Introduction to Time Series Regression
email for information on the text
The present introduction to time series regression focuses on the basics of designing and estimating an economic model. The methods are transparent regression analysis that can be done with any statistical software. While this step by step approach is slow, it does lead to understanding the principles of time series regression analysis. Canned programs may be quicker but often miss the finer points that become apparent with the present approach handling of the data.
The book is designed for working with a data set. The way to learn the concepts is to work with estimating an economic time series model. Identify a model and put together a data set. While more observations are always better, quarterly or monthly data may have seasonality that complicate the analysis. Good advice is to find yearly data and at least 40 years of history.
Suppose a model reduces to the effect of exogenous variable Xt on the endogenous variable Yt with an exogenous control variable Zt. The t subscripts refer to the time period. In general functional form the function of interest is
Yt = f(Xt, Zt),
where Xt and Zt may be vectors. The present introduction to time series regression focuses on how to estimate regressions based on this relationship. The issue for economic theory is the significance, sign, and size of the partial derivative effect of Xt on Yt holding Zt constant. Lags of the two independent variables Xt-i and Zt-i can be considered.