How Automaidan was “Judged”: Analysis of Court Statistics (2013-2014)

Directions, Judiciary, Open data, Research, Statistics, Video

When fabricating cases against Automaidan in the beginning of 2014, judges invocated the article 122-2 of the Administrative Offence Code of Ukraine – failure to stop a vehicle at policeman’s request. Therewith the fabricated protocols bear no records of pursuit of a car that presumably failed to stop, no references to application of strobe beacons and loudspeakers. There was also no mentioning of the article 130 – driving under the influence (quite frequent for real cases of failure to stop a vehicle at policeman’s request). We decided to analyze the data from the Unified State Register of Court Decisions, to find out the statistically significant anomalies.

In order to do this, we downloaded almost 6000 court decisions, and after some filtering, analyzed 5680 court decisions.


Interactive visualization that we developed allows not only to analyze data set online, but also to access the original texts of court decisions on the website of Unified State Register of Court Decisions.

Comprehensive analysis of judicial practices involving the article 122-2 of AOCU has statistically proven the occurrence within the period from the beginning of January 2014 and up to 21.02.2014 of a distinct anomaly in penal practices for the aforementioned cases. In contradiction to previously common practice where violation of the article 122-2 of AOCU was punished by a fine, in most Automaidan cases (from 52%) judges tended to revoke activists’ driving licenses.

It is worth mentioning that all such cases are lacking the instance of pursuit, and in particular the use of sound-and-light alarms. This anomaly can not be explained by any circumstances except for the deliberate pressure on judges delivering such decisions. Therefore the collected statistics confirms the investigative lead that adoption of these decisions was organized and influenced.

Read the full text of the research, including with description of methodology and dataset available for download: