation (basic pay and allowances) at all pay grades. [5] Continuing inflation will decrease real military pay even more. The next pay raise is scheduled for the fall of 1981; if recent trends continue, it will be substantially less than the 1981 inflation rate.

Consider what has happened to educational benefits. Vietnam-era veterans' benefits (the GI Bill) could be as much as $13,900 for single veterans and $18,990 for veterans with two dependents, after only two years of service. These benefits are now unavailable except for veterans with active duty before January 1977. [6] Under the current Veterans' Educational Assistance Program, the service member contributes from $25 to $100 per month (to a maximum of $2,700) to a non-interest-bearing educational fund; for every $1 contributed, the government adds $2. Under this program the government's maximum contribution to educational benefits is only $4,800 for two and $5,400 for three or more years of military service.

In fiscal 1981, however, the Army is testing the recruiting effectiveness of four alternative packages of sharply higher educational benefits, available only to high school graduates of above average mental ability who enlist in a critical Army specialty, typically in combat arms. In two of these packages (the least generous of which was introduced in fiscal 1980), the soldier must contribute to the educational fund described above; in the other two packages, he does not do so. In one of the packages the benefits will be inflation-adjusted, can be transferred to dependents, and can be redeemed for 60% of their cash value. In these experimental packages the government's maximum contribution ranges from $6,800 to $12,800 for a two-year enlistment, $9,400 to $17,400 for a three-year enlistment, and $11,400 to $17,400 for a four-year enlistment. [7] In each Army recruiting district, only one of the four packages is available: The least generous is available to roughly half the population, and the others are distributed among the remaining half. Each is geographically dispersed. One can imagine the morale problems that will arise when a soldier who enlisted in a low-benefit district learns about others who happened to enlist in a high-benefit district and realizes that he could have done the same had he been fully informed about his alternatives. There may also be some resentment from members of other services, especially if their recruiters told them that all services have identical benefits.

Clearly educational benefits for most military personnel are now far lower than when the draft ended. (Even the experimental Army benefits are lower for two-year enlistments.) Their educational purchasing power is lower still, and Congress has not raised cash pay to compensate. Meanwhile, Congress has compounded the difficulty of recruiting by greatly expanding educational grants and loans available to civilians (Federal Guaranteed Student Loans and Parent Loans, Basic and Supplementary Educational Opportunity Grants, National Direct Student Loans, and college work-study programs), with no requirement for military service. [8] In fact, military educational benefits generally reduce the civilian grants and loans a veteran can receive. (Beginning in 1981, however, loans can be forgiven in exchange for Army military service on active duty or in the reserves.)

Educational benefits are an inefficient way to recruit a skilled military. They appeal only to people planning on higher education and civilian careers; they discourage military careers because to be a full-time student and obtain the maximum monthly benefit, one must usually leave military service. They don't appeal to potentially fine soldiers who have little interest in higher education. They don't appeal to those who distrust complex deferred compensation schemes from a government that frequently changes or "clarifies" its promises, ex post facto, in a society with rampant inflation. Thus the point is not that educational benefits should be restored to earlier levels: The demise of the GI Bill should have been accompanied by a real increase in cash pay (not by more complex programs with smaller benefits), by enough to replace whatever number of recruits would not have volunteered without the GI Bill. The increase in cash would not need to equal the reduction in the GI Bill's educational benefits because most veterans did not use their full benefit. (Even fewer were eligible for the maximum benefit.)

The Structure of Military Compensation

The Gates Commission criticized not only the amount but also the structure of military compensation:

Military pay today is a conglomeration of current and future pay and benefits that are difficult to enumerate and even more difficult to measure and evaluate. Military pay lacks visibility. It functions as a continuous source of controversy. It is inequitable. It is inefficient in attracting and retaining desired personnel. [9]

The commission recommended replacing the current mix of taxable basic pay, taxable reenlistment bonuses, and nontaxable allowances for subsistence and quarters with a simple, visible, fully taxable salary (including proficiency pay differentials for those working in critically scarce occupations) that could be compared with civilian wages. [10] The commission also wanted military pensions to vest (at least partially) after five instead of twenty years so that pensions would attract first reenlistments. It also suggested that enlisted personnel] should have the same limited "right to resign" that officers have: discharged on request, after a term long enough to pay back the government's cost of training, unless that request is to avoid foreign or sea duty or is during a national emergency. [11] Finally, the commission urged less reliance on non-cash compensation:

Compensation in cash has an inherent advantage. It allows each individual to decide how he or she will use whatever he earns. He can thus get the full value of whatever costs are incurred by the government in paying him. When he is compensated in non-cash form, however, the value of what he receives is often less to him than its cost to the government. Meanwhile, he is encouraged to consume more of particular goods or services than he otherwise would. Non-cash pay also tends to result in inefficient patterns of compensation by favoring some individuals (heavy users of these items) over others, independent of performance. [12]

Visibility and the family bias. Congress has ignored almost all of these recommendations. Military pay remains a conglomeration of payments and benefits that are lacking in visibility, controversial, inequitable, and inefficient. People in military service never actually see their total pay, not even on a pay stub or W-2 form. They substantially underestimate their total pay; the underestimate is most severe among junior personnel. [13] Military pay is perceived to be even less competitive with private sector pay than it really is.

