BANNER: (Source: @nikaaleksejeva/DFRLab)

Sputnik appears to be targeting Russian-speaking audiences in the Baltic states with negative reporting about NATO, in part by inflating overall story numbers with negative opinion pieces, a content analysis of 278 articles on NATO-related safety incidents and instances of disorderly conduct reveals. The incidents described in the pieces were objectively negative; however, Sputnik’s sheer volume of news reporting on them, coupled with its many negative opinion articles, ultimately rendered the incidents’ newsworthiness to be hyperbole.

The negative coverage is probably intended to foment anti-NATO sentiment among the Russian-language audience in the Baltic states.

Sputnik produced the largest number of negative articles on three topics. The first was an accidental air-to-air missile launch by a Spanish Eurofighter EF2000 jet in Estonia on August 7, 2018. The second topic, a fire at the Adazi training ground in Latvia during NATO enhanced foreign presence exercises, took place on June 10, 2018. The third focused on multiple incidents of NATO soldiers’ public drunkenness and disorderly conduct in Lithuania.

Local public broadcasters — ERR in Estonia, LSM in Latvia, and LRT in Lithuania — and Delfi, the most popular online media outlet in the Baltic region, included more pro-NATO and neutral coverage of events than Sputnik.

Sputnik’s reach is almost nonexistent in the Baltics’ local language information spaces. Sputnik does, however, have wider reach in the region’s Russian-language information space than do the local public broadcasters, especially LSM in Latvia and LRT in Lithuania.

The content analysis compared the disposition of public broadcasters’ and Delfi’s reporting on the three topics to that of Sputnik in each of the Baltic states both Russian and the local language.

On the third topic, NATO soldiers misbehaving in Lithuania, there were at least nine news stories identified. Among them was a February 2017 incident in which local authorities detained drunk Czech soldiers for aggressive behavior in Klaipeda; a June 2017 incident in which four German soldiers were hospitalized after a fight in Jonava, Lithuania; and a June 2018 incident in which drunk American soldier fought two drunk Azerbaijan nationals in Kaunas, Lithuania.

The DFRLab’s analysis focused on two core traits of each written item: its portrayal of NATO (“negative,” “neutral,” or “positive”) and item type (“news,” “opinion,” or “analytic”). On the former, the sentiment was assessed based on the opinion or editorial stance on NATO. If editorial statements or opinions were identified in a “news” item, the item would be classified as positive or negative; in contrast, a “news” item was categorized as neutral if it cohered solely to facts with no editorial statements or quotes of commentators’ opinions. “Opinion” articles, by comparison, were largely based on opinions from political experts or commentators. Finally, if an article featured opinions from multiple experts, the article was categorized as “analytic.”

Sputnik dominated coverage of the accidental missile launch in all three Baltic states. It was also, by an overwhelming margin, the source of the most reporting on NATO soldiers’ disorderly conduct in Lithuania.

Of all the outlets, Sputnik produced the majority of negative media coverage on NATO activities. While it also did some neutral, fact-based reporting, it had no NATO-positive coverage. By contrast, the Estonian public broadcaster ERR.ee and commercial media outlet Delfi.ee engaged in positive, as well as negative and neutral reporting. For instance, ERR.ee reported on the Estonian Minister of Defense Yuri Luik’s comment that the NATO mission in the Baltic states should not be discontinued as a result of the Spanish fighter jet incident.

Negative articles were almost half, or 45.68 percent, of the articles analyzed overall. Over a half, or 52.76 percent, of all negative articles across media outlets in all languages were opinion pieces. Sputnik published 79 percent, or over three quarters, of the negative opinion pieces.

The DFRLab has previously reported on Sputnik’s use of expert opinions to shape attitudes about NATO.

Of the three topics, the accidental missile launch had the largest proportion of opinion pieces, as opposed to news or analytic pieces. Opinion pieces comprised nearly half, or 44.44 percent, of all reporting on the missile launch. In the case of the Adazi training ground fire, opinion pieces comprised nearly a quarter, or 23.53 percent, of all reporting. Just 14.29 percent of the coverage on the NATO soldiers’ disorderly conduct in Lithuania constituted opinion pieces.

