The Kalman filter is used to derive updating equations for the Bayesia
n data density in discrete time linear regression models with stochast
ic regressors. The implied ''Bayes model'' has time varying parameters
and conditionally heterogeneous error variances. A sigma-finite Bayes
model measure is given and used to produce a new-model-selection crit
erion (PIC) and objective posterior odds tests for sharp null hypothes
es like the presence of a unit root. This extends earlier work by Phil
lips and Ploberger [18]. Autoregressive-moving average (ARMA) models a
re considered, and a general test of trend-stationarity versus differe
nce-stationarity is developed in ARMA models that allow for automatic
order selection of the stochastic regressors and the degree of the det
erministic trend. The tests are completely consistent in that both typ
e I and type II errors tend to zero as the sample size tends to infini
ty. Simulation results and an empirical application are reported. The
simulations show that the PIC works very well and is generally superio
r to the Schwarz BIC criterion, even in stationary systems. Empirical
application of our methods to the Nelson-Plosser [11] series show that
three series (unemployment, industrial production, and the money stoc
k) are level- or trend-stationary. The other eleven series are found t
o be stochastically nonstationary.