Doing data differently: why data standards are political beasts

by Tom Orrell

“‘All right’, said Deep Thought. ‘The Answer to the Great Question…’


‘Of Life, the Universe and Everything…’ said Deep Thought.




‘Forty-two,’ said Deep Thought, with infinite majesty and calm”.

Douglas Adams, The Hitchhikers Guide to the Galaxy

Devoid of context, numbers have no meaning; that’s the message that the above excerpt imparts. It’s a message that rings true when it comes to the use of data in the development sector too. At the Joined-Up Data Standards project that I work on together with colleagues from Development Initiatives, we work to contextualise data within the development sector by seeking ways to make data standards interoperable with each other and  over the course of the past year we’ve produced a series of discussion papers that explore particular interoperability challenges.

It’s always been an underlying assumption within the project that interoperability challenges are political as well as technical in nature and over the course of the last year, I’ve noticed an overlap between the work we do and what the Doing Development Differently (DDD) community seeks to achieve.

The DDD Manifesto acknowledges that “genuine development progress is complex: solutions are not simple or obvious, those who would benefit most lack power, those who can make a difference are disengaged and political barriers are too often overlooked.” I believe that this applies to the process of how data standards used in the development sector are set. We need more critical analysis of the policies and processes that lie behind data standard development to examine whether they actually benefit those who ‘lack power’. At the very least, we need to understand the biases and politics that exist in the information systems that we increasingly rely on for the data and information we use as ‘evidence’.

There are three interrelated observations that I think are key to beginning to understand and identify where and how ‘political barriers’ materialise in the data standards field:

‘Data silos’ don’t build themselves, humans construct them

Firstly, I often hear the term ‘data silo’ used to describe data published to standards that aren’t interoperable with one another. Silos emerge when standard-setting institutions (including various UN agencies, the World Bank, IMF and others) produce standards without consideration of what data is already out there. There are legitimate historical reasons for why this happens but as we move towards a more data-driven approach to development, it’s a practice that needs to change. This is a basic point but one that is often overlooked and the implication is clear – in order to break down ‘data silos’ we need to first open up the discussions and decision-making processes within these institutions. This is a much harder and more nuanced political challenge that will take time to achieve.

Data standards reflect the biases of the institutions that develop them 

Linked to the above, another point that’s obvious but oft overlooked is that data standards reflect the biases of the institutions that produce them. The World Bank’s Global Data Editor sums it up nicely: “while classification schemes are convenient for analysis and communication, every one comes with a set of limitations, biases and cultural overtones.” These biases need to be properly understood and acknowledged when they arise if data standards that serve the needs of those who lack power are to be developed. For instance, we need to push for the standards that lie behind the SDG target and indicator frameworks to deliver the information that developing countries themselves need, rather than spend a fortune developing the tools and capacity needed to report on the SDGs for their own sake, an end that serves international institutions rather than developing states – a point we recently argued in another blog.

People who set data standards are not the people who publish data

Finally, there’s sometimes a disconnect between the conversations that take place between people who develop data standards and people who publish data – sometimes within single institutions or governments. In the UK for instance, the Cabinet Office leads on engagement with the Open Government Partnership, where discussions on how to meet the data needs of the SDGs regularly occur. At the same time, the Office for National Statistics is responsible for compiling data on the UK’s own progress towards meeting the SDGs and the Department for International Development provides assistance to other states to meet their SDG commitments. These three different organs of government complement each other’s work around similar themes but they are involved in very different international processes. There needs to be greater coordination between international processes where different stakeholders meet. Unless this happens, inevitably, new silos of policy-makers will form, bringing us back round to the first point.

In sum, those of us within the DDD community who work with ‘data’ and digital technologies as tools for development need to be aware of these issues if we want to make sure that new sources of, and ways of collecting, data don’t inadvertently adversely impact those who lack power.












Leave a Reply

Your email address will not be published. Required fields are marked *