preview

Propensity Score Matching

Decent Essays

Propensity Score Matching In order to identify the impact of the community based health insurance scheme, the Propensity Score Matching (PSM) method was implemented. Individuals were divided into two groups, the treatment group; those enrolled in the CBHI and the control group; those not enrolled in the CBHI scheme. The propensity score is the conditional probability of participating in the treatment group given certain observable variables (Imbensand Wooldridge, 2008). For each individual participating to the treatment the propensity score matching identifies a matched non-participant. The individual from the control group is then matched with a treated individual that is closest in terms of the propensity score, a method which is called Nearest Neighbour Matching. In order to do that, a probit regression was used to estimate the participation decision. The treatment effect, where treatment is binary and defined as having CBHI with Di being equal to one if the individual i is in the treatment group and zero otherwise, is given by …show more content…

However, since it is not possible to observe the outcome of the same individual in both treatment and control states at the same time, i.e. counterfactual problem or missing data problem, according to Holland (1986), estimating the treatment effect by using (1) is impossible. An alternative solution would be to estimate the average treatment effect (ATE) or/and the average treatment effect on the treated (ATT). The (population) average treatment effect is the difference between the expected outcomes of participants and non-participants ƮATE = E (Ʈ) = E[Y (1) – Y (0)]

Get Access