We should begin our reply to Paul Warwick by stating how very flattered we are by the diligent attention that he has given to our work on government formation. In particular, we are delighted that he has taken up and applied the simulation technology that we suggested as a way to explore cabinet stability, a matter on which we were beginning to feel like voices wailing in the wilderness.

This is not the place to bore readers of the Journal with detailed points of issue between us, of which there are obviously many. We want here to concentrate upon some larger issues provoked by Warwick's critique. These involve a debate on cabinet stability and duration between a school of thought, in Warwick's corner, that is essentially empiricist in its outlook and a school of thought, in ours, that takes analytical modelling as being more fundamental. This is a theme touched upon by Kaare Strom in his recent review essay on parliamentary democracy.