Question: What if you randomly assigned half the youth on the list to a treatment group who was offered the opportunity to participate in the program? If you compared the average income after the program among those offered the opportunity to participate to the average income among those randomly assigned to the control group, what would your estimate of the treatment effect of the program be?

(e) Suppose that the true impact of the program is to increase participant's income by 2 dollars per week. If you compared the average income (after the program) among those who participate to the average income of those who do not participate, what would the difference be? (So, this question is talking about the difference in actual observed incomes, not incomes in the absence of the program.)

Here is the context to the question:

You are interested in evaluating a program that offers entrepreneurship training to unemployed youth. The non-governmental organization (NGO) implementing the program has compiled a list of unemployed youth who might be interested in the program. To assess the impact of the program, they suggest comparing the incomes of youth who enroll in the training program to the incomes of youth who were not interested in the program. You learn that 50 percent of the unemployed youth on the NGO's list of potential participants come from poor households; the remainder come from non-poor households. Unemployed youth from poor households are typically unable to find even short-term, casual work; so all the poor youth on the NGO's list are interested in the training program. However, non-poor youth can often earn money doing casual work for their parents or relatives; as a result, only half of the non-poor youth on the NGO's list are interested in the program.