Presentation on theme: "Econometric Analysis of Panel Data"— Presentation transcript:
1 Econometric Analysis of Panel Data IntroductionPanel Data DefinitionUnbalanced PanelBalanced Panel:Short Panel:Long Panel:Panel Data AnalysisUnobserved HeterogeneityCross Section and Time Series CorrelationUsing Stata (stata_intro)
2 Introduction Panel Data Panel Data Analysis Definition (Wikipedia Encyclopedia)Examples of Panel DatasetsPanel Study of Income Dynamics (PSID)Penn World Table (PWT)Panel Data AnalysisA Primer for Panel Data Analysis (Yaffee)
3 Using Stata Declare Panel Data and Variables xtset (or tsset)xttabPanel Data Analysis: xt commandsxtdesxtsumxtdataxtlinePanel Data Regressionxtreg
4 Using Stata Hypothesis Testing Advanced Topics xthausman xttest0 xtregarxthtaylor (Hausman-Taylor Estimator)xtivreg (Instrumental Variables Estimation)xtabond (Arellano-Bond Estimator)
5 Example: Investment Demand Grunfeld and Griliches i = 10 firms: GM, CH, GE, WE, US, AF, DM, GY, UN, IBM; t = 20 years:Iit = Gross investmentFit = Market valueCit = Value of the stock of plant and equipment
6 Example: International Comparison of Economic Growth R. Summers and A. Heston, "The Penn World Table (Mark 5): An Expanded Set of International Comparisons, ," Quarterly Journal of Economics 106, 1991,G. Mankiw, D. Romer, and D. Weil, "A Contribution to the Empirics of Economic Growth,“ Quarterly Journal of Economics 107, 1992,
7 Example: International Comparison of Economic Growth yit = Real per capita GDPsi = Average saving rate (over )ni = Average population growth rate (over )g+d = 5%COMi = 1 if communist, 0 otherwiseOPECi =1 if OPEC, 0 otherwise
8 Example: Returns to Schooling Cornwell and Rupert Data, 595 Individuals, 7 YearsThese data were analyzed in Cornwell, C. and Rupert, P., "Efficient Estimation with Panel Data: An Empirical Comparison of Instrumental Variable Estimators," Journal of Applied Econometrics, 3, 1988, pp See Baltagi, page 122 for further analysis.
9 Example: Returns to Schooling LWAGE = log of wage = dependent variable in regressionsEXP = work experience WKS = weeks worked OCC = occupation, 1 if blue collar, IND = 1 if manufacturing industry SOUTH = 1 if resides in south SMSA = 1 if resides in a city (SMSA) MS = 1 if married FEM = 1 if female UNION = 1 if wage set by union contract ED = years of education BLK = 1 if individual is black
10 Example: Wage Equation Koop and Tobias The data is available in two parts:Part 1: Time-Variant Data (17,919 obs.)id= Person id (ranging from 1 to 2178)ed = Educationlwage = Log of hourly wagepexp = Potential experiencetrend = Time trend
11 Example: Wage Equation Part 2: Time-Invariant Data (2178 individuals)ability = Time invariant abilitymedu = Mother’s educationfedu= Father’s educationd = Dummy variable for residence in a broken homesiblings = Number of siblings.
12 Example: U. S. Productivity Munnell  Productivity Data 48 Continental U.S. States, 17 Years:STATE = State name,ST ABB=State abbreviation,YR =Year, 1970, ,1986,PCAP =Public capital,HWY =Highway capital,WATER =Water utility capital,UTIL =Utility capital,PC =Private capital,GSP =Gross state product,EMP =Employment,