The Opportunities and Limits of Societal Verification

The Opportunities and Limits of Societal Verification, by Kelsey Hartigan and Corey Hinderstein, makes the case that work done by non-government parties (societal verification) has an important role to play in arms control verification. The article discusses various models for societal verification, its challenges, and how it can be utilized by governments. The article concludes that the best way for societal verification to be used by governments in arms control verification is through the use of networks of outside experts. These experts will serve as “canaries in the coal mine”, whose findings get the attention of government officials that have the final say. The article also makes the suggestion that public (open source) information should also be used by the government. However, because the article doesn’t focus on outside experts, it is vague in discussing important details of how outside experts can be utilized, how they can be helped by the government, and what are potential pitfalls of utilizing them.

The focus of the article is pretty broad. It primarily discusses opportunities for arms control verification that have arose from the popularity of the internet. Namely, a vast amount of data that is important for verification is available on the internet and this data can be accessed by many people not affiliated with the government. This is relevant for arms control since many non-government weapons experts in places like academia can easily find data, such as photos, on sensitive military equipment from traditional and social media, online. These experts can use this online data to discover arms control treaty violations and other important facts.

One example of this that the article mentions is the investigation of North Korean Transporter Erector Launchers (TELs) by the Arms Control Wonk network/blog. This was a case where academics and others outside of the government compared photos of TELs from a North Korean military parade to photos of Chinese TELs from social media to uncover the transfer of TELs from China to North Korea, in violation of sanctions. This transfer was not publicly known until it was discovered by these non-government experts.

This scenario clearly demonstrates that outside experts have important contributions to make to arms control verification. Thus it would be interesting to discuss how outside experts can be helped by the government, and what are possible downsides of using their work. However, the article choses not to focus on these issues and instead discusses seemingly less important topics.

An example of this is the subsection on “Data Management”. The subsection begins with the claim that “it will be essential to develop a framework” for data collection and dissemination in a “consistent, user friendly format”. It only becomes clear what this means when the subsection later suggests “WordPress” (a popular blogging platform and program) as a possible solution for this problem. Thus, it appears to be saying ‘blogs should be used to publicize research’. The rest of this subsection also illustrates another issue I had with the article as a whole: it uses buzzwords seemingly for the sake of using them. Specifically, the subsection adds that “Innovations in cloud computing” and advances in “big data” will help with challenges in societal verification, without discussing these challenges in any depth.

I think it would have been more useful if the article discussed the relationship between government and outside experts in greater detail. In particular there were a few topics related to this, that seem worthwhile exploring, but were not discussed.

One of these is the motivation of the outside experts. Although some outside experts are currently motivated to do societal verification, maybe more research would be done if the government provided incentives for societal verification. These incentives could be monetary, for example, by providing a reward to researcher that discover a sanctions violation. However other kinds of incentives might effectively motivate more researchers as well.

Another topic that wasn’t really discussed is the public nature of the discoveries, and the challenges this poses. Because sources are revealed in societal verification, this allows the offending government to prevent similar disclosures in the future. For example, in the TEL case discussed above, North Korea now knows not to display sanctions violating equipment in photos of military parades, since blog posts containing the pictures have revealed a sanctions violation. However, if the violation were discovered by an intelligence agency using the same sources, North Korea may never learn how their sanctions violation was discovered. Although the article does discuss techniques like censorship as one way governments can frustrate societal verification, it doesn’t really discuss this cat and mouse game aspect of societal verification. — Jonathan