University of Wisconsin-Madison Department of Agricultural and Applied Economics Development Economics Preliminary Examination 6 – 10 August 2007 Answer any three of the following five questions. All questions have equal weight. Please type your answers. Your answers must be submitted to Ian Coxhead, Development Prelim Chair, 413 Taylor Hall, no later than 5 pm on Friday, 6 August. 1. The figure below is a stylized representation of the HIV/AIDS epidemic. In South Africa, the HIV prevalence curve seems to have recently peaked at nearly 25% of the population. It is estimated that cumulative AIDS-related deaths will reach 5 million by 2010, with another 1.3 million AIDS-sick individuals. While the economic and other impacts of the epidemic are usually thought to lag the prevalence and AIDS-related death curves (as shown in the figure), this question asks you to consider whether the anticipated effects of the epidemic would be expected to immediately generate economically costly changes in behavior. Epidemic Curves, HIV, AIDS & Impact Numbers HIV prev alence A Impact A2 A1 AIDS cumulativ e B B1 T1 T2 Time a. Thinking of lower income households in a country like South Africa, please briefly describe several realms of behavior that might be expected to change in anticipation of the full effects of the HIV/AIDS epidemic (e.g., at a time like T1 in the figure above). In describing these areas, please reflect on the economic costs (short- and long-term) of the behavioral changes. A good answer to this question will be embedded in the more general literature that analyzes decisionmaking by low-income households. b. Please develop a multi-period model of one of the behaviors you described in part (a.). While you should attempt at least a preliminary analysis of your model, you will be primarily be evaluated on the creativity and potential insight of the model you propose. c. Finally, please describe an empirical strategy that you would use to reliably test the implications of your model. 2. Consider an economy with an average income level of y that is comprised of N people and M discrete income levels (y1, … ym…yM). Let n(ym) denote the number of individuals with income level ym. The Gini coefficient for this economy can be written as: G 1 N N | yi y j | 2 N 2 y i 1 j 1 The Gini coefficient has been the focal point for much analysis of growth and distribution, including the analysis of Alesina and Rodrik (“Distributive Policies and Economic Growth,” Quarterly J of Econ (109)2: 465-490) who econometrically find that high initial income inequality (measured by G) dampens subsequent economic growth using cross-country data. However, Esteban and Ray (“On the Measurement of Polarization”, Econometrica 624 1994, 819-851) suggest that the following polarization measure may be a more informative measure of income distribution: m m P K | yi y j | n( yi ) n( yi ) n( y j ) , i 1 j 1 where K is a normalization term, | yi y j | n( yi ) defines the ‘effective alienation’ between persons with incomes I and j, and the parameter α determines the polarization sensitivity of the measure (note that if α=0, P becomes proportional to G). The question asks you to think about the desirability of polarization measures as a basis for the analysis of economic growth. a. Under what type of income distribution dynamics will measures G and P be most different? That is, when might the two measures evolve in different directions? b. Using a political economy model (e.g., that of Alesina and Rodrik), explore the linkage between income distribution (and levels of G and P) and aggregate rates of economic growth. c. From the theoretical perspective your put forward in part (b.), how do you interpret the econometric results reported by Alesina and Rodrik? 3. In “National policies and economic growth: a reappraisal” (Handbook of Economic Growth, ed. Durlauf and Aghion, 2005) William Easterly rejects the World Bank claim that policy reforms in developing countries can accelerate growth and cut poverty. He argues instead that the statistical significance of associations between national policies and growth results rely on extreme values and moreover, “The [econometric] evidence suggests that macroeconomic policies do not have a significant impact on development after accounting for the impact of institutions.” This view echoes other heterogeneous perspectives, such as in Rodrik, “Growth strategies”, also in Handbook of Economic Growth reflecting a pessimistic view of ‘Washington Consensus’ policy packages. a. Evaluate Easterly’s claim by reference to empirical analyses of economic growth. Is there any means by which to explain or to reconcile this econometric agnosticism with the many theoretical models– including one presented by Easterly himself in the same article – which show direct links between policies and growth? b. Consider the implications of Easterly’s claim for attempts to promote long-run economic development. Are there any robust general policy prescriptions for sustaining growth? Present and defend your answer(s) to this question in the form of advice to national policy makers or to the key decision-makers in a multilateral institution such as the World Bank. 4. Neoclassical economic theory asserts that international migration together with remittance ‘backflows’ must improve welfare in both sending and receiving countries, but that result is of course conditional on many assumptions. This question asks you to focus on the welfare effects of migration in a typical sending country, which you should think of as a low or middle-income developing economy. a. Assume the economy is a price-taker in international markets, is free of distortions and policies, and that all aggregate factor endowments are exogenously determined. Use a simple real general equilibrium model to investigate the static effects of an exogenous outmigration “shock”, in which aggregate remittances from outmigrants are some fraction 0 of their earnings in the pre-migration equilibrium. Evaluate changes in the structure of production and trade, aggregate income, and the welfare of a representative consumer. Do this first for the simplest case of two traded goods, and then for the extension to three goods—importables, exportables and non-tradables. Briefly discuss and compare the results. b. To confirm the logic of your trade-related results, reconcile the effects of outmigration and remittances in a simple statement of the current and capital accounts of the balance of payments. c. It is widely observed that outmigrants from poor countries are not taken randomly from the labor force, but are disproportionately drawn from the upper tails of the educational, skills and entrepreneurship distributions. Extend the supply side of your model to identify the expected impact of outmigration by relatively more productive workers on prospects for long-run growth. Establish a baseline result using standard neoclassical assumptions, then extend the supplyside model or relax its assumptions in ways that might generate endogenous growth outcomes and recompute the effects of the outmigration/remittance experiment. Provide brief but convincing empirical explanations for any amendments you make to the neoclassical model. What are the key parameters governing the long-run results? d. How do the results in part (c) differ from those from the short-run model in part (a)? Discuss structural aspects of the results and supply a brief normative (i.e. policy-oriented) discussion. 5. The Gates Foundation asks you to propose a project to reduce the total number of households in Paraguay with income per household-member below one dollar a day. You (amazingly!) have panel data on demographic characteristics of the people and the households’ income and expenditures. You are given a budget to work with and are told that you can make targeted (non-negative) transfers to households. a. What is your (most simple) proposal? Specify a rule indicating how much money should be given to each household based on data regarding household size and income. b. You have heard that the Gates Foundation is going to show your proposal to a board of expert reviewers including Anthony Atkinson, Michael Carter, Garance Genicot, Dean Karlan, Shelly Lundberg, James Robinson, and Robert Townsend. Summarize what you expect each of these reviewers’ objections to be. c. Suggest seven modifications to your proposal which would address the seven objections you describe above. d. Of course, you may not be able (or want) to implement each of these modifications. Design a modified proposal which you find the most satisfying and discuss why you have chosen that design as your final proposal. e. At the same time that you would like to help both the Gates Foundation, and the people of Paraguay, you would also like to help your own academic career. The Gates Foundation does not want to carry out a randomized experiment. Still, you are sure that there will be some ‘natural experiment’ to emerge when your proposal is implemented. You would like to take advantage of that to write a nifty empirical paper on poverty alleviation. Discuss a natural experiment that might arise, the question you could use it to answer, and the empirical identification strategy you would use to test your hypothesis. Try to think of the objections that other economists might come up with to your identification strategy and give rebuttals to them.