6 Conclusion
In this paper, we present a new data set for years of schooling intended to reduce
the measurement error present in existing series. We concur with Krueger and Lindahl
(2001) in the debate on errors in the measurement of data on schooling and
the implications for the estimated impact of schooling on income. Furthermore, our
results confirm those of De la Fuente and Domenech (2002, 2006), who highlight the
low quality of schooling data even for the subgroup of high-income OECD countries