Dear colleagues, I am studying a policy of conditional transfers among federated entities. The policy has already been evaluated and has a positive effect on the outcome indicator of municipalities, specifically the scores of students in a large-scale assessment. However, since the increase in education spending was significantly lower than the results achieved, there are reasons to believe that the policy has increased their efficiency.
With this preamble, I intend to measure the causal impact of the policy on the level of efficiency of municipalities. However, I am not sure which methodology to adopt. I have considered two approaches:
1. Two-stage evaluation: first, I calculate the efficiency score (using DEA), and then I use this measure as an outcome indicator in a DiD framework.
2. Simultaneously estimate everything via SFA, using the DiD design to model the mean inefficiency in the maximum likelihood function, taking as a starting point the panel data model by Batesse and Coelli (1995).
Personally, I am more inclined towards the second approach. However, I have not found many references in the literature, so if anyone could guide me, I would be immensely grateful.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS