Course Syllabus: ECONOMETRICS & APPLICATIONS UVM EC 200/SPR 2014/RM: L309 LAFAYE/MWF: 4:05-4:55 PM Prof: John F. Summa, PhD. Office: Old Mill, Rm. 236 Office Hours: Tuesday & Thursday 8:30-9:30 (or by appointment) Phone: (802) 846-7509 (c) E-mail: jfsumma@uvm.edu Required Text: Introduction to Econometrics, Stock and Watson (Pearson, 3rd edition, 2006). Supplemental Readings/Texts: There will be occasional supplemental readings provided by the instructor. These readings will be made available at Blackboard (Bb), or handed out in class. Course Overview & Objectives: Econometric techniques and methods, known as regression analysis, have a wide range of applications in today's world. Essentially, regression analysis allows economists, and other social scientists, to identify and measure causal effects through quantitative analysis. For example, how much does a tax cut increase investment? Or does your skin color or gender have anything to do with your wage level? We need data to answer these questions and econometric techniques to process the data scientifically. This course will equip you with the analytic toolbox required to answer these and many other questions that are relevant to policymakers, as well as lawmakers. Objectives: The course begins with an overview of regression analysis (and a definition of econometrics through representative research questions and cases), and then develops a full understanding of the regression technique known as Ordinary Least Squares (OLS) both single and multivariate forms. Once the key regression properties, coefficient estimation, and model specification are learned, the classical model assumptions are spelled out. This is followed by examination of classical model assumption violations and possible remedies. The seminar course then moves into extensions of the basic model. Students will acquire the ability to undertake their own data-driven research project, a key requirement of the course. See below for more details on course requirements.
Learning Methodology: Class lectures will follow closely the required text. The textbook provides all the essentials for an introduction to the key concepts and practical applications of econometrics. Lectures will drive the learning in combination with weekly econometric data assignments. Assignments will require use of either Gretl, an open source software solution, which is available for free download, and is now installed on classroom computers, or STATA. STATA software is an industry standard and is installed on classroom computers. Online tutorial videos to be provided for STATA. Learning Methodology: No formal training is provided in STATA but basic programming steps will be covered by the instructor. As each chapter lecture is completed, exercises will be assigned directly from the text. Datasets required for exercises will be provided by the instructor if not available from the author's website (see Bb for link to author website and data sources). Exercises are designed to give a hands-on understanding of econometrics and aimed at getting students prepared to do their own research projects (required for completion of the course). Supplemental readings may be assigned periodically to maximize understanding of difficult concepts. Students will be required to present oral reports on regression assignments at intervals specified by the instructor. Research Projects: As part of the course requirements, students must undertake a datadriven research project using regression analysis (econometrics). The instructor will guide the students toward this end, providing basic model structures that students must adapt to datasets and theoretical questions. The instructor will include a set of models for students to work with, but students must be able to incorporate the relevant data to drive the model. Results must be reported and related to theoretical literature. Course Outline: I. Econometric Questions and Data (Chapter 1); pp. 1-11 A. Economic questions we examine B. Causal effects and idealized experiments C. Data Sources and Types
II. Review of Probability (Chapter 2); pp. 15-50 A. Random Variables and Probability Distributions B. Expected Values, Mean, and Variance C. Two Random Variables D. The Normal, Chi-Squared, Student t, and F Distributions E. Random Sampling and the Distribution of the Sample Average (Mean) F. Large-Sample Approximations to Sampling Distributions III. Review of Statistics (Chapter 3); pp. 65-91 A. Estimation of Population Mean B. Hypothesis Tests Concerning the Population Mean C. Confidence Intervals for the Population Mean D. Comparing Means from Different Populations E. Differences of Means Estimation Using Experimental Data F. Using the t-statistic When the sample Size is Small G. Scatter plots, the Sample Covariance, and the Sample Correlation IV. Linear Regression with One Regressor (Chapter 4); pp. 107-131 A. The Linear Regression Model (LRM) B. Estimating the Coefficients of the LRM C. Measures of Fit D. The Least Squares Assumptions E. Sampling Distribution of the OLS Estimators V. Hypothesis Tests and Confidence Intervals (Chapter 5); pp. 144-166 A. Testing Hypotheses About One of the Regression Coefficients B. Confidence Intervals for a Regression Coefficient C. Regression When X is a Binary Variable D. Heteroskadasticity and Homoscedasticity
