ODI Logo ODI

Trending

What we do

Search

Newsletter

Follow ODI

Interview: interrogating the power relationships that shape data for sustainable development

Written by Louise Shaxson, Vissého Adjiwanou, Tom Moultrie

Quito, Ecuador, host city of the UN's Habitat III, 2017.

Assessing progress towards the Sustainable Development Goals (SDGs) requires data, and lots of it. When the SDGs were designed in 2015, they were accompanied by the promise of an explosion in the amount and types of data available: a ‘data revolution’. The 2014 report of the Independent Expert Advisory Group (IEAG) on Sustainable Development which advised the then UN Secretary General identified two global challenges to this revolution:

  1. Gaps in knowledge stemming from limitations in digital and traditional datasets.
  2. Disparities between those with and without the data they need.

It called for increased innovation, leadership and coordination to mobilise resources to overcome inequalities between countries and between data poor and data rich people.

What was less understood then was the radical change happening in the big data ecosystem because of the changes in the 3‘C’s:

  • digital crumbs emitted from digital devices and phones
  • the capabilities to uncover patterns in this data
  • the combinations of communities (not just researchers and policy-makers but civil society organisations and the private sector) who would use and interpret the data.

Demographers in particular sounded a cautionary note around the data revolution.

While big data brings new opportunities, traditional forms of data such as censuses are tried and tested, particularly to ensure that the poor and vulnerable are counted.

They also highlighted the need to combine digital and traditional datasets, and new data science techniques with traditional methods to understand population structures and dynamics, their causes and consequences.

Five years on from the creation of the SDGs, Louise Shaxson sat down with Professor Tom Moultrie of the University of Cape Town and Professor Vissého Adjiwanou of the University of Québec in Montréal to ask what progress has been made and where the global data community stands on addressing the data divide.

This will form part of a new report assessing a programme designed to strengthen relationships between demographers and the data revolution.

How well do you think the global data community has addressed the key recommendations of the IEAG report?

Tom: Despite the ambitions of the SDGs, whole regions of the world risk being left behind in terms of data.

We haven’t engaged with the political economy of the data revolution. Who asks the questions for which data is needed? Who provides the data, and for what purposes? Who controls how that data is used?

Digital data doesn’t just miraculously arrive to be harvested and harnessed. We need to understand how representative it is, how it can be analysed alongside traditional forms of data, and whether it can inform policy debates robustly.

It’s great that we have satellite data on communities that are distant from major roads, for example, but that data by itself tells us nothing about what other forms of poverty and vulnerability those communities suffer.

Vissého: The mantra ‘better data leads to better lives’ is increasingly outdated. Of course we need better data, but we also need to combine data sources in new ways.

Social media has given the poor and the left behind much stronger voices than was imaginable a few years ago. We can capture these via big data, and this gives us new opportunities to find solutions that suit the Global South. But in the euphoria of big data we mustn’t forget two things.

First, traditional data and the work of national statistical offices remains essential. Second, we from the Global South need to be able to work with our own data, and to do research that is important to our communities.

Tom: I worry that it’s too easy for data science to produce compelling models which researchers could be tempted to substitute for the hard work of real data collection.

Most funding for data science flows to well-funded research labs in the Global North. If there aren’t the skills in the Global South to challenge the results of these models, we risk a downward spiral of local capacity.

In effect, I’m seeing well-funded researchers in the Global North say to us: ‘here’s a solution to your problem we have designed and developed using your data which tells you the answers we think you want to know’.

Maybe that approach is useful in the short-term, but in the long-term it disempowers knowledge and insight from the Global South, and risks decapacitating local organisations.

How can researchers in the Global South counter the data divide with the Global North?

Vissého: There has been very little emphasis on building skills in developing countries to work with new forms of data. It’s part of a wider problem where globally the fields of demography and data science have been quite separate.

It’s only recently that ‘digital demography’ has emerged to bring them together. The programme Tom and I have been working with has begun to build this capacity in the Global South. But the problem can’t be solved with a single programme.

There needs to be a concerted effort across academia to strengthen the links between demography and data science: not just in the Global North and not just in English.

There’s a real danger that Francophone Africa and other non-English speaking parts of the world are being left even further behind.

Tom: And even where the data revolution is being pursued in partnership with developing countries, we need to question the longer-term intellectual sustainability of those projects.

Organisations like the UN Statistical Commission and Paris21 work to ensure that national statistical systems can utilise big data, but do the skills exist for developing countries to take ownership of their data for the purposes they require? Or will we see a gradual erosion and attrition of knowledge in those countries, resulting in increased dependence on those well-resourced centres in the North? We can do things like get population estimates at micro levels of granularity, but are we building the skills necessary to use, enhance and reimagine those systems so that they continue to serve local needs?

It all comes back to the question of who gets the long-term benefits of our data.

Vissého: For big data or digital demography research to benefit the Global South, it’s imperative that students from those countries are trained to use them.

The research combining different data sources I mentioned earlier is important work, but I don’t think any of the students who contributed to it were from sub-Saharan Africa.

We urgently need to build our own capacity to use all these new techniques, working on our own data, and developing our own solutions.

Where do go from here?

The problem Tom and Vissého described is far reaching. The types and quantities of digital crumbs will continue to explode. The danger is that the ability to analyse big data could become increasingly skewed towards the Global North.

So, in addition to using smart financing, as Shaida Baidee and Johannes Jütting have recently argued, we need to urgently strengthen data capabilities across the Global South. Doing this will help a build a wider range of communities with the ongoing capacity to develop locally owned solutions.

In order to frame these efforts, we need to begin with a deep interrogation of the power dynamics that shape how we gather, analyse and use data for sustainable development.


The International Union for the Scientific Study of Population have published Spanish (PDF) and French (PDF) versions of this blog.