Greene (2003 p. 768) remarks that the ML estimates β j of the regression coefficients in a censored-regression model are often approximately equal to the OLS estimates B j divided by the proportion P...

Greene (2003 p. 768) remarks that the ML estimates β_j
of the regression coefficients in a censored-regression model are often approximately equal to the OLS estimates B_j
divided by the proportion P of uncensored observations; that is, β_j
= B_j=P. Does this pattern hold for the hours-worked regression in Table 20.7 (page 639), where P = :69?