Adjustments in the demand for skills have been proposed to follow a secular pattern which favors high skilled workers. In this perspective, upgrading low skilled workers is potentially associated with substantial gains for the individual and for the society, but employers are reluctant to train low skilled, who in their turn are unwilling to participate due to financial constraints and/or a perception of low quality and/or returns to training. If this is a market imperfection, a remedy is suggested by public provision of formal education at upper secondary and tertiary level and with participants eligible for financial support. It would alleviate financial constraints, improve the quality of training and (information on) the returns and also circumvent employers reluctance to engage low skilled in training. However, the costs involved are potentially large and the economic return to formal adult education (AE) for low skilled, a crucial measure to assess if expenses should be increased or decreased, is a virtually unexplored issue.
The purpose of this paper is to address these uncertainties by analyzing Swedish register data 1990-2004 of low skilled siblings aged 24-43 in 1994. Sweden is a suitable country for this kind of study as the last decades have seen considerable public investments in AE. The empirical strategy is to estimate annual earnings difference-in-difference-in-differences (DDD) regression models with family fixed effects. Earlier studies of AE in Sweden which have presented incoherent results are reconciled (Ekström, 2003, Albrecht, Van den Bergh and Vroman, 2004 and Stenberg and Westerlund 2008). US studies have been based on accomplished course credits at community college, i.e. mainly tertiary level education. Jacobson, LaLonde and Sullivan (2005) found a year of studies was associated with a 9 per cent earnings gain for men and 13 per cent for women. A major difference compared with the present study is that participants generally had no public financial support. Thus, their results are not necessarily applicable to low skilled and/or to a situation where study allowances decrease the opportunity costs. One may also expect the returns to AE are lower in European labour markets which are characterized by more compressed distributions of wages as well as of skills than in the US.
The main finding is that a year of AE increases earnings by 4.4 per cent, but calculations imply that private returns alone are unlikely to cover the total costs for society. The necessary payoff to cover the total costs is attained if the social returns to AE exceed the private returns by a factor of around 1.5. The reasoning is stable to assumptions regarding the individual discount rate, years until retirement and/or the impact of AE on reductions in the public transfer payments.