"In contrast to traditional empirical approaches which attempt to isolate causal relationships between individual variables, systems methods examine the effects of multiple actions on multiple outcomes over time, often through the application of simulation models. Accordingly, they are able to capture the feedbacks and non-linear behavior inherent in complex systems. Without an understanding of such phenomena, urban policies often fail to achieve desired goals over the long-term, and frequently create unanticipated and undesirable secondary effects. For example, floodwater levees reduce transient risk of flooding, but in so doing have multiple unintended consequences: they concentrate water, leading to increased risks from larger-scale floods that exceed levee specifications; they decrease public perception of risk, leading to urban development in risky areas; and they diminish the experience of urban managers with small floods. Each of these factors tends to increase vulnerability."
Quoted from: https://ehjournal.biomedcentral.com/articles/10.1186/s12940-016-0107-2