Measuring Socioeconomic Status
Socioeconomic representation has been included in the annual Population Representation in the Military Services since the FY 1986 report. However, there were no reliable socio-economic data to report at that time. Available data included the zip code of a recruits current address and associated statistics from census data. While this type of data is useful for demographic trend analysis and advertising and marketing research, it is not reliable for comparing socioeconomic representation in the military to that of the general population. For example, applicants and recruits may not come from the background indicated by the zip code for their current address (i.e., these individuals may move away from home to go to college or to work). 
In FY 1999, the Survey of Recruit Socioeconomic Backgrounds, first administered in March 1989, was again administered at recruit training centers. Participants answered questions about their parents education, employment status, occupation, and home ownership. While income is a widely used measure of socioeconomic status, research provides evidence that recruit-aged youth are not accurate at estimating their parents income.  Therefore, home ownership was included as a proxy for income.
Several researchers have devised a summary statistic for socioeconomic status.  The socioeconomic index (SEI), derived from predicted prestige scores based on levels of income and education within occupations, is one means of defining socioeconomic status. Stevens and Cho  developed such scores for each 3-digit occupation code in the 1980 Census, revising earlier work by Duncan, and Featherman et al.  More recently, this index has been revised by Hauser and Warren  to incorporate prestige ratings from the General Social Survey conducted by the National Opinion Research Center,  as well as occupational income and education data from the 1990 Census. This report uses a version of the SEI that incorporates income and educational data about both males and females; it is termed the Total Socioeconomic Index (TSEI). TSEI scores for recruits can be calculated using parental occupational information reported in the Survey of Recruit Socioeconomic Backgrounds.
In FY 1999, the Survey of Recruit Socioeconomic Backgrounds was given to both active duty and Reserve Component recruits without prior military experience. Approximately 14,100 active duty and 3,500 Reserve Component enlisted accessions provided information on the marital status, education, employment, and occupation of their parents.  The survey requested information on the parents with whom the recruit was last living, whether they were biological parents, stepparents, or other legal guardians. Throughout this discussion, these will be referred to as recruit or DoD parents.
For civilians, similar information is collected by the Bureau of the Census. These measures include marital status, highest level of education, home ownership, employment status, and occupation. For comparison, information is provided for parents of civilian youth between the ages of 14 and 21, inclusive, who were living at home. These data are taken from the Current Population Survey (CPS), an ongoing survey conducted by the Bureau of the Census for the Bureau of Labor Statistics.  They will be referred to as CPS parents.
Comparisons between DoD and CPS parents should be tempered by the fact that the DoD group does not include officer accessions. Since Active Component officer accessions represent nearly 8 percent of total Active Component accessions, adding officer socioeconomic measures could produce a moderate change in the overall DoD results. However, for most of the variables reported in this section, including officer data would produce little change in the reported values, because the civilian and military distributions are quite similar. Specific areas in which adding officer data might change the comparisons will be noted in the following discussion.
 Stevens, G. and Cho, J.H., Socioeconomic Indices and the New 1980 Census Occupational Classification Scheme, Social Science Research, 14 (1985), pp. 142168.
 See Duncan, O.D., A Socioeconomic Index for All Occupations, in A.J. Reiss, Jr. (Ed.), Occupations and Social Status (New York: Free Press, 1981), pp. 139161; Featherman, D.L., Jones, F.L., and Hauser, R.M., Assumptions of Social Mobility Research in the U.S.: The Case of Occupational Status, Social Science Research, 4 (1975), pp. 329360.
 Hauser, R.M. and Warren, J.R. Socioeconomic Indexes for Occupations: A Review, Update, and Critique (Madison, WI: Center for Demography and Ecology, June 1996).
 Nakao, K. and Treas, J., Updating Occupational Prestige and Socioeconomic Scores: How the New Measures Measure Up, in P. Marsden (Ed.), Sociological Methodology, 1994 (Washington, DC: American Sociological Association, 1994), pp. 172.
 Navy recruits who said that they were in the TARS program were counted as active duty recruits.
 To facilitate comparison between the military and civilian data sets, the CPS data were weighted to match the military data in terms of age. CPS sample weights were ratio‑adjusted to age distributions, in 5-year intervals, of recruits parents. Consequently, the adjusted CPS data contain the same percentage of parents in a specific gender and age group (e.g., male parents age 4044) as the military data set. When sample sizes are large, small differences in magnitude can be statistically significant. For comparisons between DoD and CPS parents, any difference greater than about one percentage point is statistically significant; the comparable figure for comparisons between Services or between active duty and Reserve Components is 3 percent.