Furthermore, because quarters allowances are substantially larger for married personnel [14] and because family-oriented, non-cash benefits (subsidized housing, medical and dental care, exchange and commissary privileges) appear so large relative to cash pay, military compensation is particularly attractive to persons with large, sickly families--the very opposite of the unencumbered single soldier favored by Senator Nunn. [15]

Thus the complex military pay structure discourages recruitment twice. First, the "lack of visibility" of financial compensation makes military service appear financially worse than it really is. Second, payment in family-oriented benefits instead of cash discriminates against the single, unattached individual who might be most receptive to military service--and who might make the best soldier, at least in the junior ranks.

Current versus retirement compensation. The recruiting problems created by the pay structure would matter less if military compensation were generous--but, at least for recruits, it is not. Without retirement pay, total compensation (basic pay and allowances, federal income tax savings due to allowances being nontaxable, department of defense contributions for social security, medical benefits, and exchange and commissary privileges) of a typical young enlisted person is somewhat below the median earnings of comparably-aged civilian white male high-school graduates. [16] For a single enlisted man, compensation is even lower. Including the present value of retirement benefits raises the total compensation of a typical twenty-year career enlisted person well above the median, [17] but retirement benefits have little effect on initial enlistments or first reenlistments because few of these people will reenlist for the required twenty years. [18]

Furthermore, deferring so much compensation as retirement pay that takes twenty years to vest, coupled with relatively small financial rewards for remaining in the service past twenty years, creates a military in which "careers are both too long and too short." [19] Talented individuals unwilling to put in twenty years quit early; others who contribute little still "hang on" for twenty years; and the best performers are discouraged financially from staying more than twenty Years. [20]

Another problem is that the present military pay structure is remarkably insensitive to the speed with which promotions occur. Even if less productive career personnel are promoted slowly, they earn almost as much as more productive personnel who are promoted more rapidly because "pay is more a function of years of service than of pay grade" (i.e., rank). [21] Consequently military pay probably is too generous for marginal people and not attractive enough for exceptional people.

Compensation reform versus conscription. In short, the present military pay structure is incredibly inefficient in attracting and retaining the most qualified military personnel. One estimate is that revising the military compensation structure to reduce this inefficiency could save over $3 billion annually. An additional $2 to $7 billion might be saved yearly by decreasing the proportion of inexperienced first-term personnel, reforming training programs, substituting relatively cheap capital equipment for increasingly expensive labor (particularly in support activities), and contracting out various tasks now done by civil servants. [22]

However, military pay has now fallen so far behind the private sector that some catching up is needed. Assuming only relatively minor changes in the compensation structure, the Congressional Budget Office recently estimated that recruiting and retention problems could be solved for less than $1.5 billion in 1981, plus whatever future raises are needed to keep military pay competitive with the private sector. [23]

On the other hand, reinstating military conscription would save little. Even if everyone with less than two years' service were a conscript, all other military personnel would still be volunteers, who must be paid enough to persuade them to stay. Furthermore, even conscripts should be paid something. The less conscripts are paid, the larger the implicit tax they bear, the more they will attempt to avoid service, and the more the government will have to spend to enforce the draft. [24] Furthermore, equity and social harmony would require that conscripts he paid something because fewer than one of every four young men and fewer than one in every seven young men and women will be called to serve in the 1980s. [25] If conscripts continued to receive present levels of compensation, restoring the draft could save "only a few hundred million dollars a year." If the 1970-72 relative pay increase for recruits were revoked, restoring the draft could save less than $2 billion a year. Even under the cruel and unrealistic assumption that two-year conscripts were paid absolutely nothing beyond their subsistence, restoring the draft could save at most $5 billion per year. [26]

If these estimates have any validity at all, the question should not be whether Congress will be willing to appropriate the funds necessary to sustain the all-volunteer armed forces. Rather, the question should be why Congress would want to reactivate the draft when it could so easily strengthen the financing of volunteer forces simply by reforming military compensation and by adopting other policies to exploit the savings that volunteer forces permit. Compensation reform could raise military effectiveness while saving taxpayers' money, by increasing retention rates and reducing training costs.

Allegations of Failure

Every instance of volunteer force "failure" levied by Senator Nunn is either fallacious or due primarily to military policies enacted or condoned by Congress.

