'Classical' econometric theory assumes that observed data come from a stati
onary process, where means and variances are constant over time. Graphs of
economic time series, and the historical record of economic forecasting, re
veal the invalidity of such an assumption. Consequently, we discuss the imp
ortance of stationarity for empirical modeling and inference; describe the
effects of incorrectly assuming stationarity; explain the basic concepts of
nonstationarity; note some sources of non-stationarity; formulate a class
of nonstationary processes (autoregressions with unit roots) that seem empi
rically relevant for analyzing economic time series; and show when an analy
sis can be transformed by means of differencing and cointegrating combinati
ons so stationarity becomes a reasonable assumption. We then describe how t
o test for unit roots and cointegration. Monte Carlo simulations and empiri
cal examples illustrate the analysis.