This study uses the British National Child Development Study to examine the effect of educational attainment on social capital at the individual level. Social trust and membership of voluntary groups are considered as two basic indicators of social capital. We employ the IV analysis and non-parametric bound analysis to tackle the problem of education endogeneity. Both approaches reveal that the OLS estimator of the educational effect suffers from an upward bias in the study of group membership. We do not observe any significant bias in the educational effect on social trust. Our empirical findings indicate that education has a positive influence in promoting social trust and membership of voluntary groups.
Key words: Educational attainment, social capital, IV analysis, nonparametric bound analysis.