Reserve forces. Senator Nunn states that "the selected reserve--those reservists organized into units that would augment active military forces in a mobilization--are more than 30,000 below authorized strength." [27] These reservists receive only one day's basic pay for each drill session attended; they receive no quarters or subsistence allowances (except while on active duty for training), or any significant medical, exchange, or commissary privileges; their retirement plan is much less generous (even on a per-day-worked basis) than the retirement for active duty careerists; and therefore their hourly compensation is much lower than that received by active duty personnel--although service in the reserves must compete with opportunities for civilian overtime. [28]

Senator Nunn states that the individual ready reserve (basically a list of former servicemen with less than six years of active and selected reserve duty) is "500,000 below mobilization levels for the Army alone." He fails to mention that these "reservists" need not train or maintain their skills and typically receive no compensation.

He states that "manpower problems in the event of mobilization are so severe that the military services are not now capable of meeting our national emergency requirements in terms of manpower." Given the sad condition of the reserves, this assessment is probably true. Rather than revitalize the reserves, [29] "the Senate Armed Services Committee has urged that peacetime registration of young men be reinstituted," even though many months after the onset of an emergency would be needed to activate training facilities and properly prepare the first draftees for war. Congress has now authorized registration of all young men aged eighteen through twenty.

Compared with well-trained and -equipped reserves, registration is useless and expensive. One government estimate is that peacetime registration could save fifteen days of the seven months needed to prepare a draftee for war. Another government study states that peacetime pre-mobilization registration would save at most seven days, compared with standby procedures for efficient post-mobilization registration, and cost an additional $14 million annually [30] -- not counting all the government's enforcement costs, or any of the resources consumed in efforts to avoid registration. Nevertheless, in 1980 Congress authorized registration of all young men aged 18-20.

Racial composition. Under the heading "The Trend in Characteristics of Those Who Do Serve is Adverse," Senator Nunn notes the large proportion of blacks entering and staying in the enlisted ranks, particularly in the Army where "29% of the total enlisted force was Black" [31] at the end of fiscal 1978. Then he writes: "The issue is not why Blacks are attracted to military service, but rather why other groups are not attracted."

The probable reason is that larger proportions of these other groups have increasingly better nonmilitary opportunities than many blacks do. In this connection the senator does not mention the 20% decline in real basic recruit pay since 1973 or the sharply reduced military educational benefits, coupled with massive increases in federal aid to civilian students. Nor does he mention congressional failure to establish a visible compensation structure flexible enough to correct imbalances in the supply of recruits going to various occupational military specialties. That only part of military pay is easily visible probably discourages non-black enlistments more than it discourages black enlistments because civilian wages are higher (and unemployment rates are lower) for non-black youth than for black Youth.

Of course, high black participation in the military raises "the image of black men dying for a white man's war." [32] But most of the military is white, even in the Army's combat arms. [33] And a major conflict would greatly expand the military, inevitably increasing non-black participation in the fighting and in the casualties. Admittedly, if combat forces do become predominantly non-white, casualty lists containing disproportionate numbers of Blacks and other minorities would be especially awkward during a regional conflict in Africa or other non-European theater. This problem is best avoided by avoiding American military intervention in such conflicts.

In any case, blacks could not possibly benefit from a return to military conscription. Those kept out of the military (to make room for draftees) would have to rely on their relatively poorer civilian alternatives. Those drafted against their will probably would be forced to serve at wages lower than their civilian opportunities.

Recruit quality. Another of Senator Nunn's concerns is the "large increase in the second lowest mental category" (IIIB) recruited into the Army, Marines, and Navy. [34] He fails to mention that the services deliberately increased this category so that they could cut the proportion of recruits from the lowest legally acceptable broad mental category (IV) to less than half of what it was before the Vietnam War and to one-fifth of what it was at the peak of the Vietnam War. (However, there is some evidence that recent test scores may exaggerate mental abilities, at least at the lower end.) [35] These changes resulted from policy decisions designed to improve quality rather than from a spontaneous deterioration in quality, as Senator Nunn implies.

The senator critically observed that in 1978 "only 5%" of the Army's recruits were from the top 8% of the population (mental category I). [36] But most category I individuals would go on to college, qualifying them to serve as officers.

The upper two mental categories (I and II) may better indicate the quality of the enlisted ranks. These upper groups, containing the top 35% of the population, provided 34% of all the services' recruits in fiscal 1978 and 29% in 1979. [37] Could more be expected, particularly while military compensation is falling further and further behind inflation and civilian compensation?

Senator Nunn also complains that the proportion of recruits who have graduated from high school has not kept pace with the increasing proportion of high-school graduates in the whole population. [38] But decreased recruitment of high-school graduates is the price the military paid for its huge cuts in mental category IV recruitment. Many of the rejected category IVs were high-school graduates, while many of the higher mental category IIIBs recruited instead were not. Of course, the trade-off might not be necessary if military compensation were more visible and if less of it were deferred until retirement.

