Social insurance contributions: frequency vs. earnings
Abstract: The ability of contributory pension policy to replace earnings in old age is heavily impaired when cumulative contribution frequencies are low. With evidence from Chile, this paper finds that cumulative contribution frequency is unrelated to average relative earnings for relative earnings above a threshold. However, below the threshold, lower relative taxable earnings are correlated with strong reductions in contribution frequency. Stratification by earnings of the frequency of contributions pushes low earners into partial pensions or none. Exclusion of data for individuals with less than 15 years of contribution, as in some OECD countries (e.g. Spain and U.S.A.), would conceal stratification in Chile. Stratification remains strong for different averaging periods and for younger cohorts. When young participants look forward and assess the risk of having low contribution frequencies in the future, they find that this risk reaches a peak at the third quintile of relative earnings, where it is twice as large as in the first and fifth quintiles. Exploring causality below the threshold, the paper finds that relative earnings causes cumulative contribution frequency, in the cross-section of lifetime outcomes. The discussion of mechanisms suggests an explanation of general applicability: enforcement inequality. This is linked to a short-term view of informality.