Economic and Social Determinants of Infant Mortality in Developing Countries - Assignment Example

Topic: Economic and social determinants of Infant Mortality in Developing countries No offence, but plz write as a non-native speaker. PART This part must have a total of four pages (excluding the bibliography), and it must have two sections: _ Select / describe / justify a research topic (two pages) _ Literature review (two pages) o Select / describe / justify a research topic (two pages) Describe the goal of your research study. For example, envision addressing the following questions: What does the study do? How are you planning to address your question? What kinds of data would you need to answer your question? How are you planning in using the data to address your research question? What kind of public policy recommendations you envision providing to policy makers based on your research findings? Finally, justify why your research topic is an important one. For example, envision submitting your research topic to a government agency for research funding, in which you could address the 5 following questions: Why is research necessary on this topic? How will it help policymakers develop good policies or change existing policies based on the outcomes of this research? How will this research contribute to the growth of scientific knowledge? As you complete this section, please make sure that you describe your research topic, concepts, and questions in terms of economic language. In other words, THINK AND WRITE LIKE AN ECONOMIST by using the tools and terminology that you already learned in past economics courses. For example, where applicable, you should use terms such as “opportunity costs,” “incentives,” “decision at the margin,” “moral hazard,” “hold-up problem,” “externalities,” “technology,” “production function,” “utility,"“substitutability,” “elasticity,” “market power,” etc. o Literature review (two pages) You must provide a literature review of the economic concepts (I stress: economic concepts) The Economics scholarly journals and/or books should be reputable and relatively recent. PART 2_ Data description (two pages) o Data description (two pages) First, provide an overview of the data that you will be using to conduct your econometric study. Questions that you should address are: _ What are your data? _ What type of data are they? (For example: Cross-section, panel, time-series?) _ What is the source of your data? _ How did the data provider collect the data? (For example: Random sampling, stratified sampling, telephone interviews?) _ What is the underlying “target population” of your data? _ What is the “unit of observation” in the data? (For example: If you are studying the relation between crime and punishment at the city level, and you have data on the crime rate and punishment rate for many different cities across the U.S., then your unit of observation is “city.”) _ What are the strengths and weaknesses of the data? How will these strengths/weaknesses affect the ability to successfully address your research question? Next, provide useful descriptive statistics and charts of your data; then, briefly discuss stylized facts about your key variables. This can be accomplished by displaying and discussing some of the following: 7 Table of descriptive statistics _ Histograms _ Box plots _ Scatter plots _ Table of correlations _ Etc. PART 3: This part must have four pages (excluding tables, figures, and bibliography), and it must have two sections: _ Hypothesis to be tested and econometric model (two pages) _ Empirical findings, robustness checks, and policy implications (two pages) o Hypothesis to be tested and econometric model (two pages) As part of your research topic, at this stage you should have already established a key dependent variable, “Y,” and some key independent variables, “X”s, that affect the Y. For example, if you are studying the determinants of crime rate in inner cities, then Y may be potentially defined as the “number of crimes per 10,000 residents” in inner cities (i.e., the “crime rate”), and X may contain, among others, “spending per capita on police officers,” “income per capita,” “median age of population,” and “unemployment rate” in inner cities. Your research topic from Part 1 may postulate, for example, that a particular independent variable, say “spending per capita on police officers,” has a (causal) impact on the “crime rate.” This is an example of a hypothesis. For example: “My hypothesis is that “spending per capita on police officers” reduces the “crime rate.” In other words, when “spending per capita on police officers” increases by a positive amount, one should expect that “crime rate” will decrease. This is because … (etc etc)… According to economic theory, this is a very plausible assumption (see for example, Cooper (2009)), and one that needs to be empirically tested… Furthermore, I believe that “spending per capita on police officers” affects “crime rate” in ways that have not yet been pointed out by the literature, such as … (etc etc)… Etc…” “At the same time, it may be also the case that “median age of population” also affects “crime rate”, which means that I should also incorporate the variable “median age of population” as a control variable in my analysis. Otherwise, I may be neglecting an important determinant of “crime rate” in my analysis, 8 which may bias my estimates of the regression coefficient on “spending per capita on police officers” in my regression, etc…” You must also state how your hypothesis fits into the existing