Special issue blog series on Policy Expertise in Times of Crisis

Cosmo Howard and Bernadette Hyland-Wood

COVID-19 showed the world that statistical data are indispensable for government decision making, especially in times of crisis. Yet, data systems to support public policy were often found wanting during the pandemic.

In our recent article in Policy & Politics, we argue that contemporary data sharing systems often behave like natural ecosystems for several reasons. They can grow and adapt organically, their members are highly interdependent, they are steered by powerful ‘keystone actors’ like government agencies and big tech companies, and they can be harnessed to provide benefits for society. This makes modern data systems different from older models, which relied on a single official statistical agency or a tight network of data providers to supply statistics for policy making. To test the operation of data ecosystems in practice, we investigated how Australia’s infectious disease data ecosystem functioned during COVID-19. We found strong evidence for organic growth and adaptation, but we also saw that keystone actors like government departments sometimes restricted access to data, which hampered the work of some data analysts within the ecosystem. Furthermore, there was often insufficient leadership by keystone actors to ensure that data ecosystems functioned coherently, resulting in disagreements over interpreting the data, as well as gaps in data coverage.

Continue reading →