Code:
sum export import gdp distance landlocked RTA lpi0 lpi1 lpi2 lpi3 lpi4 lpi5 lpi6 Variable | Obs Mean Std. Dev. Min Max -------------+--------------------------------------------------------- export | 1,336 828.4553 3016.076 0 41549.71 import | 1,336 832.2992 3950.458 0 58532.57 gdp | 1,336 4506.944 16476.49 1.97454 196236.7 distance | 1,336 8808.857 4342.772 216.4326 19228.99 landlocked | 1,336 .2095808 .4071619 0 1 -------------+--------------------------------------------------------- RTA | 1,336 .0905689 .2871025 0 1 lpi0 | 1,254 2.881038 .5772499 1.598322 4.225967 lpi1 | 1,254 2.693316 .5977922 1.111111 4.20779 lpi2 | 1,254 2.756885 .6817719 1.237654 4.439356 lpi3 | 1,254 2.848861 .5226796 1.362654 4.235 -------------+--------------------------------------------------------- lpi4 | 1,254 2.830826 .6080419 1.394253 4.31065 lpi5 | 1,254 2.886995 .629167 1.513605 4.377678 lpi6 | 1,254 3.256001 .5821565 1.665079 4.795714
1. In my data, the cases in which export and import value is 0 account for 18% of total observation . So is this necessary for me to use other regression not OLS as OLS can make bias estimation ?
2. As I saw in some theses, the authors used PPML,OLS truncated and Heckman with high of zero trade bilateral . I have tried with PPML and run my data as following as :
Code:
ppml export ldis lgdp llpi0 landlocked RTA, cluster(country1)
Thank so much
0 Response to Using PPML, OLS truncated and Heckman for data with high of zero trade bilateral
Post a Comment