* Code Provided by Rainer Winkelmann for Table 8.8 in his textbook with Prof. Boes
* 2-12-16
use MOBFIN.dta, clear
poisson djobs edu exp1 exp1sq union german qualwc ordwc qualbc
cloglog d edu exp1 exp1sq union german qualwc ordwc qualbc
tpoisson djobs edu exp1 exp1sq union german qualwc ordwc qualbc if d==1, ll(0)
* Moving to including only variables examined in Table 8.8 in W&B textbook
poisson djobs edu exp1 exp1sq union german
cloglog djobs edu exp1 exp1sq union german
tpoisson djobs edu exp1 exp1sq union german if d==1, ll(0)
* As Prof. Winkelmann mentioned to me in his e-mail message, he is not sure
* why these estimates do not exactly match up with those printed in Table 8.8
* in the book but then he says:
* "There are some minor differences in the hurdle part which I am not quite sure
* where they come from (I used a different software at the time, this comes from
* my count data book). Also, the regressions need to include occupational status,
* which unfortunately was not mentioned in the discussion of table 8.8"