It is true, as Senator Nunn states, that military manuals have been rewritten to lower grade levels. [39] But he presents no evidence that rewriting has diminished the effectiveness of training manuals; indeed, given their traditional lack of clarity, one suspects the contrary. In any case, the decline of reading skills is a nationwide trend not limited to military recruits. Publishers have been reducing the reading levels of all sorts of books, including college textbooks. In fact, military recruits may read somewhat better than other young civilians. In 1977, for all the services together, proportionately more recruits read above the tenth-grade level than did the young civilian non-college population from which most of these recruits come. And even though Army recruits read more poorly than other recruits, proportionately more Army recruits read above the eighth-grade level than did the civilian non-college population. [40]

Despite these data on the mental categories from which recruits come and on their reading skills, it still could be argued that the quality of recruits is too low. After all, the peacetime armed forces will be the cadre for any expanded military force. But conscription would lower, not raise, recruit quality.

When a Service is drafting, it conscripts college youth who would not voluntarily enlist, but it cannot turn away minimally qualified volunteers to draft higher qualified non-volunteers. In the draft years, there were about 5% more accessions in the highest levels (mental groups I and II) in the armed forces than under the all-volunteer force, but there were also more than three times as many in the lowest category used by the military (mental group IV). [41]

Senator Nunn and his congressional colleagues must face the fact that the only feasible way to improve the quality of a given number of recruits is to make military compensation either more efficient or more generous or both.

Attrition and discipline. Other concerns expressed by Senator Nunn are attrition during first enlistment and discipline. [42] But the Senator does not mention that attrition is largely controlled by the services:

Some of the increased attrition that has been observed under the All-Volunteer Force is of concern to the Department of Defense and actions are being taken to lower it; however, the capability to reject early those individuals who do not measure up to service standards is considered desirable and has had a net positive effect on morale. During the draft years, it was often necessary to retain malcontents and poor performers because of fears that their release from duty would be considered as rewarding poor performance and would encourage others who were serving involuntarily to seek a similar escape. Simple justice required that when some were being forced to serve against their will others could not be excused simply because they did not adapt to the service. As a result, incarceration was frequently selected as the solution to a malperformance problem rather than early release. [43]

[Since 1970] the greatest attrition increase has occurred in the 0-6 months of service period due to introduction of the trainee and expeditious discharge type programs. These programs were begun to facilitate the release of individuals who did not adapt to military life. While it is important that the services be able to release malcontents and people who do not adapt to military life, we have gone too far and are now releasing many persons who could have productive careers in the military. [44]

For example, non-prior-service male attrition during first enlistment was 26% for fiscal 1971 recruits (including draftees), peaked at 37% for fiscal 1974 recruits, and is scheduled to decline to less than 29% for recruits enlisting after fiscal 1978. For the same group, attrition during the first six months of service peaked at 16% for fiscal 1976 recruits, but it is scheduled to decline to less than 11% for recruits enlisting after 1981.

Increased attrition during the early years of the all-volunteer force caused--or at least permitted--substantial decreases since 1974 (when total disciplinary actions peaked) in both main categories of disciplinary action in every service. Nonjudicial punishment is down. Courts-martial (summary, special, and general) are down, not only below their peaks but also below their pre-Vietnam rates. Desertions are down in every service except the Navy,[45] where congressional refusal to build enough ships to meet the Navy's worldwide, congressionally mandated obligations has caused mushrooming unauthorized absences and desertions among grossly overworked below-decks personnel. Contrary to the implications in Senator Nunn's presentation of the Navy's desertion problem, [46]

there is no evidence to suggest that this Navy problem is attributable to the All-Volunteer Force. Conversely, the down trends in desertion in the Army and Marine Corps would indicate that the Navy's problems are independent of the All-Volunteer Force. Difficult living conditions, long working hours, and rigorous inspection schedules in the Navy have all contributed to higher desertion rates. Much of the increases, in fact, have occurred in relatively few engineering ratings where working conditions are unusually arduous. [47]

Readiness. Congress having created a compensation package designed to appeal more to people with dependents than to single, unattached individuals, Senator Nunn worries that young couples--or singles--with dependent children may not be ready to fight on short notice, particularly if they are assigned to Europe. After all, who will take care of the children if the grandparents are on the other side of the Atlantic?

He describes the plight of junior enlisted personnel assigned to Europe: "not entitled to government quarters"; having "inadequate" medical care, so that "for many stationed in remote locations, even minor medical treatment or emergency care means an overnight trip to a large city"; living "far from the post without other American neighbors and not in touch with German neighbors"; and "unprepared for high rents, car insurance, and utility costs" because "there are few opportunities for jobs for dependents." These junior enlisted personnel are apparently so poor that "many . . . do not have telephones." A "senior enlisted soldier may have to drive 60 miles in Paul Revere fashion telling people there is an alert." If they do not get the word or do not bother to respond (perhaps because of a spouse who is "very concerned over irregular military hours"), their military commander may be "reluctant to impose" a monetary fine as a penalty because of their "financial difficulties." [48] If this description is true and not a caricature, the responsibility clearly rests with Congress because all of the inadequacies are matters clearly under congressional control, as is the pay structure's lack of appeal to single, unattached individuals.

