Value assumptions in integrated assessment models

Kuvaus

Integrated assessment models (IAMs) have in recent decades become a crucial tool in providing knowledge of the potential wider impacts of a warming climate and the outer bounds of climate policy. The third working group of the IPCC bases its claims on the deployment of these models. As did the authors of the recent IPCC report on the 1.5°C target stipulated in the Paris Accord.
Yet the models have been controversial. Among the points of contention one finds long standing disputes over the discount rates, worries about the consequences of simplistic representations of the climate system, and the impacts on model outputs on seemingly arbitrarily constructed damage functions (see e.g. Pindyck 2013; 2018; Stern 2013). More recently how so-called negative emissions technologies are represented in the models and the wider implications of this for how scenarios that are produced using the models are to be assessed has been a point of contention (see e.g. Anderson 2015; Anderson and Peters 2016; Carton 2019; Vaughan and Gough 2016; Smith et al. 2016).
The role and influence of values—social, ethical, cognitive and epistemic—and conflicts and tradeoffs between such values is a central (if not always entirely explicit) theme in this critique and raises important philosophical concerns about the appropriate role of value judgements in integrated assessment modelling and how to adjudicate between different (kinds of) values when they are in conflict.
This paper is an attempt at contributing to a burgeoning discussion on the specific role of values and value assumptions in integrated assessment modelling and the way these models can inform policy and decision making (see e.g. Intemann 2015; Biewald et al. 2015; Kowarsch 2016; Winsberg 2018).
The papers targets the particular context in which IAMs occur. The focus is on three main points. First, there appears to be conflicts that arise out of the integrative and interdisciplinary nature of the models themselves. Here differences in cognitive and epistemic values e.g. guiding what kinds of simplifications are appropriate and how uncertainties are to be dealt with come to the fore. This is considerably complicated by the fact that it is precisely this integrated and interdisciplinary nature of the models that is supposed to account for their epistemic superiority (Rotmans and van Asselt 2001).
Second, the aims, constraints and affordances offered by the modelling context itself easily comes into conflict with the demands of policy- and decision makers that make it hard to convey how the models are to be informative about concrete situations.
And third these aims and uses all taken together are important for how we are to assess the influence of value assumptions in the models. The conventional wisdom in the philosophy of science is that models are primarily tools that can be used in a variety a ways and hence should be assessed according to the degree to which they achieve those aims. Furthermore, inductive risk considerations associated with the stakes associated with the specific decision situations have an influence on how model output is to be (epistemically) evaluated (Douglas 2009). Given the plurality of decision situations and the many different uses of IAMs we should not expect there to be one standard for assessing IAMs.
Finally, the paper contains a discussion on some of the ideas in the literature on how to deal with values in IAMs (and models generally) focusing in particular on the role of transparency and democratic endorsement (Douglas 2009; Elliott 2017; Intemann 2015). An argument is made that the complexity both of the models themselves and the social-epistemic context in which they are developed presents major obstacles to these approaches.

Aikajakso	17 syyskuuta 2019
Tapahtuman otsikko	Ethics and Values in Assessments: Divergent values in sustainability assessments: love them, leave them, or change them?
Tapahtuman tyyppi	Konferenssi
Sijainti	Berlin, Saksa
Tunnustuksen arvo	Kansainvälinen