literature that you discussed in Part 1. Additionally, you should clearly describe how your hypothesis is novel in the sense that it has not yet been analyzed by previous researchers. For example: “The economic literature on my research topic currently provides no clear answer as to whether and how “spending per capita on police officers” affects “crime rate,” and whether this effect (if existent) has remained constant over time. For example, many authors, such as Smith (2004) and O’Brien (1998), have established a clear link between “income per capita” and “crime rate,” and other authors, such as Taylor (2001) and Doe (1992), have even analyzed the relation between “unemployment rate” and “crime rate,” but the answer as to whether “spending per capita on police officers” and “crime rate” are related in any shape or form remains to be addressed... Etc…” “As such, my research tackles a potentially novel relation that may provide new insights to the literature and to policy makers. Etc…” Please note that if the relation between “spending per capita on police officers” and “crime rate” has already been studied within a particular context, it does not mean that you cannot study this relation within a different context. For example: “Many studies of the relation between “spending per capita on police officers” and “crime rate” have shown that “spending per capita on police officers” reduces the “crime rate” in inner cities. For example, see the findings of Porter (2008) and Malley (1999). However, there have been no studies on the potential effect of “spending per capita on police officers” on “crime rates” in rural areas… Etc…” “My research topic, thus, will make a unique contribution to the literature by analyzing the relation for rural areas. Rural areas are very relevant in these particular markets, because … (etc etc)… Therefore, there are potential important policy recommendations to be made … (etc etc)…” Next, you must specify the econometric model that you will use to test your hypothesis. This may be accomplished by providing the following: _ Describe and write down the econometric equations you will use to test your hypothesis. _ Describe the estimation method you will use to test your hypothesis. For example, will you be using OLS regressions? Or, fixed-effect panel data regressions? Or, a difference-in-difference panel data estimator? Or, an instrumental variable method (if so, then describe your instruments)? Or, a binary discrete-choice model? Etc… _ What kind of econometric issues you envision encountering. For example, will your estimates likely suffer from “omitted variable bias” (and, if so, what are the potential omitted variables)? Or, will your estimates likely suffer from “high but imperfect multicollinearity” and therefore will be imprecise? Or, will they likely suffer from “simultaneity bias”? Are there non-linearities in the relation between your variables of interest (e.g., should you use a quadratic specification, or the log some the variables)? Etc… _ Are you applying any transformation (e.g., taking logs, shares, polynomials) to any of your variables? If so, why? 9 o Empirical findings, econometric issues, and policy implications (two pages) Display and discuss your empirical findings. Do they support your hypothesis? Why or why not? Do they support economic theory? Are your results “statistically” significant? Are your results of “practical” relevance? It is highly recommended that you display the empirical results from your regressions in “side-by-side” tables; that is, you should summarize your econometric findings in side-by-side tables that display different regression specifications in a side-by-side manner. What are some of the econometric issues with your model? How did you tackle them? For example, do your estimates still suffer from omitted variable bias? If so, what are some of the candidate omitted variables? Can you provide information on the likely “direction” / “sign” of the bias? For example, is it likely that your current estimates have a “downward” or an “upward” bias? Why? Or, do your estimates suffer from simultaneity bias? Or, do your estimates suffer from “high but imperfect multicollinearity”? What have you done to diagnose and address any of these potential issues? How do your findings compare to findings from the existing literature? And, equally important, what are the policy implications of your empirical findings? What type of policy recommendations can you provide to policy makers based on your findings? As you describe your econometric model, empirical findings, and econometric issues, you should keep in mind the econometric techniques that you learned during your econometrics course (if you took one) and during the econometrics lectures for this course: _ OLS multivariate regression (and the search for “causality”) _ Hypothesis testing (t-stat, F-test) _ Adjusted R2 (and its limitations) _ Heteroskedasticity-robust standard errors (“robust”) _ Multicollinearity (and its “symptoms”) _ Omitted variable bias _ VIF _ Practical versus statistical significance _ Detecting and handling “outlier” data points _ Standardized variables _ Dummy variables _ Interaction variables _ Log variables _ Polynomial regressions _ Pooled OLS with panel data _ Fixed effects with panel data _ Time effects in panel data _ Difference-in-difference panel data estimator _ Clustered standard errors _ Instrumental variables and “valid” instruments _ Simultaneous causality bias _ Discrete choice estimators _ Censoring, truncation, and sample selection methods Read More