The Importance of a Critical Approach to Open Data

Carleton University undergraduate students in Ottawa, Canada, are studying the link between communication and open government. In the class, the students examine how communication can be used to improve governance and to foster a more collaborative relationship between governments and citizens. This series of blog posts is related to a range of topics concerning the issues that challenge open government. These posts will appear on the OGP blog over the coming weeks. We hope you enjoy them.

Open dataBy opening up data and making it sharable and reusable, governments can enable informed debate, better decision making, and the development of innovative new services. Technical specifications: Polici... itself can serve to perpetuate the same barriers which it claims to bridge. The hype around open data, and specifically and the current lack of critical discourse, is not only disingenuous but prevents actors from engaging with the problems of open data, and finding much needed solutions.

In their paper “The New Ambiguity of Open Government”, Harlan Yu and David G. Robinson understand open data—in the context of open government—to mean data that are “accessible in order to promote government transparencyAccording to OGP’s Articles of Governance, transparency occurs when “government-held information (including on activities and decisions) is open, comprehensive, timely, freely available to the pub... More and accountability”. While open data disclosures are generally viewed as a positive component of ideal open government practice, a critical perspective is necessary to fully understand the data and the methods of disclosure.

Introducing Critical Data Studies

Rob Kitchen and Tracey Lauriault provide a useful definition for critical data studies, which they describe as “[exploring] the ways in which data are never simply neutral, objective, independent, raw representations of the world, but are situated, contingent, relational, contextual”. Building off of this theoretical basis, I think the following measures are essential for better understanding and using open data in a way that will provide greater benefits to citizens:

1. Demystification. Promotion of the fact that data are not neutral and are products of biases which they interact with during their creation, collection, dissemination and manipulation.
2. Accessibility. Ensuring that open data projects are accessible. The label of “open” should not apply if accessibility is not at the forefront of a project.
3. Curation. Proper and meaningful curation of data from those with expertise and knowledge about the data involved.
4. Usage. The inclusionOGP participating governments are working to create governments that truly serve all people. Commitments in this area may address persons with disabilities, women and girls, lesbian, gay, bisexual, tr... of tools, applications and visualizations with open data sets.

Open Canada

One statewide open data initiative that illustrates the inherent problems of open data and the need for an integrated critical data studies approach is the Open Canada website. The mere fact that  the open data portal exists and undoubtedly contains valuable information somewhere is an insufficient standard by which to assess the effectiveness of an open data portal. As a university student who has worked with some of the site’s datasets, I can assure you that it is nearly impossible to find a dataset that coincides with my research. Is this intentional? Perhaps not, but the claimed goal of “transparency” goes unmet when there is so much data that nothing can easily be found.

What is most striking about the open data portal is the lack of curation and clear technological barriers that prevent citizens from truly understanding the data. To name an example of one such barrier, the site uses GIS files. The average citizen does not have any use for GIS files, nor would they necessarily know what a GIS file is, what software is required to open it, and how to manipulate the file in that software.

“Raw” Data does not Mean “Neutral” Data

You would not be alone in thinking that “raw” data is synonymous with neutral data. However, if we think about this concept through the lens of critical data studies, data can never truly be “raw”. Data in its rawest form is still the product of various social regimes with which it has interacted throughout its construction. Radical asymmetries of power and technology shape data even at its earliest stages. And just as you should avoid feeding Gordan Ramsay raw food, you should also avoid presenting raw data to non-expert citizens.

This question of who controls the production of data is a very real issue, especially in Canada. The former head of Statistics Canada, Wayne Smith, resigned this past September after Shared Services Canada – the federal government’s centralized Information Technology service – established complete control over the data holdings of Statistics Canada. For Smith, this shift represented an utter loss of independence in Statistics Canada’s operations and rendered its data holdings vulnerable to manipulation not only by Shared Services Canada but also by anybody who influences Shared Services.This is perhaps the most striking example, at least in Canada, that illuminates the need for a critical understanding of open data and their socio-cultural associations in the context of open government.

From an open government perspective, it is not enough to simply have an open data portal and slap the label “transparency” on it. The public also needs easy-to-use tools and informed curation in order to maximize the potential of open data, as well as a critical understanding of the context in which data are produced and disseminated. These steps create a more open form of data, and potentially, a more open government.