This article develops a small-scale evolutionary agent-based model to investigate the interplay between heterogeneous agents, institutions and technological change. By ac-knowledging the concept of behavioural dispositions, we differentiate between changers, neutrals, and deniers. The composition of the population is endogenously determined taking into account that reasoning is context-dependent. As we increase the degree of interaction between agents, a bi-modal distribution with two different basins of attrac-tion emerges: one around an equilibrium with the majority of the population supporting innovative change, and another with most agents being suspicious of innovation. Neutral agents play an important role as an element of resilience. Conditional on their share in equilibrium, an increase in the response of the respective probability functions to growth results in a super-critical Hopf-bifurcation, followed by the emergence of persistent fluctuations. Numerical experiments on the basin of attraction also reveal the birth of an additional periodic attractor. A long-run cycle of technological and institutional change may coexist with locally stable fixed points. Our results indicate that economies are more likely to be path-dependent than what conventional approaches usually admit. (c) 2022 Elsevier B.V. All rights reserved.