2021 Shortlist

Rings of Corruption. How Kyrgyz businessmen fakes competition for 83 million soms

Country/area: Kyrgyzstan

Organisation: Kloop.kg

Organisation size: Small

Publication date: 15 Jul 2020

Credit: Ekaterina Reznikova, Rinat Tuhvatshim, Khakim Davurov, Aziza Raimberdieva, Alexey Gulyaev, Atai Narynov, Anna Boiko, Arseniy Mamashev

Project description:

Kyrgyz businessmen imitate competition and win state tenders without any obstacles, because there are no other participants besides them. The law prohibits entering into contracts with such companies and calls it a “conflict of interest.”

For this project we scrapped and combined the databases of state procurements and legal entities registered with the Ministry of Justice. 

Impact reached:

The whole project was not just publishing story, we conducted a series of online training sessions for all who were interested in learning Neo4J graph database (around 20 participants not only from Kyrgyzstan).

The explaining video has watched more than 200 000 times (both on Facebook and Instagram). 

As a result of this project, after combining two databases we received a huge amount of our own data. Since we are small media we do not have enough resources to verify and work on everything. 

So we decided to launch online investigation school, where we teach people journalism and investigations, while they train by working on with our data. Now we have around 100 students (althoug around 33 of them are actively participate, make all tasks and attend online classes).

Techniques/technologies used:

We scrapped two databases: state procurements and legal entities, combined them and used Neo4J graph database to show the connection between companies involved in state procurements. We also created an interactive interface where everyone can try and see how links are built in our database.

What was the hardest part of this project?

Three challenging things:

1. The site of Kyrgyz state procurements itself: everything was filled in manually by employees (and often with errors and mistakes), so we double-checked the information

2. Every case needs to be veryfied to make sure that the companies are really controlled by the same people

3. For some reasons at the final point we decided to create our own interface for exploring connections in our database. 

What can others learn from this project?

This is the unordinary approach for a journalistic stories and investigations. Such a data-oriented approach might be useful for other journalists.

Project links: