Education versus Cash Redistribution: The Lifetime Context (1979)

In what ways should the public allocation of educational resources be influenced by distributional considerations? Jencks (1972) argued that, since the correlation between education and experience-specific earnings was so low (around 0.3), education should not be considered an important instrument of redistribution. For even if all men had the same education, the variance of log earnings (within experience groups) would be reduced by only around 10 per cent.1 Therefore, he argued, the main emphasis should be on redistribution by means of cash.