“We need more data”! The politics of scientific information for water governance in the context of hydraulic fracturing

• Published in: Water alternatives 2018-02, Vol.11 (1), p.142-162
• Main Authors: Michele-Lee, Moore, Karena Shaw and Heather Castleden
• Format: Article
• Language: English
• Subjects: accountability; Canada; Data; Data collection; Data management; Governance; Hydraulic fracturing; Indigenous knowledge; Indigenous peoples; Indigenous Peoples' knowledge; Knowledge management; Literature reviews; neoliberalism; science policy; Transparency (optical); Two-Eyed Seeing; Water consumption; Water governance; Water use
• ISSN: 1965-0175; 1965-0175

Proposed and actual developments of hydraulic fracturing, as a high-volume water user, have proven contentious in recent years. However, one point of agreement has emerged amongst all actors with regards to water use and hydraulic fracturing: we need more data. This consensus fits with a longstanding reification of the role of data in water governance, and yet we argue it hides a politically contested terrain. Based on a literature review, an empirical Delphi study and a workshop with a diverse array of participants from across Canada, we explore the data needs related to water governance and hydraulic fracturing. We then investigate three areas of deficiency that point to a lack of trust and oversight as well as the exclusion of community and Indigenous knowledge. We argue that in an era of neoliberal approaches to water governance, issues of trust, accountability and transparency all link back to a diminished role for data management within existing water governance arrangements. The challenge is that simply collecting more data will not help decision-makers navigate the complexity of water governance. Our findings suggest a growing call by participants for greater engagement by governments in data collection and knowledge management, new funding mechanisms for data collection and re-thinking how and what to monitor if including multiple ways of knowing and values.