E. The Theoretical Foundations of Ordinary Least Squares F. Using the t-statistic in Regression When the Sample Size is Small (Regression/research project model selection deadline: 2/5) VI. Linear Regression with Multiple Regressors (Chapter 6); pp. 179-203 A. Omitted Variable Bias B. The Multiple Regression Model C. The OLS Estimator in Multiple Regression D. Measures of Fit in Multiple Regression E. The Least Squares Assumptions in Multiple Regression F. The Distribution of the OLS Estimators in Multiple Regression G. Multicollinearity VII. Hypothesis Tests and Confidence Intervals in Multiple Regression (Chapter 7); pp. 214-240 A. Hypothesis Tests and Confidence Intervals for Single Coefficient B. Tests of Joint Hypotheses C. Testing Single Restrictions Involving Multiple Coefficients D. Confidence Sets for Multiple Coefficients E. Model Specification for Multiple Regression F. Analysis of the Test Score Data Set Midterm Exam: Chapters 1-7: Date TBA VIII. Nonlinear Regression Functions (Chapter 8); pp. 252-294 A. A General Strategy for Modeling Nonlinear Regression Functions B. Nonlinear Functions of a Single Independent Variable C. Interactions Between Independent Variables D. Nonlinear Effects on Test Scores of the Student Teacher Ratio (Regression/research project model data files deadline 3/5)
IX. Assessing Studies Based on Multiple Regression (Chapter 9); pp. 312-339) A. Internal and External Validity B. Threats to Internal Validity of multiple Regression Analysis C. Internal and External Validity When Forecasting with Regressions D. Test Scores and Class Size Example X. Regression with Panel Data (Chapter 10); pp. 347-369 A. Panel Data B. Panel Data with Two Time Periods ('Before and After') C. Fixed Effects Regression D. Regression with Time Fixed Effects E. The Fixed Effects Regression Assumptions and Standard Errors F. Drunk Driving Laws and Traffic Deaths (Regression project regression results and report deadline: 4/2) XI. Regression with a Binary Dependent Variable (Chapter 11); pp. 381-404 A. Binary Dependent Variables and the Linear Probability Model B. Probit and Logit Regression C. Estimation and Inference in the Logit and Probit Models D. Application for the Boston HMDA Data XII. Instrumental Variables Regression (Chapter 12); pp. 419-453 A. The IV Estimator with a Single Regressor and a Single Instrument B. The General IV Regression Model C. Checking Instrument Validity D. Application to the Demand for Cigarettes E. Where Do Valid Instruments Come From? (Three Examples) (Regression/research project paper draft deadline: 4/18)
XIII. Experiments and Quasi-Experiments (Chapter 13); pp. 469-503 A. Potential Outcomes, Causal Effects and Idealized Experiments B. Threats to Validity of Experiments C. Experimental Estimates of the effect of Class Size Reductions D. Quasi-Experiments E. Potential Problems with Quasi-Experiments F. Experimental and Quasi-Experimental Estimates in Heterogeneous Populations (Regression/research project paper final draft deadline: 4/30) XIV. Regression Analysis of Economic Time Series Data (Chapter 14); pp. 517-568 A. Introduction to Time Series Regression and Forecasting B. Auto regressions C. Additional Predictors and the Autoregressive Distributed Lag Model D. Lag Length Selection Using Information Criteria E. Non stationarity I: Trends F. Non stationarity II: Breaks Please note: The instructor reserves the right to alter the course outline and course readings and requirements at any time. [Final exam: May 2, 13:30-16:15, LAFAYE RM L309] Grading Policy: Grades will be based on a midterm exam (20%), a non-cumulative final exam (20%), analytical and empirical exercise assignments (15%), attendance and participation (10%) and regression project and paper (35%). The instructor, without advance notice, may give an occasional quiz (short-answer type) based on assigned readings. It is the responsibility of the student to be prepared for each class in terms of readings and assignments. Any late homework assignments receive a zero (one exception only).
Research Project/Paper: Research project and paper requirements will be handed out during the semester, and will include important deadlines and paper specifications. The instructor encourages students to begin to organize project data as early as possible, since data management -- issues related to data use and availability -- can become very time consuming and difficult to resolve quickly. The more time you have the better. Attendance Expectations: Students are required to attend every class and will be held responsible for material presented in class. Exams will be based on readings/exercises and material presented in lectures, as well as any other content presented in class. If you miss a class, it is your responsibility to acquire the material presented and assigned in that class. Email Policy: The instructor cannot guarantee a timely response to e-mail inquiries/communication (inside of 48 hours), although the instructor does try to respond to all inquiries as quickly as possible. Electronic Devices: No electronic devices are permitted during class time unless pre-arranged with the professor or needed during lab classes. UVM Code of Academic Integrity: Violations of the UVM's Code of Academic Integrity are any acts which would have the effect of unfairly promoting or enhancing one's academic standing within the entire community of learners. Such acts are serious offenses and will not be tolerated. Any suspected violations of the Code will be forwarded to the Center for Student Ethics & Standards. To read the Code of Academic Integrity go to: http://www.uvm.edu/~uvmppg/ppg/student/acadintegrity.pdf. UVM Diversity Statement: The University of Vermont holds that diversity and academic excellence are inseparable. An excellent university, particularly one that is a public land grant, needs to actively seek to provide access to all students who can excel at the institution, without respect to their backgrounds and circumstances, including, among other differences, those of race, color, gender, gender identity or expression, sexual orientation, national and ethnic origin, socio-
economic status, cultural and/or geographic background, religious belief, age, and disability. There is, moreover, a compelling national interest in a higher education sector rich in diversity and opportunity, and a clear state interest in making the educational benefits of this diversity and opportunity accessible to all.