I have a doubt in understanding the concept of input and output models in DEA.
I want to assess the efficiency of the states with respect to number of accidents and Strict rules (e.g., heavy fine).
What would be the input and out put variables in this context?
Could anyone suggest me which DEA approach would be best fit to assess state wise efficiency.
The implications may suggest to minimize the death rate by maximizing the impositions of rules (may be followed by the frontiers).
In DEA, you might enter accidents (in reverse scale) as Output, and rules as Input. Thus, each state has an objective of minimizing the number of accidents (maximize the inverse of accidents) subject to rules. If 2 states run the same rules, a state with less accidents can be considered as more efficient (output technical efficiency concept); similarly, if 2 states have the same number of accidents, state with less strict rules could be considered as more efficient (input technical efficiency concept).
Generally, the outputs are those variables you want to have more in the system, something for which the higher values are desired, such as the reduction in the number of deaths or the reductions in the number of accidents. The inputs are those variables for which the lower values are desired, such as the total cost of the system or the total number of staff necessary for the system.
You should determine first the input and output variables for your study based on literature.
Next, the CRS and VRS frontier of DEA analysis generally defined by assumption you make for DMUs (States), eg. If you assume states are homogeneous then choose CRS (CCR) model. I recommend you to go for both scales and see the difference in efficiency scores then decide which scale can be more specific.
Next, the orientation you can choose considering which variable (input or output) can be more control by DMUs.
- Badges
- Science topic
- Similar topics
- Econ
- Manufacturing
- Industry