Questions1. Causal Inference [15 points]Many microfinance programs are introduced in a staggeredmanner. That is, one group of locationsget the program today, another gets the program in 6months, and so on. Suppose you are asked byyour country’s government to evaluate the impact onpoverty of a microfinance scheme that wasprovided to region A in January 2000, region B in July2000, region C in January 2001, and regionD in June 2001. Suppose that (i) you have data on povertyrates in all four regions in January ofeach year 1999-2002, and (ii) if microfinance has anyimpact on poverty, it will occur within 6months.(a) Describe a difference-in-difference procedure toestimate the impact of microfinance onpoverty rates. What will you measure? When? What will thetreatment group(s) and controlgroup(s) be?(b) Suppose that in January 2000, households in regionsB, C and D were notified that theywould not receive the program until the later scheduleddate. As a result, these householdsbegan to work even harder and earn more income thanbefore they learned about theforthcoming program. This is an example of what type ofbias/effect? Will this effect leadyou to over- or under-estimate the true impact ofmicrofinance access on poverty? Whatwould be one solution for addressing this bias?(c) Suppose that region A is poorer than region B, whichis poorer than region C, which ispoorer than region D. This suggests that policymakerschose to place the program in poorerregions first. This is an example of endogenous programplacement, which can bias ourestimates of program impact. Suppose that you can go backin time and design a randomizedcontrol trial (RCT) to study the impact of microfinance.How would the design compare tothe approach in part (a)?(d) Suppose that region B is adjacent to region A andsome households in region B gain accessto the program as early as January 2000. This is anexample of what type of problem inprogram evaluation? How will this affect the estimatedimpact of the program if we justfocused on the difference-in-difference estimate forregions A and B? [Hint: how will YB1-YB0 look in this case compared to a case where region B isvery far from region A and nohouseholds in B receive the program until July 2000?]2 of 3(e) Suppose that it is now the year 2005, the programexists in every region, and you areinterested in study the impact of microfinance onbusiness creation. Someone notices thatthe microfinance institutions in region B only providecredit offers to households withlandholdings less than 0.5 hectares. How would you usethis rule to estimate the programimpact? Describe the empirical strategy: define thetreatment and control groups, state therelevant assumptions, etc.2. Ethnic Diversity, Conflict, and Income [10 points](a) Suppose there are three ethnic groups in country C:group 1 has 30 people, group 2 has 10people, and group 3 has 50 people. What is the degree ofethnic fractionalization? Nowsuppose country C splits into two countries where countryC1 has 10 people fromgroup 1,10 people from group 2, and 10 people from group 3.Country C2 has the remainingpopulation. Did the splitting create more or lessfractionalized countries compared to theoriginal country C? From the perspective of the Estebanand Ray model, which country (C1or C2) will have more conflict in equilibrium?(b) What is the degree of ethnic fractionalization inyour country? Discuss one way thatfractionalization might (indirectly) affect economicgrowth in the Solow model?(c) Suppose you are interested in understanding theeffects of income on conflict. Someresearchers argue that commodity price shocks are a goodinstrumental variable for income.Others (like your professor) argue that price shocks arenot a good instrument. Explain whatassumptions are required for price shocks to be a goodinstrument in this context and whythese assumptions may not hold.(d) Briefly note any major conflict episodes in yourcountry since 1990. Report the number ofbattle deaths since 1990 as recorded in the WDI.(e) What effect might conflict have on economic growth?Explain through the lens of the Solowmodel supposing that the country is below its steadystate before the conflict. Make adistinction between short-term and long-term effectsfocusing on the potential impacts onthe following variables: savings, population, capital,and technology.3. Demographic Transition [10 points](a) Construct a demographic transition plot for yourcountry with the birth rate, death rate, andpopulation level all on the same graph over the period1960—2010 (there may be missingdata; connect points using straight line; put populationon the right Y axis and birth/deathrates on the left Y axis). Discuss (i) the historicalpatterns, and (ii) potential implications foreconomic growth in the next decade.(b) Now, construct a graph showing the birth rate and theinfant mortality rate. Describe thepatterns and offer a potential explanation for therelationship.3 of 34. Population and Fertility Choice4.1. True or False? [6 points]State whether the statement is true or false [2 points],and briefly explain why [1 point](a) The populations of Europe and North America grew at acombined rate between 1650 and1930 that significantly exceeded the population growthrates of developing countries at thattime.(b) If country A has a population growth rate that islower than country B, then the averagewoman in country A has fewer children than hercounterpart in country B.(c) Total population levels in a country should be lowerat the end of a demographic transitionthan they were at the beginning.4.2. This is a question on joint families, externalities,and fertility choice. Suppose that Joseand Maria are the heads of a nuclear family, making theirfertility decisions. For simplicity,suppose there is no gender bias and no infant or childmortality. The following table detailsthe costs and benefits (in dollars, say) of differentnumbers of children. [6 points](a) Based on the information in the table, how manychildren would Jose and Maria have inorder to maximize their net benefit?# of children Total benefit ($) Additional cost peradditional child ($)One 500 100Two 750 100Three 840 100Four 890 100Five 930 100Six 950 100Seven 960 100Eight 960 100(b) Now consider two identical nuclear families: Jose andMaria (as above), and Jorge andRosa. Jose and Jorge are brothers and the two couples forma joint family. Both coupleshave exactly the same costs and benefits of havingchildren as in the table. Now suppose that50% of the upbringing costs of each child (e.g., childcare) can be passed on to the otherfamily. Each couple makes independent decisions, takingonly its own welfare into account.Now how many children will each couple have?(c) Explain the reason for this seemingly paradoxicalresult, using the concept of externalities,and try to understand why larger families (eitherintegrated across generations or betweensiblings in thesame generation) will tend to have a larger number of children per couple.