Continue with the program described in Q4. Suppose instead the government decides to hire independent Economic Consultant to assess the effectiveness of the program. The Consultant decides to run an RCT in a state to figure out the causal impact of the program. She suggests to randomly assign 1/2 of the counties of the state to the treatment group and rest in control group, though individuals are allowed to freely move across the counties. At the end of two years, wages of the individuals in treated counties will be compared with those in control counties.

is the unit of randomization in this experiment. Moreover, this design solve the selection issues that may affect the causal estimate of the effectiveness of the program.