The readiness argument, of course, also raises the question, ready for what? Current U.S. policy is to be prepared to fight "one and a half" wars, or one major and one minor war. If our armed forces are not in a sufficient state of readiness for such conflicts, perhaps we should question the policy. The traditional American foreign policy, dating back to George Washington and Thomas Jefferson, is one of nonintervention in the affairs of other countries. If the United States did not attempt to be the world's policeman, we would not need such a large, ready army. A policy of nonintervention would be in keeping with the ideals on which America was founded, it would save billions of dollars and untold numbers of lives, it would reduce the threat of nuclear war, and it would grant to other countries the respect we expect them to grant us, thus avoiding situations like the Iranian hostage crisis.

It is true that conscription makes foreign intervention easier. It would be difficult to attract volunteers for another Vietnam-style war in El Salvador, Angola, or any other trouble spot. Conscription enables military and political leaders to intervene overseas with less regard to public opinion. But why should our leaders have the ability to involve America in a war that Americans would not voluntarily fight?

Patriotism. Senator Nunn fears that making military service financially rewarding enough to attract the needed number of quality recruits may impair, if not destroy, the military's patriotism. [49] But recall that the only alternative to volunteers is coercion. Doesn't a society demean military service if it pays so little to the minority defending it that they must be forced to do so? Especially when only a small minority of youth serves, can the draft's blatant inequities--with or without a lottery--fail to damage morale and fighting effectiveness? What evidence is there that involuntary servitude or governmentally imposed financial hardship strengthens patriotism or love of country?

The argument that . . . today's volunteers are less motivated by patriotism that they would be under a draft is simply illogical. . . . Under either the draft or the All-Volunteer Force, the military will take all the qualified young men who are motivated to enlist by patriotism. . . . The alternatives are to attract youth to the military by maintaining the pay and quality of military life at levels that are competitive with other occupations or to force youth to serve by coercion through some form of a draft. . . . Our most recent experience with the draft certainly does not suggest that conscription fosters patriotic feelings among the nation's youth. . . .

The implication that the draft fostered the ideals of duty, honor, and country is hard to accept. To many draftees, the "calling" came from no higher than the local draft board and was based more on a desire to avoid punishment than any sense of patriotism. This is evidenced by the humor of World War II and Korea and by the hostility that developed during the later years of the draft between the draftees and the career force personnel who were referred to disparagingly as "lifers." The career force has always been volunteers and generally is more interested in patriotism. Nor is it clear that the low pay given to enlisted personnel during the draft era contributed to the development of any patriotic sentiment on their part. . . .

To a large extent, military service should build patriotism, not start with it. In peacetime that is more difficult than during a war of national survival, such as World War II, but it is occurring on a daily basis. All of the evidence indicates that today's active force will stand up to any enemy and today's reserve force is perhaps more highly motivated and dedicated than any draft-motivated reserve in history. But overt patriotism is generally played down in today's society and the military reflects those norms. Perhaps the nation is better off with a military that fights well, rather than one that cheers well. [50]

The Case against Conscription

Essentially Senator Nunn is arguing that we cannot afford to pay military personnel the compensation necessary to attract the needed supply. This argument confuses budgetary outlays with real costs. Whether volunteers or conscripts, military personnel are removed from the production of nonmilitary goods and services. The civilian output lost to society is basically the same in either case. The primary difference is who bears the defense burden: society in general through taxes or those who happen to be of draft age in years during which the youth population happens to be small or military activity happens to be extensive.

The foregoing actually oversimplifies: We cannot afford military conscription, much less the chimera called "universal" national service because it would inevitably decrease both our nation's military capability and the economy's productivity.

Under conscription, the military cannot legally exclude low quality volunteers to make room for high quality draftees: Conscription lowers the quality of military personnel. And--especially when only a small minority of youth are "called to serve"--conscription cannot avoid destroying morale. For these reasons alone conscription must reduce military effectiveness.

Furthermore, individual capacities and aptitudes differ for military as well as for civilian occupations. Military forces can be most effective and the civilian economy most productive when individuals choose for themselves among competing military and civilian alternatives. A conscripted romantic poet may not make the best soldier. A conscripted machinist may be less productive in the military than in the civilian economy.

In a hostile world, a country whose economy is barely growing cannot afford conscription's wastes. It had better adopt efficient military compensation and other personnel policies generous enough to make its volunteer armed forces work, and work well.


