The Doing Business report has crashed spectacularly. This month, the World Bank announced it was canceling its flagship publication following an audit alleging that employees tweaked rankings under pressure from country officials. Now, amid calls for the Bank to revamp the report or axe it for good, it’s perhaps time for all of us to think about how we can safeguard a data system that commands public trust.
The Bank is one of a few key global institutions that are central to the data system we have built. The World Development Indicators are the go-to source for researchers, policy makers and advocates around the world, and the Bank was an early champion of open data and the data revolution for sustainable development.
The Doing Business project was always the most political part of the Bank’s data operation. It had long been controversial in both its ideological focus on deregulation and its measurement decision making, but many believed in the value of the Bank’s flagship product and the reputation of the World Bank in assessing the global business environment. When data is so closely linked to key economic policy decisions, as two American legislators wrote this week, it’s essential that “data be and be seen as unimpeachable.” Now a little tweaking of the data—a couple of cells in a spreadsheet, perhaps—has embarrassed some of the most powerful people and institutions in the world.
Whenever explicit or implicit biases come to light, trust in institutions is a casualty. This matters because institutions like the World Bank have a key role to play in collecting and safeguarding the quality of data and shared facts upon which global issues can be debated and, hopefully, resolved. The World Bank has taken a positive first step by canceling the Doing Business report, but its leaders are now faced with larger questions of how to rebuild public trust, especially around data, indicators, and information. With the latest World Development Report emphasizing the role of trust in the social contract for data, we’ll be looking forward to seeing what other steps the World Bank takes to address data integrity.
The World Bank, and the other big institutions that make up our global data system, have to be part of the solution, but they can’t do it alone. Given the huge power imbalances in this industry, global development organizations are especially vulnerable to issues arising from constituents’ lack of trust, and they need to listen and learn from people and institutions with less power in the system.
This is why the Data Values Project places an emphasis on uplifting and encouraging organizations that are working to build and restore trust among people who feel the effects of data and its role in politics and decision making. An increasing number of organizations are seeking new approaches. Open Heroines, a community of women and non-binary people who work in open government, open data, and civic tech, commissioned a report seeking to understand power dynamics and challenges within their community. In doing so, they’ve created a model for how to engage with a diverse range of stakeholders in the open data space.
Unfortunately buried among the headlines from the past week is an alarming but all-too-common narrative of bullying and toxicity within the Doing Business team. We talk a lot about protecting people in data. But it’s also important to remember that, because data is an expression of power, safe and appropriate spaces for people who collect, analyze, publish, and use data are also needed. Large institutions are by no means the only culprits here, as recent years have seen a spate of mishandled or ignored harassment and bullying incidents among data for development and civic tech NGOs. (Sunlight Foundation is one example, but the issue is widespread. Open Heroines has also collected stories of women’s experiences across many institutions.)
Data is political, because it informs the decisions that affect all our lives. Trust in the process by which that data is produced is vital to the sustainability of the system. The Doing Business Indicators offer a lesson in what can go wrong - a lesson we should all learn from as we think about how data is collected, analyzed, and used in our own organizations and in those we support.
