When risk reduction backfires: a systematic review of the safe development paradox

Hydrological hazards cause significant impacts worldwide. Yet, risk reduction measures (e.g., levees, insurance, and other structural and non-structural interventions) can unintentionally exacerbate the impacts they aim to mitigate. Indeed, when such interventions disregard the complexities of human–water interactions, they can produce adverse outcomes, including the safe development paradox (SDP) and the levee effect (LE), wherein risk-reduction measures paradoxically increase risk by fostering a false safety feeling. Despite growing attention to these socio-hydrological phenomena, empirical evidence remains fragmented.

To consolidate existing knowledge, we reviewed 56 studies published between 2001 and November 2025 that investigated the SDP and LE in specific case studies. Specifically, we analyzed the methodological approaches used, the variables considered, and the extent to which they provided evidence for or against the occurrence of SDP and LE.

Most studies (69.6%) presented conclusive evidence of the SDP or LE through three primary mechanisms: (a) intensified development in protected areas; (b) reduced preparedness and a false safety feeling; and (c) increased damage resulting from rare and extreme events. Only 5.4% of studies reported mitigation or absence of the SDP or LE, highlighting the role of individual preparedness, existing policy frameworks, and risk awareness as potential mitigating factors. Surprisingly, 42.9% of studies focused exclusively on exposure, ignoring vulnerability or behavioral dimensions associated with false safety feeling. This tendency was especially pronounced in the 2024–2025 papers, 68.8% of which considered exposure alone. However, we argue that exposure alone is insufficient to confirm or refute the SDP or LE as it neglects coping capacity, risk perception, and individual adaptation. Consequently, increases in urbanization or population within protected areas cannot, by themselves, confirm the SDP or the LE.

Most studies (44.6%) examined only the effects of structural measures, disregarding the influence of non-structural measures and individual adaptation. Moreover, flood studies dominated, with few articles addressing landslides, mass movement, and other sediment-related hazards.

Therefore, advancing the understanding of these socio-hydrological dynamics requires integrating preparedness, vulnerability, and risk perception into multi-hazard assessments. Furthermore, the role of non-structural measures in generating unintended consequences should be further studies. This comprehensive approach would enable a better understanding of the diversity of scenarios where the SDP and LE can manifest.