Agent-based models, with a history reaching back to the 1940s, have been cited as a useful technique for planning economic development and simulating the effect of economic crashes. These models offer an insightful alternative to the traditional techniques of mathematical modelling. Understanding how different designs of agent-based models change simulation outcomes will be useful for modellers of economic and other simulation scenarios. The work presented here examines how a computer simulation of an agent-based model responds to disruptive events, in the context of an economic model. Agents within the model interact by producing, selling and buying goods. A series of experiments compare system stability in two scenarios: one where a top-down rule is applied to the pricing of goods and another where decision-making is at the individual agent level, a bottom-up approach. These two approaches are termed system-adaptive and self-adaptive. Results draw the conclusion that a self-adaptive function can provide greater stability, but this depends on whether the measured variable is a primary or secondary variable to the adaptive function. Considerations are presented for future work which could consider the impact adaptive functions have on secondary variable measurements.