Search for a command to run...
Flood protection infrastructure can paradoxically increase vulnerability through the ‘levee effect' or the ‘safe development paradox'. In some cases, protection can reduce the perceived risk of flooding, leading to the intensive development of floodplains. This may necessitate additional investment, as short-term rational optimization can create long-term vulnerabilities through path-dependent processes that are difficult to reverse. However, the empirical evidence is limited. This study uses the Shinanogawa River Basin in Japan as a case study to examine this feedback cycle and the dynamics of the coupled human–water system over the centuries. The study reconstructs economic analyses to provide a systematic, long-term, river-basin-wide examination. This retrospective analysis aims to extract lessons that can be applied to current and future flood protection and development planning in the context of climate change and evolving social values, rather than criticizing past decisions in hindsight. The completion of the Okozu Diversion Channel in 1924 reduced flood risk in Nigata City, enabling river reclamation in the 1930s and 1940s. While this project contributed to the development of Nigata City, it also narrowed the river, reducing its flood capacity. Although these reclamation projects appeared profitable at a project level, they created hidden liabilities, necessitating the construction of the Sekiya Diversion Channel in 1972. Together, these interventions resulted in a net system loss of 12 trillion JPY, or 80 billion USD, at maximum. System-level analysis reveals reclamation created a hidden liability requiring compensatory infrastructure costing 365 billion JPY, or 2.4 billion USD (2015 prices, 4% discount rate). This analysis demonstrates how conventional project-level cost-benefit analysis can mask system-wide implications and liabilities. The study traced interventions and social responses, revealing that the temporal mismatch between human decision-making timeframes (years to decades) and the frequency of hydrological events (decades to centuries) was the driving force behind this dynamic. Flood management strategies must consider the long-term by incorporating the uncertainty caused by climate change and the changing environmental and cultural values of water.