This paper investigates inefficiency/merger dynamics in the English higher education sector from 1996–97 to 2008–9. Merging can lead to greater efficiency, and this is the motivation for encouraging merger in the English higher education sector in a period of austerity. But inefficiency can also contribute towards the decision to merge, meaning that the relationship is two-way. Although there is literature on the effects of merging in higher education, the dynamics are typically not modelled. This paper examines the differences in efficiency between pre-merging, post-merging and non-merging universities, and explores the evolution of the effects over time following the merger. We develop a dynamic model of inefficiency and merger the estimation of which relies on Bayesian techniques organized around the use of Markov chain Monte Carlo. We find that typically merger delivers efficiency gains in the first instance, but these plateau soon after merger. Moreover, there is a wide dispersion around mean efficiency and there is a substantial group of mergers where the probability of efficiency improvement is relatively low. The policy implication is that merging universities is not a universal solution to improving their efficiency.