Similar to the full country model, R&D lowers the significance of the other variables, and proves to be insignificant. This is likely due to a lack of sufficient data for this variable. The probit model shows similar results, which strengthens the robustness of our results. The difference lies in FeedstockP and PopInAgric, which are significant in the probit model. Similar to the full model, though, we observe that the significances are significantly higher than those in the OLS model, and have inconsistent signs; it therefore likely suffers from the same shortcomings. Oil is statistically significant in column 5, but proves to be negative, which goes against our intuition. Also, adding the Oil variable increases FeedstockP’s statistical significance. We will observe whether we have this puzzling outcome for Oil in our non- OECD model.