6 Steps to Get Started on Decolonizing Data for Development

Statisticians entering data into the database for further processing and analysis. Turkmenistan. Photo: World Bank


If you work in data for development, you have likely heard about the potential of “data for good”. You might have even attended a webinar on how taking a more data-driven approach as a practitioner or social impact organization could help you enhance positive impact for the stakeholders you serve. Data is increasingly being leveraged as a tool to support international development interventions. Its reach spans across sectors; from data analytics for rural development, to artificial intelligence for healthcare and agriculture and blockchain for digital identities. Originally developed for the private sector, data-based technologies aim to save time, increase accuracy, enhance precision, and broaden impact. 

Despite good intentions, data is sometimes collected, analyzed and used in ways that can replicate or even amplify existing injustices and inequalities. This happens because data and data-driven technologies are often treated as neutral tools. However, since these tools are developed by human beings within a social and cultural context, values and world views can be embedded within them. Decolonizing data for development approaches can help us to unpack this.

What is decolonizing?

Decolonizing is a restorative justice movement that seeks to address the ongoing influence of colonial legacies. The precise meaning of ‘decolonizing’ is varied and contested in both academia and in practice. Despite this, many agree that it is about critically examining how colonial power structures continue to produce inequalities today and the changes we can make to address those inequalities. Coloniality has its roots in the Age of Exploration that started in the 15th century. During this period European nations explored the world in a quest to increase their knowledge, wealth, and influence. They later established settlements and colonies, with many colonies only gaining independence within the last six decades. Coloniality is not only about the violent acquisition of land, but also the cultural and psychological impact of centering European and Western ways of viewing the world, both past, and present. It shapes our current realities in international development because former colonies are considered to be the ‘developing’ or ‘Global South’ countries, and former colonizers are the ‘developed’ and ‘Global North’ countries that hold the majority of the world’s wealth and global power, which of course was extracted through colonialism.

“Decolonizing the Internet’s Languages Conference – reflections and actions” organized by Whose Knowledge? at a MozFest 2019 in London, UK.

What is decolonizing data for development and why does it matter?

Decolonizing encompasses various perspectives and approaches and is an emergent conversation across different fields such as education, journalism, media, and international development. In the context of data for development, decolonizing approaches focus on how data-based technologies are reproducing and reinforcing colonial structures of inequality. 

Concepts like ‘data colonialism’ and techno-optimism help us understand the role of data and technology in entrenching those inequalities. ‘Data colonialism’ highlights how the ways in which we seek to extract value through data mirror past colonial resource extraction patterns. This is a phenomenon in which companies that are typically based in high-income countries profit from personal data collected across the world. These companies dominate the data and digital technology supply in low-and middle-income countries (LMICs) and can hinder countries from enhancing their data economies and capabilities. This can lead to greater dependency and impoverishment

More than 500 women greeted UN Women Executive Director Phumzile Mlambo-Ngcuka in Gujarat, India. This is an “information fair” that brought together local women and elected leaders, where community members shared their experiences of using technology to catalyze change.
Photo: UN Women/Gaganjit Singh

Techno-optimism assumes that data, like technology more broadly, is neutral and that the use of data for development is a universally good thing. It sees digital tools and data – particularly in their newest and most advanced forms – as something that will inevitably change the world. Techno-optimists believe that digital tools and data can and should be unquestioningly leveraged for social impact if only practitioners seize the opportunity. Since this narrative is pervasive, it can trickle into our work and how we communicate what we do, sometimes without our being fully aware of it. While data advances do lead to change, the change is not guaranteed to be ‘good’ unless we approach it critically. It is probably safe to assume that most data for development practitioners don’t want to be complicit in the continued oppression of others. So where do we start to contribute to positive change? 

Where do we start?

1. Ask the difficult questions BEFORE you start a data for development project

Before we use data for development, we should ask why we are using it. Is it because it’s trendy and will make our project look futuristic? Are we being techno-optimistic? Are we using technology and data as the only means to an end? 

2. Apply a decolonial lens across the data lifecycle

Consider decolonial, intersectional, and inclusion, diversity, equity, and access (IDEA) principles in the design from the beginning and throughout the data lifecycle. For example, data collection should not disempower the people who contribute it, just because it might be inconvenient for data for development practitioners to have the people who generate data to own and have a say in its usage. Such considerations cannot be an afterthought. It cannot be an item on a performative list to tick off as these types of actions often do more harm than good. 

3. Acknowledge that data is always situated in a particular context

Data is neither neutral nor value-free. What is collected – and not collected – will always be embedded in broader social, economic, political, cultural, and historical contexts. Because of this, intersectional analyses must be undertaken when processing data to understand and mitigate how the broader social milieu impacts the outcomes anticipated from the usage of data for development.

4. Consider the narrative being used to communicate about data

Ask: Who is the data about? How are they being portrayed in the data? Who is telling the stories using data generated by others? How are we communicating these stories? Are we perpetuating global inequalities in our data work?

5. Have a plan and start with short-term goals

Shifting to a decolonial lens doesn’t happen overnight. Start with short-term goals and build from there, ensuring the potential to harm others is minimized or altogether eliminated. If you don’t feel confident to do this on your own, consider training or hiring a consultant to support your efforts.

6. Work together

Recognize that decolonizing data for development will require a collective effort. This means that organizations need to see this work as everyone’s responsibility and not the work of a couple of individuals. If you’re a social impact organization (SIO) just beginning your data journey, we recommend completing our Data Maturity Assessment to see your overall progress with data maturity and to obtain free resources on how SIOs can implement more intersectional approaches in their data for development work. This is a useful first step in decolonizing data since to understand where you can help, there is a need to know your own strengths and weaknesses – and to work to address the weaknesses in order to enhance the potential impact of your data for development work.


Just as the process of colonization took centuries and didn’t end when empires were abolished, the process of decolonizing will take a long time. It can sometimes be overwhelming when we look at the sheer amount of change that needs to occur. We can all start by taking small but meaningful steps to educate ourselves, be more critical and reflective in our own work, and strive towards outcomes that empower all stakeholders involved in data for development interventions.

About the Authors

data.org In Your Inbox

Do you like this post?

Sign up for our newsletter and we’ll send you more content like this every month.

By submitting your information and clicking “Submit”, you agree to the data.org Privacy Policy and Terms and Conditions, and to receive email communications from data.org.