[1] Sam Nunn, "Those Who Do Not Serve in the All-Volunteer Armed Forces," The Journal of the Institute for Socioeconomic Studies 4 (Autumn 1979): 10-21. The quotations in this and the following paragraph come from pages 19-20. Senator Nunn is a member of the Senate's Armed Services Committee and was chairman of its subcommittee on manpower and personnel in the Ninety-sixth congress.

[2] Sam Nunn, "Another Look at the All-Volunteer Force," The Washington Post, Monday 28 March 1977, p. A19. After discussing AVF "problems," he concluded: "I reject the notion that the only alternative to the AVF is a Vietnam-style draft. . . . The government should begin now to explore possible alternatives to the AVF, such as a national service program for all our young people."

[3] Thomas S. Gates, et al., The Report of the President's Commission on an All-Volunteer Armed Force (New York: The Macmillan Company, 1970), Tables 5-II and 5-III, pp. 52, 58. For an analysis see Richard V. L. Cooper, Military Manpower and the All-Volunteer Force (A report prepared for the Defense Advanced Research Projects Agency by the Rand Corporation, R-1450-ARPA, September 1977), pp. 109-10.

[4] After adjusting for inflation, real basic military pay (in 1967 dollars) in grade E-1 was $307/1.277 = $240 in January 1973, and was $501/2.632 = $196 in February 1981, a decrease of 20.8%.

[5] See John R. Thompson and Richard W. Hunter, "Subjective Criticism in Perspective," in Appendix G in America's Volunteers: A Report on the All-Volunteer Armed Forces (Washington: Office of the Assistant Secretary of Defense (Manpower, Reserve Affairs, and Logistics), 31 December 1978), p. 238.

[6] Charles J. Zwick, Report of the President's Commission on Military Compensation (Washington: U.S. Government Printing Office, April 1978), pp. 159-60.

[7] In the first two packages, called the "Super" and "Ultra" VEAP, respectively, the government adds (with no additional matching requirement) to the basic VEAP fund an Education Bonus: for a two-year enlistment, either $1,600 or $4,400 plus $100 or $300 for each contributing month beyond twelve, to a maximum of $2,000 or $8,000 (giving the total maximum government contribution of $4,800 plus either $2,000 or $8,000); for a three-year enlistment, either $2,600 or $4,800 plus $100 or $300 for each contributing month beyond twelve, to a maximum of $4,000 or $12,000 (giving a total maximum government contribution of $5,400 plus either $4,000 or $12,000); for a four-year enlistment, either $3,600 or $4,800 plus $100 or $300 for each contributing month beyond twelve, to a maximum of $6,000 or $12,000 (giving a total maximum government contribution of $5,400 plus either $4,000 or $12,000).

The third package, called the Non-Contributory VEAP and sponsored by the U.S. Senate, is like the Super VEAP above, except that the soldier's monthly contributions to the basic VEAP fund are paid by the government (not more than $2,400 for two-year enlistments and $2,700 for longer enlistments).

The fourth package, known as the "Mini GI Bill" and sponsored by the House of Representatives, gives the soldier for each year of service (up to four) $1,200 for tuition and $300 per month for a nine-month academic year, adjusted for inflation, transferable to dependents, or redeemable in cash for 60% of its value.

Thus comparing the regular non-Army VEAP and the experimental Army educational benefit programs, the government's maximum contributions are:

  Non-Army Army
Enlistment Basic VEAP Super VEAP Ultra VEAP Non-Contributory VEAP Mini GI Bill
2 year $4,800 $ 6,800 $12,800 $ 9,200 $ 7,800
3 year 5,400 9,400 17,400 12,100 11,700
4 year 5,400 11,400 17,400 14,100 15,600




According to a statement by Charles W. Moskos (Department of Sociology, Northwestern University) to the U.S. Senate Budget Committee on 3 March 1980, Veterans' Educational Assistance Program benefits (basic VEAP) will grow to only $87 million by fiscal 1984. Federal aid to civilian college students already has reached $4.4 billion in fiscal 1980 and is projected to rise to $7.0 billion in fiscal 1985. Moskos argues that aid to civilian college students should require some national (including military) service; he also favors restoring generous GI Bill educational benefits for military service. See his "Saving the All-Volunteer Force," The Public Interest (Fall 1980), pp. 74-89, particularly pp. 85-89; also see Morris Janowitz and Charles C. Moskos, "Five Years of the All-Volunteer Force: 1973-1978," Armed Forces and Society 5 (February 1979): 171-218.

[9] Gates Commission Report, p. 60.

[10] Ibid., pp. 57, 60. Of course, the military salary should be large enough to compare favorably. Ibid., Appendix B, p. 201.

[11] Ibid., pp. 61-64.

[12] Ibid., p. 62.

[13] Cooper, Military Manpower, pp. 365-66; Zwick, Commission on Military Compensation, pp. 102-3. The Zwick Commission argued that because a statistical study showed the probability of reenlistment to be affected equally by perceived compensation and by the difference between actual and perceived compensation (a suspiciously surprising result), low perceptions of total pay may not impede recruiting or retention. But these reenlistment results should not be used to predict the behavior of prospective recruits (who have no experience with the compensation system).

