Research
-
The development of lethal autonomous weapons systems (LAWS) has raised concerns about legal, operational, and accountability issues when these systems violate international humanitarian law. Literature on LAWS frames the “accountability gap” as a lack of legal clarity needing doctrinal fixes. This paper proposes a different view—accountability is not missing but fragmented across a complex web of actors, each with a partial and discretionary share of responsibility. Drawing on organizational theory and studies in science and technology, the paper argues that the supposed accountability gap in LAWS development results from two semi-independent action loops. One loop belongs to private-sector developers, focusing on model training, validation, and system reliability. The other loop concerns military commanders, who focus on tactical intent, rules of engagement, and mission success. The friction at the intersection of these loops obscures control, not any single design or command failure.
This paper makes three novel contributions. First, it introduces the concept of a “mutual expectation trap,” where developers shift moral judgment about engagement decisions to downstream commanders. Meanwhile, commanders assume that complicated model behavior has been made safe upstream. Second, it highlights a situation where responsibility is divided among data labelers, software developers, procurement officers, and strategic overseers. This allows each node to shift blame laterally through handoffs between professionals. Third, it explores whether the dilution of accountability works not as a problem but as a structural facilitator of rapid defense innovation. This setup allows states to deploy hidden capabilities because no single actor faces personal risk from systemic failure.
-
Between July 2018 and December 2019, the Trump administration launched a trade war with China by levying four major tranches of country-specific (Section 301) tariffs against China, raising concerns about trade rerouting. While prior research has observed limited evidence of Chinese trade rerouting through Vietnam due to the trade war tariffs, there has yet to be a cross-national analysis assessing whether Chinese trade rerouting through multiple countries is a systematic response to China-specific tariffs. In this thesis, I estimated the effect of Section 301 tariffs on patterns of Chinese trade rerouting through third-party countries between 2018 and 2022, and found that the tariff imposition slightly increased the volume of Chinese trade rerouting. Specifically, I discovered that products subjected to tariffs in 2018 exhibit significant trade rerouting effects, whereas products tariffed in 2019 show no such effect. I also observed that trade rerouting is evident when examining trade quantities, but not when analyzing trade values. Next, I used a combination of quantitative and qualitative methods to analyze the mechanisms driving these discrepancies. The mechanisms analysis finds that differences in trade rerouting between the 2018 and 2019 cohorts are primarily explained by two factors: first, the 2018 cohort included products strategic to the Made in China 2025 industrial policy (“strategic products mechanism”); and second, intermediary countries with stronger preexisting trade ties to both China and the US were more likely to serve as conduits for rerouting 2018 cohort products (“close ties mechanism”). Finally, the analysis revealed more limited evidence of rerouting in trade quantities compared to trade values, attributable to variations in pass-through rates and product complexity.