Rethinking Digitalization and Climate: Don't Predict, Mitigate

TL;DR

This paper challenges the focus on predicting digitalization's climate impact, arguing that digital carbon footprints are inherently unpredictable and emphasizing mitigation strategies instead.

Contribution

It critiques current predictive approaches and advocates for shifting from prediction to mitigation in managing digitalization's climate effects.

Findings

Digital carbon footprints are inherently unpredictable.

Prediction-based assessments rely on flawed assumptions.

Mitigation offers a more viable approach to address digitalization's climate impact.

Abstract

Digitalization is a core component of the green transition. Today's focus is on quantifying and pre-dicting the climate effects of digitalization through various life-cycle assessments and baseline sce-nario methodologies. Here we argue that this is a mistake. Most attempts at prediction are based on three implicit assumptions: (a) the digital carbon footprint can be quantified, (b) business-as-usual with episodic change leading to a new era of stability, and (c) investments in digitalization will be delivered within the cost, timeframe, and benefits described in their business cases. We problema-tize each assumption within the context of digitalization and argue that the digital carbon footprint is inherently unpredictable. We build on uncertainty literature to show that even if you cannot predict, you can still mitigate. On that basis, we propose to rethink practice on the digital carbon footprint from prediction to mitigation.