[14] For an incomplete list of pays, bonuses, and allowances, see Zwick, pp. 9-10, 18-20. For discussion, see Martin Binkin, The Military Pay Muddle (Washington: The Brookings Institution, 1975); also see Sar A. Levitan and Karen Cleary Alderman, Warriors at Work: The Volunteer Armed Forces (Beverly Hills, California: Sage Publications, 1977), chapter 2; and Cooper, pp. 361-64; Zwick, pp. 123-32. For a variety of unpersuasive reasons, however, the Zwick Commission did not favor changing to a fully taxable "salary." Ibid., pp. 106-10. For a more recent discussion, see Martin Binkin and Irene Kyriakopoulos, Paying the Modern Military (Washington, D.C.: The Brookings Institution, 1981).

[15] Nunn, "Those Who Do Not Serve," pp. 18-19.

[16] Using 1974 military and civilian data, Cooper, pp. 376-78; see also pp. 369-70. For similar results using 1978 pay rates, but apparently including only cash compensation and excluding both military and civilian fringe benefits, see Zwick, pp. 114-16.

[17] Assuming a 3% real (after inflation) interest rate, including the present value of retirement benefits, raises the total compensation of a typical twenty-year career enlisted person well above the seventy-fifth percentile of the earnings of comparably aged civilian white high-school graduates, and raises the total compensation of a typical twenty-year career officer well above the ninetieth percentile of the earnings of comparably aged civilian white college graduates. Cooper, pp. 376-78.

[18] Zwick, pp. 27, 49-51, presents survey data confirming this intuitive fact.

[19] Cooper, pp. 376, and 349-51.

[20] Zwick, pp. 26-27, 41, 42-57.

[21] This statement is less true in the very highest enlisted ranks and in the highest officer ranks. Cooper, p. 367, note 17, and pp. 370-71, 379. As did Cooper, the Zwick Commission (p. 139) recommended that a time-in-grade pay table replace the present grade and time-in-service pay table.

[22] Cooper, pp. 380, 319, note 57, 394, and also pp. 188-93, 277-303, 311-19, 343-55. These suggestions are discussed briefly in the Appendix available from the author.

[23] Restructuring retirement benefits could recoup some of this amount. Congress of the United States, Congressional Budget Office, Cost of Manning the Active-Duty Military, Staff Working Paper, May 1980, pp. 16-25, 27-32, particularly p. 17, Table 1.

[24] Cooper, pp. 83-86, estimated that in 1964 when the implicit tax paid by conscripts was between $2.1 and $4.4 billion, the costs-of-collection (including the potential draftees' costs of avoiding and attempting to avoid the draft) were between $2.6 and $3.6 billion. Under a lottery, the tax paid by conscripts would have been between $2.6 billion and $8.7 billion, and the costs-of-collection would have been between $0.7 and $3.7 billion. All these estimates were in 1964 dollars and would be more than twice as large in 1980 dollars.

[25] To maintain armed forces of approximately 2 million people should require less than half a million enlisted and officer accessions annually. The 18-year-old male population, now about 2.0 million, will fall, but even at its lowest in 1992 will exceed 1.6 million, after which it will rise again. Unless the armed forces increase greatly in size, less than 25% of the men who reach maturity in the 1980s will serve, and of course this percentage decreases the more that women serve. America's Volunteers, pp. 1-2, 69, 190, 209; Cooper, p. 192. For population projections, see U. S. Census Bureau of the Census, Current Population Reports: Population Estimates and Projections. Series P-25, No. 704, July 1977: Projections of the Population of the United States: 1977 to 2050, Table 8, Series II.

For an argument that women's participation in the services may be becoming excessive, see Seth Cropsey, "Women in Combat?" The Public Interest (Fall 1980), pp. 58-73.

[26] Cooper, chapter 11, particularly pp. 257-58, 261-62, and also p. 383, note 1. America's Volunteers, pp. 165-66 (also see pp. 177-80), estimates the "few hundred million" to be $250 million.

[27] The quotations in this section concerning the reserves all come from Nunn, "Those Who Do Not Serve," pp. 11-12.

[28] Recall that the total compensation (excluding retirement) of an active duty young enlisted person, including all the benefits the reservist doesn't get, is somewhat below the median earnings of comparable male civilians.

[29] For suggestions see Martin Binkin, U.S. Reserve Forces: The Problem of the Weekend Warrior (Washington, D.C.: The Brookings Institution, 1974).

[30] The fifteen days estimate comes from America's Volunteers, p. 152. The seven months is mentioned by Nunn ("Those Who Do Not Serve," p. 12), but he does not see this statistic's anti-registration implication. For the seven days estimate, see Director of Selective Service, Draft Registration Report, issued 16 January 1980, and released by Senator Mark Hatfield in the Congressional Record 126 (27 February 1980), pp. 1917-21, particularly p. 1918.

