UC Berkeley

The planet's changing climatology poses epistemological and practical problems for insurance institutions underwriting weather or property risks: models based on meticulously calculated empirical event frequencies will not project risk in a changing climate system. Seeking to explain the unprecedented scale of recent insured losses, media pieces regularly articulate a narrative that links climate change to an immanently insecure future. This logic has prompted some scholars to place climate change in a new category of risks generated by industrial society that are fundamentally incalculable and uninsurable. This dissertation challenges the epistemological assumptions and empirical validity of the "uninsurability hypothesis" using the case study of (re)insurance and catastrophe modeling for North Atlantic tropical cyclones. In so doing, it turns a critical eye on the depoliticized discourse of climate change emergency.

The research analyzes the development of insurance institutions and definitions of climate change risk over time, applying the theory that risks are reconstructed phenomenon of multiple contingency which always embody contested classificatory and causal stories. Research included over forty extended interviews with academic, regulatory, and private sector employees; observation at thirteen industry, academic, and regulatory conferences; and qualitative and quantitative analysis of corporate and regulatory documents and datasets. The findings trace new constellations of science, value, and fear that are emerging within the (re)insurance industry as it attempts to assess and manage climate risks and secure new paths to accumulation. Three major themes emerge.

First, the dynamics of climate change are being integrated into circuits of insurance and financial capital. The perception of climate risk may buoy the (re)industry's business prospects in the short term by reproducing uncertainty and allowing firms to exclude certain risks from all-perils coverage and repackage them into new products. Climate risks may be incorporated into the central contradictory dynamic of the catastrophe reinsurance market, which requires the continual recurrence of catastrophic losses and devaluation in order to sustain pricing and accumulation in the long term. Meanwhile, investment capital is accessing new risk premiums from the insurance sector through catastrophe bonds, the market for which demonstrates a strategic and selective attempt to capture "returns on place" by finance capital, rather than an "escape" from uninsurable places on the part of (re)insurers.

Second, within both the industry and scientific community, the question of how climate change is influencing catastrophic losses or will do so in the future is far from settled, despite its representation as a closed "matter of fact". Furthermore, most (re)insurers do not currently account for climate change in their daily underwriting and pricing, and often cite the possibility of compensating for climate effects through future annual adjustments to prices and policies. This apparent contradiction between discourse and practice is the result of a complex set of institutional, political, and economic factors rather than a systematic attempt to deceive the public or exaggerate risks.

Third, privatized economies of science - and particularly probabilistic catastrophe models - are central tools for climate risk management through (re)insurance markets. The expertise of PhD-credentialed scientists is increasingly used in industry contexts to publicize climate change risks, legitimate moves towards five-year forward-looking catastrophe models, and to commodify climate risks into financial exposures and assets. These findings draw our attention the (re)insurance industry's dependence on the perpetual multiplication of fear and value via technocientific risk identification, and suggest the profound limitations of attempts to manage climate risks and anxieties through market mechanisms.