This initiative focuses on harnessing open data for anti-corruption as well as developing data literacy and furthering the state of open data. Activities will first take place in Estonia and Latvia, two small Baltic countries that are excellent for piloting innovative tools. The scope ensures datasets are specific, local and easily verifiable while working as a pilot for application in other countries later on. The idea was developed in consortium with Transparency International Estonia, Transparency International Latvia, Open Knowledge Estonia, Datu Skola (School of Data of Latvia) and the Ministry of Justice of Estonia.
Step one. Creating the full and live version of the application Opener (www.opener.ee), an early conflict of interest detection application for leveraging issues around money in politics. It harvests three datasets from the public procurement registry, e-company registry and the registry of political party financing, visualizing the connections between winning contractors and party donors. The team already has access to all the mentioned datasets and a proof of concept in the form of a prototype to exhibit the necessity and achievability of the described initiative.
The development of Opener is agile, i.e. first of all a minimum viable product is developed which will then be gradually upgraded, also to include algorithms for red-flagging potential conflicts of interest situations. This logic of action ensures that the tool allows you to continuously improve it based on feedback from users.
In addition, ministries and other organizations will be able to build on the interface. Participation platforms built on open data (like the collective addresses platform rahvaalgatus.ee) can also make use of the Opener and its data.
Step two. Together with the Opener's minimum viable tool, the first cases of corruption will also be visualised in collaboration with investigative journalists. In this way, we show the benefits of Opener to a wider audience and create an interest in both the tool and the training of its use. Here is an illustrated example of a case of corruption: the facts available, such as public competitions, won procurements, company owners, board members, tax debts and media coverage (this one-time case analysis was done at the Social Impact Data hackathon in 2017 in Estonia): https://docs.google.com/document/d/154hv8ixfY-9zVCVneXOxHauCCj-0YnmwGfzhkOqE9sE/edit
Step three. Increasing data literacy by training and instructing journalists and other target groups in Estonia and Latvia. The trainings and consultations are organised by both the Latvian Data School (Datu Skola) and the Open Knowledge Estonia, which brings the international format of the data school to Estonia. The data trainings are designed for non-technical experts to promote the spread of data-driven journalism, advocacy and policy making.
Through a constant feedback loop and case studies brought forward by data literacy courses the lead organizations will advocate for more datasets to be published in open and machine-readable format, making way for more transparency and more effective anti-corruption work. The mapping of anti-corruption datasets in the Nordic region was done in 2019 by Transparency Latvia, proving that even if available, most datasets do not comply with open data standards. See more here: https://delna.lv/wp-content/uploads/2019/11/NB7_OD4PI_Final_cmp.pdf
As vital anti-corruption datasets are made publicly available, they will be added to Opener (such as beneficial ownership registry that is open in Latvia but behind a paywall in Estonia).
Once the application is functional, it can be replicated in other countries as well as cross-referenced internationally to expose illicit financial flows and business networks.