Opinion pieces comprised a significant proportion of negative reporting on the topics of the accidental missile launch. Neutral news items were the single highest portrayal-plus-type coverage of the Adazi fire, but the overall negative coverage in both opinion, news, and analytic formats far outweighed the neutral reporting. For both topics, Sputnik produced a higher proportion of negative coverage than it did neutral coverage. Removing the negative opinion articles, however, would yield more neutral news rather than negative news pieces.

In the items on NATO soldiers’ disorderly conduct in Lithuania, the media outlets under scrutiny published a greater amount of negative news reporting than neutral reporting. This might be the reason why the ratio of opinion pieces was the smallest in this case. This suggests that, given the already disproportionately negative news coverage, Sputnik did not feel compelled to distort an already imbalanced negative:neutral ratio further by adding negative opinion coverage.

Negative reporting on NATO-related accidents dominated the Russian-language information space due to extensive reporting by Sputnik. In contrast, neutral reporting dominated the local language space in Estonia, Latvia, and Lithuania. A representative share of Sputnik’s neutral coverage was in local languages as well. Overall, Sputnik was responsible for the highest proportion of negative coverage across both the Russian and the local language spaces.

Negative portrayals of NATO dominated Sputnik’s reporting on the cases across all languages. The chart below shows Sputnik’s total January 2019 traffic in comparison with the public broadcaster and Delfi’s total January 2019 traffic in each state, according to the web traffic analysis tool SimilarWeb.

Sputnik has modest reach in the local language spaces. In the Russian-language space, however, Sputnik has significant reach but still less than that of the local versions of Delfi. In Estonia’s Russian-language space, Sputnik and ERR, the local public broadcaster, have similar reach. In Latvia and Lithuania, the respective local broadcasters, LSM and LRT, have less reach in Russian than Sputnik. Because LRT does not have a specific subdomain for its Russian-language-reporting, the DFRLab could not use SimilarWeb to analyze its traffic.

In contrast, the majority of Delfi’s reporting on the three topics consisted of neutral news. Delfi’s opinion coverage of the accidental missile launch in Estonia, however, was mostly negative.

The DFRLab calculated the number of articles that mentioned NATO between February 15, 2018, and February 15, 2019, across all media outlets and their respective local language versions. Local public broadcasters in Latvia and Lithuania reported on NATO in Russian significantly less than Sputnik. Only in Estonia did the local public broadcaster, ERR, cover NATO-related stories in Russian more often than both Sputnik and Delfi.

As a result, although Delfi has the largest reach in the Baltic states, Sputnik, as Russia’s state-funded media, is more effective in communicating a cohesive narrative about NATO-related topics to a Russian-language across all Baltic states than the individual Latvian and Lithuanian public broadcasters, LSM and LRT. The Estonian public broadcaster ERR, however, garners similar reach with Estonia’s Russian-speaking population as the Russian version of Sputnik Estonia.

Sputnik’s opinion articles dominated the overall negative coverage across the three cases: the accidental missile launch in Estonia, the Adazi training ground fire in Latvia, and the reports of NATO soldiers’ disorderly conduct in Lithuania.

The incidents described in the pieces were objectively negative. Sputnik, however, combined significant volume of news coverage related to the incidents with a significant volume of negative opinion articles. While opinion pieces constituted less than half of Sputnik’s negative reporting in all of the cases analyzed, the combination of facts-based news reporting and negative opinion coverage had the ultimate effect of exaggerating the the incidents’ newsworthiness.

Sputnik’s overall traffic was modest in comparison to that of the local versions of Delfi in the Baltic states. Sputnik had more traffic in Russian, however, than the Latvian and Lithuanian public broadcasters and had almost equal traffic to Estonia’s public broadcaster, ERR. Public broadcasters in Latvia and Lithuania also provided little overall coverage of NATO-related topics in the Baltic states over the period studied.

This data-driven analysis shows that Sputnik has not significantly penetrated the local-language media market in the Baltic states. In the Russian-language space, however, Sputnik has greater presence in NATO-related coverage than local public broadcasters in the Baltic states. Currently, Delfi provides the most significant counter-narrative to Sputnik’s coverage in the Baltics states. Although Delfi produced fewer articles about NATO from February 15, 2018, through February 15, 2019, than Sputnik Estonia, ERR, Sputnik Latvia, and LSM, its overall reach with Baltic audiences significantly surpasses Sputnik’s in both the local and Russian language spaces.

Follow along for more in-depth analysis from our #DigitalSherlocks. You can also meet them in person at the DFRLab’s 360OS in London this June.