Measurement and Cross-Country Differences

All results use tobit specication sons and daughters

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: e earnings, 1995–1998. All results use tobit specification. Sons and daughters pooled. Note: For the dependent variable, probit models based on the 1996 SIPP matched to SER were used to determine if zero earnings reflected noncoverage or nonworker status and were imputed accordingly. Fathers must have positive earnings in each year. When fathers’ earnings are from the 1984 SIPP, they must be present for all interview months and have no cases of nonresponse to earnings or income questions. Only those fathers successfully matched to their wave 4 questionnaire are kept in all samples. Standard errors are adjusted for within-family correlation when more than one sibling is present. higher than that found in the SIPP. A possible explanation for this result is that despite the imputation procedure, the topcoding of fathers’ earnings resulted in a more compressed earnings distribution among those with high net worth. In any case, the difference in point estimates when comparing the top and bottom quartiles (0.310) is still quite large and significant at the 10% level. Although the fact that net worth is measured during college-going years has been presented as a strength, it may actually be a limitation if the relevant period for measuring borrowing constraints should include earlier points in the child’s educational career. Cameron and Heckman (2001) and Carneiro and Heckman (2003) have argued that much of the inequity in human capital development arises early in life. In this case, empirical evidence of differences in intergenerational mobility by wealth at such a late age might actually be viewed as even stronger evidence in favor of borrowing constraints. One might argue that because net worth and earnings are highly correlated, any nonlinearities in the IGE may also be reflected in differences in by levels of net worth that may or may not be due to borrowing constraints. To address this issue, the same comparisons in table 10 were also done using quartiles of the earnings distribution ra...
View Full Document

This document was uploaded on 02/26/2014 for the course ECON 560 at UBC.

Ask a homework question - tutors are online