The critical level of hidden bias is reported in Table 3. The critical
level of bias ranges from 1.05–1.10 to a maximum of 1.45–1.50. The
value of 1.50 indicates that participants and non-participants differ in
their odds of participation up to 50%. The numbers of treated and number
of control are also reported in the Table 3. The results are in linewith
previous studies such as (Ali and Abdulai, 2010; Ali and Olaf, 2013).
In Bhutan Community Forestry Programhas been in place for several
years and the result from ATT (Appendix 1) shows there are some differences However, most importantly there are significant differences between
the participating and non-participating households (Fig. 3).
Themain purpose of the propensity score is to balance the covariates
before and after matching and for that a number of balancing tests are
employed to check the matching quality, like median absolute bias before
and after matching along with percentage bias reduction. The results
are presented in Table 4. The bias before matching is quite high
in all four matching algorithms i.e. nearest neighbor matching, kernelbased
among the old and new members in terms of benefits i.e. longer
the duration of participation more are the benefits and vice versa