New Grant Application
Nikolay Marinov and Harald Schoen expect to file an application to the DFG in the spring of 2017. The topic and some preliminaries appear below.
The Hybrid War of Information
1.1 Project summary
The most recent incarnation of the war for hearts and minds among the Great Powers is the hybrid war of information, levied by the Kremlin against the West. The Kremlin’s campaign is a far-reaching, sophisticated, well-funded operation that shares some features in common with campaigns that came before but also breaks new ground. We use theories of international relations, public opinion formation, and psychology to lay out an argument about the impact of the propaganda operation. We identify the issue areas in which the campaign is focused, and the characteristics of individual respondents that predispose people to respond more positively to the Kremlin’s viewpoint. We ask what explains variation in the types of arguments deployed in propaganda and counter-propaganda. For example, we want to know why some responses/stories feature conspiratorial thinking while others do not. We propose the use of content-modelling to track the features and pervasiveness of propaganda. We draw on recent advances in communication studies to see how stories spread online. This includes the explicit study of trolling, and its place in the `war on democracy’. Part of our motivation is to understand the political ramifications and electoral impact of the hybr id war campaign. We also envision collaboration with computer scientists, to understand the cybersecurity aspects of the informational war. A simple but significant goal of the project is to develop a working definition of “informational hybridity’’. The more complex goals include the development of new, interlinked theories, and the accumulation of significant new data. The new data will give rise to multiple spin-off projects.
The Impact and Effectiveness of Foreign (pro-Russian) Messages
We define a hybrid war as an asymmetric campaign, waged on multiple fronts and through various means. Russia does not have the power of its Western opponents which dictates investing most of its resources in non-traditional, high-impact channels. Buying TV anchors, paying for trolls to invade social media, and subsidizing its own media in the West is a combination of overt and covert moves that sums up to a hybrid response to the perceived Western threat. Given the interests of an authoritarian foreign power, surrounded by powerful democracies, a hybrid war of information is an optimal response.
How does information, pushed by Russia onto the Western media market, affect or change the way respondents view issues of interest?
The following aspects of hybrid wars are theoretically interesting: (1) the messages that hit respondents may contain information that is demonstrably true; (2) the messages may contain information that enables judgments on the significance of the event: by interpreting, or adding distortions to the core facts; (3) the true source of the messages may and sometimes is not revealed. These features are part of the DNA of the hybrid war of propaganda.
A hybrid war pits one state actor against another. Unlike classical propaganda (Laswell, 1927), the target is the audience in another state. That means the source of the information has to contend with powerful counter-currents in a given society. Not every targeted individual holds favorable attitudes toward foreign powers providing propaganda messages. Obscuring the real source of the information will often be optimal for the originator of the message, as a strategy to avoid priming a nationalist or patriotic effect.
1.2 Preliminary Work
We offer results on how sourcing to a message to an elite affects a respondent’s take on it, and why. We argue that respondents who hold more favorable views of a source (Western or Russian) would find such a source to be credibility-enhancing. We show results from a German panel of 2,400 participants. We are able to report evidence on a several customized question-items and also results from eight survey experiments. The evidence allows us to offer tentative conclusions about the potential effects of the information war. The argument and evidence help point the way to what can plausibly be done to arrest some of the more worrisome aspects of the developments we identify.
Preliminary Evidence from Germany
We illustrate the promise of our approach by reporting some results from the German study. The appendix provides a flavor of the analysis from one of our experiments, the Ukraine-Ouster one. Our main findings so far can be summarized as follows.
In the war for hearts and minds that characterizes the hybrid warfare between the West and the Kremlin, opinions remain relatively malleable. This is more so in the case of what we call issues, policy domains of concrete concern in the conflict. Both the pro-Western and pro-Russian narratives operate strongly in that domain. In the case of narratives aimed to attack Western multilateral and domestic institutions, we find considerably more stability. The pro-Russian angle of attack is less successful, whereas the pro-Western, reinforcing narratives remain strong.
The main fault-lines of the hybrid war of information are not necessarily religious, cultural or authoritarian. We do not find that any of these variables mediate the impact of the frames we study. By contrast, beliefs in conspiracy theory and overall political knowledge play a strong conditioning role in whether a narrative has power. The worrisome aspect of these findings is the prospect of further polarization of public views, as different groups absorb at greater rates diametrically opposed narratives, and grow even further apart on policy issues and on the performance of institutions.
The impact of sourcing narratives in this war of information is quite complex. Western elites enjoy no natural advantage, and Russia, as a foreign actor, sometimes holds an edge. The overall impact of clarifying the source of the information is ambiguous.
