The Economics of Vocational Training: Past Evidence and Future Considerations. World Bank Staff Working Papers Number 713.

A partial survey of the literature on the economics of vocational training reveals three important lessons on how evaluations may be undertaken using data on pay, inputs, and outputs. The first lesson is that social, corporate, and private returns to vocational training in developing countries appear to be high enough to justify expanding training activity. However, training in industrial institutes and vocational secondary schools is less cost-effective than more formal firm-based training, at least in Kenya, India, and Israel. Also, Latin American data indicate that school and formal institutional vocational training may be substituted for one another. The second lesson is that in some sectors a more labor-intensive method of production is economically more effective than current methods. This may, in turn, imply the need for more--not less--skilled labor and vocational training. Further, given technology and output, unskilled and skilled labor appear to be substitutes, rather than complements, in the production process. The third lesson is that sophisticated function analyses are plagued by statistical and measurement problems. If they are not resolved, estimation techniques are unlikely to be able to pick up any relationship between trained labor and output among firms. However, there are alternatives. Many input and output measures such as performance rating and downtime are available for more modest evaluations. Training can be evaluated using earnings data or output and input data. In all cases, there are technical problems including sample size, control groups, the use of longitudinal data, and difficulties caused by labor mobility. A simpler before/after plant level study of changes in inputs or output associated with training may offer the most tractable evaluation. (112 references.) (Author/CML)