Question

Please I need answer for This question and it is very important and I need solution for this issue just nu.

Q. Assume you have a dataset available that includes information on who is playing in a lottery, who wins a lottery and information on how much these individuals work over several time periods (both before and after the lottery).

a) How would you test whether winning the lottery affects labour supply and hence whether the predictions made hold in the real world? Describe the method you would use and how you would use it.

b) Would you use the full sample of individuals in the data set or only the individuals that in the first place play lotteries?

Answer 1

(a) The relationship can be found at as follows. Take the people who are playing a lottery. now we know that some people are winning and some are losing. Now we can create a dummy variable for the time period in which a person wins a lottery. So for a person who wins we will have value 0 when he does not win and 1 when he wins. For a person who does not win we have all 0s. Thus, we have two groups of data. We can regress labor hours on this dummy and see if the coefficient is negative, reflecting that after winning people reduce the labor hours.

(b) We will only use the individuals that in the first place lotteries. it is because if we include those who do no playa s well, this will introduce a lot of other unobservable variables like risk profile of the people which will be different for different groups. This will lead to endogeneity and regressions will fail to tell us anything.