1: \begin{abstract}

2:   A Markov chain is geometrically ergodic if it converges to its

3:   invariant distribution at a geometric rate in total variation norm.

4:   We study geometric ergodicity of deterministic and random scan

5:   versions of the two-variable Gibbs sampler.  We give a sufficient

6:   condition which simultaneously guarantees both versions are

7:   geometrically ergodic.  We also develop a method for simultaneously

8:   establishing that both versions are subgeometrically ergodic.  These

9:   general results allow us to characterize the convergence rate of

10:   two-variable Gibbs samplers in a particular family of discrete

11:   bivariate distributions.

12: \end{abstract}

13: