Ognjan Denkovski / 3 Dec 2020
This study brings a new perspective into how domestic and foreign actors’ propaganda methods for disinformation can be used on a social media platform – Twitter. On a case study of North Macedonia during a period close to the 2020 parliamentary election, the author applies original botnet identification techniques to identify a large network of users created in the run-up to the election and sympathetic to VMRO-DPMNE (the country’s right-wing party) and opposed to N. Macedonia’s NATO and EU integration.
The study examines the presence of (foreign) computational propaganda methods for disinformation purposes in North Macedonia, a landlocked country on the Balkan Peninsula, one of six Western Balkan countries involved in EU accession-related discourse. The study is focused on the period surrounding the 2020 election, originally scheduled for April 2020, but
postponed to July 2020 due to the covid-19 pandemic. We build on insights from North Macedonia’s first computational propaganda campaign, namely the #bojkotiram (‘I am boycotting’) campaign on Twitter, which significantly shaped discussions surrounding the 2018 name-change referendum. We apply several botnet identification techniques, including
looking out for repetitive naming patterns, large numbers of similar accounts created prior to key events and activity rates which exceed normal human behavior, usually achieved through retweeting. Using these approaches, we identify a large network of users created in the run-up to the election and sympathetic to VMRO-DPMNE, North Macedonia’s right-wing party, as well as to Levica, a far-left party opposed to North Macedonia’s NATO and EU integration. Many of the identified accounts oppose North Macedonia’s name-change, while also promoting conspiratorial content and anti-Western attitudes. Conversely, very few of the accounts identified expressed support for SDSM (North Macedonia’s Western-oriented centre-left party), the name-change or progress in the country’s Euro-Atlantic integration process. These findings are not aligned with results from public opinion polls regarding North Macedonia’s foreign policy, which show that most citizens are in favor of Euro-Atlantic integration. Moreover, we find that the network identified has extensive overlap and interaction with accounts originally created for the #bojkotiram campaign, which is still active on Twitter.
The analysis suggests that the network identified in the current study is likely run by local actors, as we did not identify any direct foreign involvement. However, even if no foreign actors directly contributed to the development of the network identified, the findings show that the conditions for easy entry by actors interested in developing disinformation campaigns in the country are present, both in terms of technical know-how and existing networks of (automated) accounts which promote anti-Western sentiments. Researchers focused on identifying and responding to disinformation campaigns on social media in the Western Balkans are advised to consider the naming characteristics and closely related account creation dates identified in this study, while paying particular attention to topics and issues pertinent to right-wing voters and parties in the region, such as those endorsed by North Macedonian VMRO-DPMNE or Serbian SNS.