Skip to main content
. 2020 Dec 7;22(2):229–242. doi: 10.1007/s10198-020-01245-1

Table 2.

Likelihood of Generic Entry—Average Partial Effects

Dep. Variable Generic Entry (1) (2) (3) (4) (5) (6)
Share L.Share Sales Sales Sales Price
Mol/ATC Mol/ATC LOP Quantile Branded
Market Size 0.0349 0.296 0.0669*** 0.00977 0.150
(0.326) (0.404) (0.0171) (0.0332) (0.220)
Quantile2 0.203***
(0.0657)
Quantile3 0.180**
(0.0771)
Quantile4 0.0116
(0.102)
Num 0.226*** 0.271*** 0.166*** 0.180*** 0.180*** 0.223***
(0.0591) (0.0724) (0.0418) (0.0394) (0.0484) (0.0606)
Difficult 0.378*** 0.399*** 0.200*** 0.230** 0.155* 0.392***
(0.127) (0.134) (0.0602) (0.0913) (0.0912) (0.114)
Combination 0.139* 0.181*** 0.0213 0.0269 0.0292 0.148**
(0.0761) (0.0621) (0.0352) (0.0517) (0.0480) (0.0739)
NICE TA Rest -0.0139 -0.0102 -0.00757 -0.0145 -0.0533 -0.0122
(0.0640) (0.0669) (0.0452) (0.0496) (0.0516) (0.0662)
NICE TA Rec 0.163 0.180 0.124* 0.102 0.0360 0.163
(0.141) (0.154) (0.0742) (0.0769) (0.0731) (0.133)
Observations 649 590 649 649 649 643
Time dummies Yes Yes Yes Yes Yes Yes
ATC dummies Yes Yes Yes Yes Yes Yes

Notes: Standard errors in parentheses. We use averaged variables over the year, except for the variable (Share Mol/ATC) for which we use data for the quarter with the highest market share. Using the average over the year provides virtually identical results. Variables Sales LOP, Sales and Price Branded are logged. Reference categories for sales quantiles is the lowest quantile, for difficulty of manufacturing is Easy, and for NICE TA is No TA issued. APEs for the time-varying variables used to parameterise the individual effect are not shown here for parsimony. Significance levels: *** p<0.01, **p<0.05, *p<0.1