The document presents the corporate and managerial policy results of quantitative models of an individual's propensity to terminate from manpower programs. The models were tested using a sample from 10 firms participating in the National Alliance of Businessmen (NABS) and the Job Opportunities in the Business Sector (JOBS) programs formed to identify, hire, and train unemployed members of the secondary labor market and to integrate them into the primary labor market. The withdrawal processes were studied using econometric simultaneous equation techniques along with multivariate statistical and probabilistic methods including reliability theory. A simultaneous model of turnover, retention, and absenteeism was developed, in order to investigate the hypothesized relationship between absenteeism and turnover. It includes the disaggregation of job satisfaction into a number of variables and also the introduction of supervisory variables. The study supports the assumption that turnover and attendance do not have identical determinants. Solely in terms of significance the findings show company policy variables to be the most important determinants of the propensity to terminate. The implications of the findings for hiring and supervisory policies relating to programs attempting to integrate disadvantaged workers are discussed at some length. A list of references is included and statistical data and analyses appended. (Author/EC)