The initial premise that open data alone would lead to a radical change in outcome for governments and citizens has proved overly-optimistic. As a recent Transparency International/Web Foundation report found, countries are often failing to do the first step of following through on their commitments to put the right policies in place, let alone ensure that they are delivering results. Where information is published, there are some important things that need to be in place if open data is going to have an impact on corruption. They are:
The existence of information in digital, shareable formats. Lots of government processes use paper forms, or generate PDFs, which can be difficult and costly to digitise for computers to use.
Information needs to be published as open data in line with the six principles of the Open Data Charter. The Charter sets out the best way for governments to publish data - it should be open by default, interoperable and timely.
The right data needs to be published. Given that corruption happens through often complicated networks, you need multiple datasets to build a picture of what is going on. The guide includes a list of 30 different datasets that can be used to fight corruption.
Datasets need to be able to talk to each other in order to build a picture of what’s going on. The list of datasets in this guide includes information on what key features the datasets have to have in order to be interoperable. The Guide helps governments understand what datasets they should prioritise for collection and release, and which data standards are out there to use.
Citizens, journalists and members of civil society have the skills and legal protection to analyse the data that’s published, share findings publically and seek a response to their findings. The direct users of raw data are likely to be a small number of engaged specialist organisations and individuals, who can act as intermediaries between data sets and the broader public.
Data publication and analysis needs to be accompanied by other investigative methods, such as interviewing sources. Insights from a dataset is often only the first step in combating corruption.
There need to be structures in place so that governments and law enforcement respond to indications of potential corrupt activity.