Our objectives are to define, in a theoretically-defensible way, the set of [TBA]; to develop hypotheses about their occurrence [TBA]; to collect data to test the hypotheses; to disseminate the findings to academics, policy-makers, and make the data widely-available and easily accessible.
We define a hybrid war as a deliberate attempt by a foreign government to change [TBA]. As social scientists, we care about explaining the conditions under which foreign interests utilize different information channels, choosing different messages. At present, we do not understand this. As a result, we can neither satisfactorily explain foreign interest in democratic politics, including in electoral contests, nor predict the adopted strategies.
Defining a Hybrid Information campaign
If the theory we work with is correct, it would be difficult, in many cases to even tell a hybrid war piece of news from a “regular” piece of news. The reason is that an anti-refugee message may be originating with a domestic actor, and may be intended to promote the interests of the domestic actor. Uncovering the outside boost to this component may be a challenge. We have conducted an experiment, asking subjects to code the “hybrid war” component of RT.com’s front page, in English and German. There is considerable disagreement on the proportion of hybrid news, ranging from 20 to 60 per cent. Furthermore, the share of subjects unable to classify a piece of information as either war or not is very high: between 10 and 60 per cent. The main criteria cited when classifying a piece as hybrid war were: news brings panic, portrays things in black and white, is very tendentious against Western actors, exaggerates divisions in West - between countries but also groups, portrays Russia in a positive light and fails to dwell on Russian problems. Also, news items in different categories were systematically different in their emphasis on hybrid war materials with World News/Politics including more than Sport/Business sections.
We can use topic-modelling and the study of online social forums to understand how the information we study spreads. We can model how RT.com topic diverge from or coincide with local market topics. We can track topics over time, to see whether pro-Russian sites abruptly change coverage on some (but not other issues) in response to specific events: such as in response to Russia’s bombing of a humanitarian convoy, to divert attention from the issue. Also, which events are subject to such spikes will be of interest.We can study the determinants of changing topic emphasis when it comes to issues, politicians and institutions – namely, whether issue coverage is more responsive to specific events, whereas attacks on institutions are more long-term and not predicted by the same factors. We can test a diversionary theory, a PR theory, or a “weaken-your-adversary” theory. We can simultaneously plot topic coverage in Western media vs RT.com and check the deviations.
Estimating the Effectiveness of a Hybrid War campaign
This problem has two dimensions: how much information enters the Western public sphere, and what do respondents makes of it.
No one quite knows the size of the problem. A recent study by the European Parliament sounded the alarm bells, by noting that Sputnik.de (a Russian) outlet has 150,000 FB likes. RT Deutsch on FB has quarter of a million likes. The study noted that EU institutions are seldom aggressive in pushing out information, and rely on the assumption that facts speak for themselves. We want to track “Russian” theses as pushed by Russian outlets, compare them to theses that are pushed by “genuinely” domestic outlets such as SDZ, and understand the relationships. Furthermore, combined with our survey design, below we can track opinion penetration over time, by approaching respondents directly. This would be the first direct and immediate measure of the size of the problem.
We use survey-experiments to learn how information, pushed by Russia onto the Western media market, changes the way respondents view issues of interest.
We will look at Russians in Germany and Latvia, and measure the opinions of that community. Evidence (from Latvia) indicates they share many views with ethnic Latvians but sharply diverge on NATO – potentially in response to propaganda.
Interviews with Elites
As part of the data-collection, the PI’s will conduct interviews with policy-makers to understand which messages concern them the most, and why. We hypothesize a number of relationships, arising from the objectives of the hybrid war operation, and the anticipated consequences on the behavior and calculus of politicians.
It is also of interest to examine when Russian elites choose to sway public opinion on an issue vs. assailing the reputation of a specific political leader. There is variation in that regard. We will collect evidence, and work on models explaining the shift from issue-specific campaigns (NATO build up in Poland is bad) to candidate-specific ones (Merkel is bad), or a mixture.
The issue of hacking is very interesting in this context. Conspiratorial narratives, and Wikileaks type of Robin Hood arguments (elites are bad, information enables the common person to push back) are central to the theories we investigate and develop. We can build game-theoretic models of which elites’ communications will be broken into, and then, which ones will be made public. The models will incorporate, as part of the sender’s strategy, the anticipated response to the revelation of information. We plan to consult and collaborate with computer scientists, to check the model’s assumptions. Based on what we learn, we will consider developing further projects, using evidence of data-theft as the dependent variable.