Fatal Choices : Public Administration & Disaster Vulnerability

Abstract

One major way that humans interact with their environment is through land use and development. As communities grow, the conversion of farms or natural areas to residential or industrial land leads to denser development. Planning experts have long advocated recognizing the consequences of such land use changes on the natural environment and local communities. Land use changes can increase disaster risk from natural hazards inherently connected to the nature of a space, such as flooding, earthquakes or wildfires. Interactions are particularly dynamic at the urban fringe, with populations newly moving into hazardous areas. Land use and development controls are therefore important tools for local governments to reduce vulnerability in these communities, an objective that should be prioritised to ensure public safety. Yet, when looking at what happens on the ground, failures do occur. The dilemma that inspired this dissertation is the puzzling behaviour of bureaucratic actors in cases where they do not implement land use and development controls to address vulnerabilities against known natural hazards despite having access to tools that would allow them to do so. The approach here is based on the assumption that there are social mechanisms connecting the organisational context of public organisations to the behaviour of bureaucratic actors (Hedström & Ylikoski, 2010, p. 61), which in some cases have unintended negative consequences. More specifically, public actors are assumed to be sensitive to perceived threats to the reputation of their organisation and defending it can, in circumstances perceived as dangerous, supersede other priorities like public safety (Carpenter & Krause, 2012). This perspective has not received much attention by researchers interested in the failures of disaster prevention efforts or in the failures of public organisations. The causal mechanisms that explain how and when political considerations change organisational behaviour and contribute to failure remain unclear. As land use and development management issues touch on basic rights and will become more contentious in the context of climate change mitigation and adaptation (Sager & Hinterleitner, 2022, p. 107), this topic is worth investigating more thoroughly. In this dissertation, three singular case studies are analysed with the help of causal process-tracing (Blatter & Haverland, 2012) to achieve two main goals: First, the cases that were selected allow for expanding knowledge regarding the impact of reputational threats on public bureaucratic actors’ organisational behaviour. An institutionalist approach is particularly useful for this purpose. This is first critically assessed by looking at the least-likely case (Busby & Bennett, 2008) of the collapse of public housing in Amatrice, Central Italy, during an earthquake in 2016, which had disastrous consequences despite the existence of specific strict seismic building codes. Assumptions about the incentives that contributed to omissive behaviour are then further fleshed out in two additional cases following an abductive approach. In the second case – on prevention efforts to reduce land use vulnerability against bushfires in Murrindindi Shire, Melbourne, Australia, before 2009 – the focus is on determining the causal reactive and anticipatory threat mechanisms that explain how reputational threats lead to omissive behaviour. The considerations of bureaucratic actors regarding the way potential implementation actions could impact their relations with relevant audiences is identified as an important explanatory factor for omissive behaviour. The robustness of the mechanisms is then tested in a third case study of public efforts to reduce land use vulnerability to flooding in the Cypress area, Houston, United States, before 2017, which allows for broader comprehension of relevant contextual conditions. Second, cases were selected where unsafe living conditions, in terms of high vulnerability due to land use, directly contributed to the fatal disaster consequences (the Central Italy earthquake in 2016, the Black Saturday bushfires in Melbourne in 2009 and Hurricane Harvey in 2017 in Houston). This makes it possible to clearly define the organisational failure in vulnerability reduction by means of planning that contributed to disaster impact in each of the cases. Thus, another goal can be achieved: increasing the amount of systematically collected evidence into the causes for public organisational failure in disaster prevention and how they related to each other to produce failure. The Murrindindi and Cypress cases are particularly instructive in this regard. The integrative research design enables fitting the puzzle pieces of case-specific and generalizable findings into a single picture about the social mechanisms impacting bureaucratic actors that implement disaster prevention policies by means of land use and development controls.