Over the past few years, China’s One Belt-One Road (OBOR) investment and trade initiative has gained a lot of attention from students, particularly those visiting UCL from China. This seems to be the result of political announcements and media coverage: ‘everyone is talking about it’ is a frequent motivation. This, in itself, is not at all bad. One thing I frequently stress to thesis students, particularly those facing difficulties in choosing a topic, is that they find something that interests them. After all, they’ll spend a quarter of a year working on that project, and a loss of motivation can be catastrophic for their degree and career goals.
So what’s the problem with OBOR?
A pattern emerged in the OBOR-related theses I supervised over several years. In the best efforts, artificially constraining the research question to the few years in which OBOR has existed created significant data issues. Not only are the public data often of very poor quality, missingness and a lack of observations severely limited the extent to which any meaningful inferences could be made. These problems were further magnified by students’ methods training, which limited the analytical options they could feasibly apply. Consequently, the best of these OBOR projects never performed particularly well compared to projects I supervised on other topics.
The distribution of outcomes on OBOR projects revealed something of a ‘missing middle’. Students who made an effort to do the best they could with the data constraints they faced, often incorporating mixed methods or extensive use of vignettes, performed well relative to those who fell back on more journalistic approaches to the topic. In this latter group, the best outcomes merely repeated news reports and speculation, while the worst fell into a 10,000-word rendition of government propaganda. Both approaches demonstrated a severe lack of research design that should be core to the MSc research project.
Solution to the OBOR problem: Don’t think about OBOR
After repeatedly failing to get students to successfully tackle a research-based approach to OBOR projects, I think I finally have a way to work around their preoccupation with OBOR. Prior to our first supervision meeting, I have my students fill out a brief proposal for their projects. This form differs from the departmental form in its focus on research design, data discovery, and where the proposed thesis fits into the existing literature. This year, I specified that students hoping to work on projects related to OBOR omit all references to OBOR. In most cases, this caused students to shift their attention from the OBOR initiative itself to the underlying relationships between actors and flows, and the resulting data-generating process. It also prompted them to look for data beyond the constraints of the OBOR timeframe.
Moving beyond OBOR itself and examining China’s economic relationship with other countries over a broader period of time presents new opportunities for students to identify useful data sources. It also opens doors for the application of methods beyond those they learned in their introductory econometrics class. Diff-in-diff, regression discontinuity, and repeated cross sections are relatively approachable analytical methods that can be employed in the context of methods they already understand. For qualitative projects, the incorporation of pre-OBOR periods allows for a larger population of potential cases.
The difficulty for students at this early point while under supervision is to retain their focus on the underlying phenomena highlighted in their proposals. A reversion to thinking explicitly about OBOR could generate the same suboptimal outcomes previously observed, while those who focus on research design should be able to produce analysis that both explains patterns in China’s outbound economic flows and the effects of OBOR on these patterns.