○ Multinomial Logit Model (both conditional and unconditional)
○ PSM (as pre-treatment for reducing selection bias)
- Kernel Matching
- Nearest Neighbor Matching(
○ ATT ? Average Treatment effect on
- Robustness Test(hidden bias adjustment)
- The Rosenbaum bound (measuring how big the difference in unobserved factors)
○ expenditure per capita , FGT (headcount / poverty gap) used for outcome variables
○ farm h/h in all 3 schemes (fairtrade, utz, organic) combined have significantly higher living standards than their matched counterparts in non-certified channels, yet no significant impact to poverty.
○ Disaggregation of 3 schemes are important, because impacts may differ considerably among schemes
○ Fairtrade contributes to significant improvements in h/h living standard and reduces the poverty gap.
- Fairtrade guarantees minimum support price(reduces downside risk)
- Fairtrade coops receive a premium(use for investments in infrastructure and training programs)
- Fairtrade coops have more freedom in terms of marketing(own certification documents and sell to any buyer)
- Fairtrade coops sell most of the certified coffee after milling (oUTZ and organic farmers unprocessed form)