Micro Econometrics

ECON30342 Micro Econometrics 10 Semester 2 Undergraduate Level 3 Martyn Andrews Unit coordinator Economics 1 FHEQ level (Framework for Higher Education Qualifications) ' Last part of a Bachelors ' European Credit Transfer & Accumulation System Rating : 5.0 This course unit makes use of the basic econometric tools learned in Econ20110 Econometrics and applies them to modern microeconometric issues. Microeconometrics is the application of regression techniques to estimating causal effects using cross-section or panel data sampled from microeconomic agents. The course will be of interest to those studying microeconomics or related areas (e.g. labour economics, public economics, and development economics) and will be useful to students who wish to pursue more advanced applied economics courses later in their career. Anyone writing an undergraduate dissertation in these areas should take this unit.  This course unit makes use of the basic econometric tools learned in Econ20110 Econometrics and applies them to modern microeconometric issues. Microeconometrics is the application of regression techniques to estimating causal effects using cross-section or panel data sampled from microeconomic agents. The course will be of interest to those studying microeconomics or related areas (e.g. labour economics, public economics, and development economics) and will be useful to students who wish to pursue more advanced applied economics courses later in their career. Anyone writing an undergraduate dissertation in these areas should take this unit.  The aim of this course to further develop students' knowledge of modern microeconometric techniques, and so develop the tools required to analyse datasets covering relating to microeconomic units (individuals, households or firms) observed at a single point in time or over time multiple time periods. It focuses on causes of bias/inconsistency, and therefore non-causality, namely endogeneity. These causes are omitted variables, measurement error, simultaneity and self-selection.  After completing this unit, students will:(i) be able to demonstrate a clear understanding of how the regression model is used when analysing cross-section data;(ii) know how to use, and interpret the output from the econometric package Stata;(iii) understand what is meant by a causal effect;(iv) appreciate the various difficulties in estimating causal effects, and understand the way in which econometric methods, in combination with economic models, can address these difficulties when using cross-sectional and panel data. Analytical skills Synthesis and analysis of data and critical reflection and evaluation. Other Computer literacy and the application of theoretical knowledge. Provisional1. Introduction: types of data; panel v (pooled) cross sections.2. Causality and unbiased estimation:<ul><li>Wooldridge's MRL1-4 & multiple regression;</li><li>Simple regression with RHS dummy. The analogy principle (estimation in the population);</li><li>Consistency.</li></ul>3. Endogeneity:<ul><li>Omitted variables bias (OVB), long and short regressions;</li><li>Measurement error (ME);</li><li>Simultaneity.</li></ul>4. Selection bias, treatment effects, and the Potential Outcomes framework.<ul><li>The ATT and ATE.</li></ul>5. Inference (hypothesis testing)<ul><li>Traditional inference (Wooldridge's MLR5 & MLR6);</li><li>Heteroskedasticity and Robust inference;</li><li>Clustering (Moulton formula only).</li></ul>6. Modelling using Multiple regression.<ul><li>Using multi-category dummies (firm-size and earnings) and fitting a quadratic (graduate earnings differential and age)</li></ul>7. Panel data and First Differenced Estimator (FD).8. Policy Analysis and the Difference-in-Difference (DiD) estimator.<ul><li>Panel & Pooled Cross Sections.Testing common Trends. Hetergeneous treatment effects.</li></ul>9. The IV estimator (basics).<ul><li>The simple regression model: one covariate and one instrument. 3 conditions for instrument validity (testable/assumptions). When z is binary (Wald estimator). Inference. Hausman test for endogeneity and weak IVs. Over-identified models (GIVE/2SLS). The Grouped Mean estimator. The Local Average Treatment Effect. </li></ul>10. Regression Discontinuity (the RD estimator). [If time allows] Synchronous activities (such as Lectures or Review and Q&A sessions, and tutorials), and guided self-study 0 Other 100% 20% mid-term assignment (1,000-words) 80% exam, (90 min)   <ul><li>Tutorial feedback.</li></ul> ECON20110 Econometrics Co-Requisite Compulsory Pre-requisites ECON20110ECON20110 Y The main reading is:<ul> <li>Wooldridge Introductory Econometrics, seventh edition (chapters 1-7, 13, 15).</li></ul>Other texts referred to:<ul> <li>Angrist, J. & Pischke, J.-S. (2009), Mostly Harmless Econometrics: An Empiricist's Companion, Princeton University Press, Princeton, NJ.</li> <li>Angrist, J. & Pischke, J.-S. (2015), Mastering 'Metrics: The Path From Cause To Effect, Princeton University Press, Princeton.</li> <li>Stock, J. & Watson, M. (2012), Introduction to Econometrics, third edn, Pearson. Acronym.</li> <li>Verbeek, M. (2017), A Guide to Modern Econometrics, fifth edn, Wiley.</li></ul>PDF files of the slides will be available for download from the course BB site before each lecture. See Blackboard Site for arrangements now there is online teaching only. 0 0 0 For every 10 course unit credits we expect students to work for around 100 hours. This time generally includes any contact times (online or face to face, recorded and live), but also independent study, work for coursework, and group work. This amount is only a guidance and individual study time will vary.