[31] Nunn, "Those Who Do Not Serve," p. 13. Between the ends of fiscal years 1978 and 1979, the black proportion of the enlisted ranks rose from 29% to 32% in the Army, 19% to 22% in the Marine Corps, 9% to 11% in the Navy, 15% to 16% in the Air Force, and 19% to 21% in the combined services. See Thompson and Hunter, pp. 191, 201, 203, 365, 372, and unpublished tabulations, courtesy of the Office of the Assistant Secretary of Defense for Manpower, Reserve Affairs, and Logistics.

The Gates Commission Report, p. 147, had predicted that in 1980 these enlisted percentages would be 19%, 16%, 8%, 14%, and 15%. The commission grossly underestimated the extent to which blacks would choose the Army instead of the other services but did better in predicting the proportion of blacks in the services overall.

[32] Thompson and Hunter, p. 246.

[33] Particular Army units do have high concentrations (as much as 50%) of blacks. One reason is that the Army has "unit of choice" and "station of choice" recruiting options that allow blacks to join units that already have many blacks, or which at least have reputations for good race relations. Levitan and Alderman, Warriors at Work, p. 175.

[34] Nunn, "Those Who Do Not Serve," p. 14. Mental category I contains the top 8% of those tested for military service; II contains the next 27%; and IIIA contains the remaining 15% who are above average. Of those below average, mental category IIIB contains the top 20%, IVA contains the next 10%, IVB the next 5%, IVC the next 6%, and V contains the bottom 9% who are legally ineligible for military service. Cooper, p. 127, Thompson and Hunter, p. 25.

[35] Thompson and Hunter, pp. 197, 370; and Office of the Assistant Secretary of Defense for Manpower, Reserve Affairs, and Logistics.

[36] Nunn, "Those Who Do Not Serve," p. 14. In 1979 the proportion of recruits from mental category I was 3% for the Army, 3% for the Marines, 6% for the Navy, 6% for the Air Force, and 4% for all services combined. In 1977 (when the GI Bill still was in effect) the Navy and Air Force obtained substantially more than 8% of their recruits from category I. Thompson and Hunter, pp. 27, 198.

[37] Thompson and Hunter, pp. 365, 370; Office of the Assistant Secretary of Defense for Manpower, Reserve Affairs, and Logistics.

[38] Nunn, "Those Who Do Not Serve," pp. 14-15. See Thompson and Hunter, pp. 30-31, 199, 369.

[39] Nunn 'Those Who Do Not Serve,", p. 14.

[40] Thompson and Hunter, pp. 33-35, 200.

[41] Ibid., p. 166. For further discussion of the quality of the all-volunteer force, see Ibid., pp. 233-36.

[42] In this connection, Senator Nunn mentions that "In the 20-year period preceding the All-Volunteer Force, the proportion of honorable discharges varied from 90 to 96 percent of those discharged. Under the All-Volunteer Force, honorable discharges have represented 86 to 89 percent of all discharges." Nunn, "Those Who Do Not Serve," p. 16. But since type of discharge is determined by military policies that change over time (especially over twenty-nine years), the relevance of this data is not clear. For a brief discussion of discharge and other disciplinary policies, see Levitan and Alderman, pp. 160-67.

[43] Thompson and Hunter, pp. 239-40.

[44] Thompson and Hunter, p. 67. The data that follow immediately below are from pp. 65-68, 208.

[45] Ibid., pp. 46-51, 205-6.

[46] Nunn, "Those Who Do Not Serve," p. 17.

[47] Thompson and Hunter, p. 50. (Abbreviations have been spelled out.) At the Hoover-Rochester Conference on the All-Volunteer Force (sponsored by the Hoover Institution, the University of Rochester, and the Sloan Foundation, 13-14 December 1979, at the Hoover Institution on War, Revolution, and Peace, Stanford University), it was stated that among below-decks engineering ratings (e.g., boiler technician) on Navy ships at sea, repeated 80-hour workweeks have not been uncommon. The statement was not challenged, although many very knowledgeable Navy Department and other defense analysts were present.

[48] Nunn, "Those Who Do Not Serve," pp. 18-19.

[49] Senator Nunn, "Those Who Do Not Serve," pp. 17-18, approvingly quotes Air Force General David C. Jones (Chairman of the Joint Chiefs of Staff) on this theme: ". . . the All-Volunteer Force . . . approach in recruiting of personnel for the services (means) moving away from the normative values of calling or profession embodied in the words 'duty, honor, and country,' toward an appeal based on monetary compensation and the values of the marketplace."

[(50)] Thompson and Hunter, pp. 243-45.

1981 The Cato Institute
Please send comments to webmaster