There are perennial calls for drafting all 18 year olds to serve in either the military or some civilian alternative. Congressman Charles Rangel has repeatedly introduced bills in Congress (the "Universal National Service Act") that would do this. The bills have never come close to passage, and are unlikely to in the future even with Democratic control of both houses of Congress. But universal national service is one of those seductive ideas that refuse to die completely, and perhaps therefore it deserves a serious analysis. It is analytically interesting and can serve as an example of the utility of a cost-benefit approach to public programs.
Roughly 4 million Americans reach the age of 18 every year. There are only 1.4 million active-duty military personnel, so only a small fraction of each vintage of 18 year olds could be assigned to the military. At their present size, our active-duty armed forces require only about 150,000 new recruits each year. So any universal national service obligation would have to be primarily an obligation to do civilian work.
Civilian national service (in the United States--thus excluding the Peace Corps, and the missionary work that young Mormon men are required to perform for two years without compensation) funded by the federal government exists already. The "AmeriCorps" program provides federal grants to a large number of service organizations, both public and private. Although these organizations pay only the living expenses of their volunteers plus a modest education grant, the federal contribution amounts to some $27,000 per volunteer.
The number of volunteers supported by AmeriCorps grants is small--well under 100,000. But of course total volunteer activity is much greater than that, and by no means limited to young persons--an affiliate of AmeriCorps is the "Senior Corps." A survey by the U.S. Department of Labor found that there were some 60 million American engaged in volunteer activities in 2006 and that the median number of hours that the volunteers devoted to such activities was about 50 hours a year. Thus, assuming that the average is not much different from the median and that a full-time job is 2000 hours a year, there were the equivalent of 1.5 million full-time volunteers (50/2000 x 60 million). That number is important because a universal national service obligation would have a substitution effect: someone required by law to provide a year of national service would be likely to reduce the amount of volunteer service that he would provide in the future. If, for example, there were a two-thirds reduction in volunteering, from 1.5 million full-tine equivalents to 500,000, and thus a loss of 1 million full-time-equivalent volunteers, universal national service would augment volunteer activities by only 3 million full-time equivalents a year (4 million - 3 million). Granted, this number would rise if universal national service had a complementary effect on volunteer service rather than or, more plausibly, as well as a substitution effect--if, that is, the year of obligatory service created a taste for such service. I find this implausible.
If 4 million persons were conscripted for one year's national service, at an annual expense of $27,000 per person, the program would cost more than $100 billion a year--probably much more, because the $27,000 figure excludes the overhead expenses of the service organizations that receive the per capita grants. The $100 billion (or whatever the correct figure is) would be a transfer payment, but it would generate costs of two types. The first would be the deadweight costs that the taxes required to fund the payment would impose. The second and doubtless greater cost would be the difference between the value of the conscripts' national service work and the value of their output in whatever jobs they would have had were it not for their national service obligations. About half the 18 year olds would (but for their national service obligation) be in college rather than working, and so the effect of universal national service on them would be to postpone their entry into the job market by a year. Their lost wages in their first job would be a rough estimate of the value of their work in that job. The starting salary for college graduates is more than $40,000, other than for liberal-arts majors, and this is about twice the starting salary for high school graduates. That is some evidence that a universal national service program would be inefficient: it would in effect reallocate a year of a college graduate's working life from after college to before college, when he would be less productive.
Against this it could be argued that the national service work that the 18 year olds would perform would have a social value in excess of its private value. But this seems unlikely for most jobs that these teenagers would perform, such as helping out in hospitals and nursing homes and picking up litter on roadsides and in parks. A possible exception is tutoring children, since education produces significant social benefits. But only a small fraction of the 4 million national service conscripts could usefully be employed in that activity.
Universal national service would also have peculiar effects on the distribution of income. The unpaid national service workers would replace low-paid service workers, pushing many of them into poverty.
Proponents argue that, all narrowly "economic" issues to one side, universal national service would confer intangible social benefits in the form of increased solidarity, as all Americans would share in the experience of working for the overall social good without compensation beyond modest living expenses. But given the heterogeneity of the jobs that the national service workers would be performing, the solidarity-enhancing effect would surely be quite limited. It would be different if the 4 million were all drafted into the armed forces for a year, but that is infeasible.
In a candid moment proponents of universal national service might respond that its real purpose is to take rich kids down a peg by forcing them to work for a year with minimal compensation. The hope would be that the experience would make the rich empathize more with the poor and therefore treat them more generously. This seems unlikely, though the issue is worth studying. A person's attitude toward issues of distributive justice is shaped by a variety of factors, including temperament, parental values--and personal experiences not limited to a year's working without pay.