Applied Time Series Analysis for the Social Sciences

REGINA M. BAKER has extensive experience teaching a time series course for the Inter-University Consortium for Political and Social Research summer program, the University of Oregon and the University of Notre Dame.

• Acknowledgments xiAbout the Companion Website xiii1 Introduction 11.1 Why Time Series and Why This Book? 11.2 Time Series: Preliminaries 41.3 Time Series Approaches: Some History, and Outline of the Book 101.4 Summary 162 Foundations 192.1 Multiple Interpretations: Some Intuition 192.2 The Lag Operator, the Difference Operator, and Lag Operator Algebra 222.3 Lag Operator Division and Infinite Series 242.4 Lag Operator Algebra: An Example 262.5 An Aside on Linear Difference Equations 273 Properties of Time Series: Mean and Variance Stationarity 293.1 Stationarity: Formal Definitions 323.2 Mean Non- stationarity: Stochastic Trend Versus Deterministic Trend 333.3 Dickey–Fuller (D–F) Tests 423.4 Unit Root Testing Strategies: Elder–Kennedy’s Simplified Approach 473.5 Transforming Unit Root Series to Achieve Stationarity: Differencing 563.6 Extensions of the Dickey–Fuller Test 583.7 Seasonal Non- stationarity 613.8 Variance Non- stationarity 623.9 Summary 644 Properties of Time Series: Autocorrelation 674.1 Rethinking Autocorrelation 674.2 Modeling Autocorrelation: Wold’s Theorem 694.3 Moving Average Processes 704.4 The Autocorrelation Function (ACF) and Sample Autocorrelation Function (SACF) 724.5 ACFs for MA Processes 764.6 Autoregressive Processes 804.7 The Partial Autocorrelation Function 884.8 Seasonality 904.9 ARMA (mixed) Processes 964.10 Summary 965 Autocorrelation: Univariate ARIMA Estimation and Forecasting 995.1 ARIMA (p,d,q)(P,D,Q) s Notation 1005.2 The ARIMA Model-Building Process 1025.3 Example: Directory Assistance (411) 1065.4 Forecasting 1135.5 Summary 1185.6 Appendix to Chapter 5: Non-linear Models and Numerical Estimation Methods 1186 ARIMA Intervention Models 1216.1 Transfer Functions: General Form 1236.2 Simplest Form: Zero-Order Transfer Functions 1256.3 Gradual Changes: First-Order Transfer Functions 1266.4 Example: The Directory Assistance (411) Series 1307 ARIMA with Continuous Explanatory Variables 1357.1 The Cross-Correlation Function (CCF) 1367.2 Prewhitening 1427.3 Example: Lydia Pinkham Advertising 1437.4 Conclusion 1478 OLS and the Gauss–Markov Assumptions 1498.1 Autocorrelation and its Consequences: A Review 1508.2 Detecting Autocorrelation 1528.3 Generalized Least Squares 1588.4 Limitations of GLS Approaches 1628.5 An Aside: Newey–West Standard Errors 1638.6 Summary 1649 Dynamic Specification: Distributed Lag Models 1679.1 Distributed Lag and Autoregressive Distributed Lag Models 1689.2 The Koyck Model and Dynamic Specification 1719.3 General- to- Specific Modeling and the ADL(1,1) Model 1759.4 Estimating Models with Lagged Dependent Variables 1789.5 Testing Constraints 18010 Regression with Non-stationary Series: Cointegration and Error Correction Models 18310.1 Non-stationary Series and Spurious Regression 18510.2 Cointegration and Cointegration Tests 18710.3 Testing for Cointegration: Two Examples 19010.4 Error Correction Models for Non-stationary Series 19310.5 Summary 19811 The LSE Approach: Encompassing, General-to-Specific Modeling, and Forecasting Success 20111.1 Competing Models of the Consumption Function 20211.2 Additional Criteria: Forecast Success and Parameter Constancy 20411.3 A Sketch of the DHSY Process 20711.4 Summary 21012 A Brief Introduction to Vector Autoregression 21112.1 VAR: Logic and Motivation 21312.2 Estimating VAR Models 21512.3 Summary 220References 220Index 221