Measurement and Cross-Country Differences

This said the estimates are within one standard

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: he only change TABLE 8.—LEAST SQUARES ESTIMATES OF REGRESSION TO THE introduced in row 4 is to broaden the sample of sons to include those with one parent who was born in Canada, more akin to the definition used by Borjas (1993). The MEAN MODELS OF EARNINGS FOR IMMIGRANT Sample Selection Rules Outcome Fathers 1. ln weekly earnings for daughters - male, immigrant - 16 to 65 years - 1981 Census 2. ln weekly earnings for daughters at age 31 - male, immigrant - 16 to 65 years - 1981 Census 3. ln weekly earnings for daughters at age 31 - male, immigrant - married or in a common-law relationship with another immigrant - children 5 to 17 years - 1981 Census - male, immigrant or spouse is an immigrant - married or in a common-law relationship - children 5 to 17 years - 1981 Census - male, immigrant - married or in a common-law relationship with another immigrant - children 5 to 17 years - 1981 Census 4. ln weekly earnings for daughters at age 31 5. ln annual earnings for daughters at age 31 FATHERS AND THEIR DAUGHTERS Least Squares Regression Results Daughters Constant Slope Coefficient R2 - female 16 to 65 years both parents immigrants 2001 Census female 16 to 65 years both parents immigrants 2001 Census female 25 to 37 years both parents immigrants 2001 Census 6.86 [0.623] 0.050 [0.092] 0.01 6.57 [0.623] 0.050 [0.092] 0.01 6.715 [0.738] 0.048 [0.108] 0.01 - female 25 to 37 years at least one parent immigrant 2001 Census female 25 to 37 years both parents immigrants 2001 Census 7.13 [0.649] 0.114 [0.095] 0.03 11.1 [0.966] 0.093 [0.091] 0.02 Notes: Earnings are adjusted for age and region as described in the text. The number of observations in all cases is seventy, corresponding to the country of birth of the father. Estimations are based on weighted least squares, with the sum of the number of sons and daughters from each group as the weight. Standard errors are presented in square brackets. All estimates are significant at least at the 5% level. The sample selection rules in row 1 are similar to tho...
View Full Document

Ask a homework question - tutors are online