#Effects of censorship on knowledge and % of heretic authors# 
Wihout censorship, knowledge quality in t=5 would have been: 43.35293 % higher 
With cenosorship, the share of heretic authors in t=2 is: 51.54841 
With cenosorship, the share of heretic authors in t=5 is: 28.65946 
 
#Decomposition of the effects of censorship (Equation 13)# 
The overall drop in quality in t=5 is:  1.818487 
The within drop  in quality in t=5 (a) is:  1.24495 , in %  is 68.46073 
The across drop  in quality in t=5 (b) is:  1.501389 , in %  is 82.56251 
The interaction drop in quality in t=5 (c) is:  -0.9278509 , in %  is -51.02323 
 
The bounds of the fixed cost of imposing censorship psi (appendix F2) are:  1.029262 1.032387 
 
The difference in quality between censored and non censored authors in t=2 is:  1.779005 
 
The difference in quality between censored and non censored authors in t=5 is:  1.289272 
 
#Macro shocks caused a decrease in average quality of 11  % in t=5 
#If macro shocks were constant over time, knowledge quality (wrt baseline) would increase by:# 
15 % in t=4; 
12 %  in t=5; 
