Interviewed by Helen Wu
Dr. Henry Siu, a professor at the specializing in macroeconomics, discusses the role of digital scholarship in economic research. In addition to providing insights on the Siu also shares his perspective on the educational role of machine learning and AI for both students and professionals, emphasizing how these technologies can make digital scholarship more accessible to all.
Could you briefly introduce your primary area(s) of research?
I specialize in macroeconomics, which focuses on why economies grow, and why they experience booms and recessions. I also research labor economics, studying employment, wages, and income distribution.
How would you define “digital scholarship” as a macroeconomist?
In economics, digital scholarship is fundamental to our work. We use analytical techniques, which involve a lot of computational work. For us, digital scholarship also includes big data analysis, using comprehensive datasets and sophisticated statistical and econometric methods to extract information.
Relative to other disciplines in Arts, economic history aligns more closely with traditional historical studies and uses digital methods intensively. Many historians focus on digitizing historical materials and extracting valuable data; economic historians combine that with economic modelling and econometric analysis. So, I would say that digital scholarship is integral to all areas of economics.
One of your past digital scholarship projects is the VSE COVID-19 Risk/Reward Assessment Tool. Could you briefly explain the project and the methodology you used?
At the start of the COVID-19 pandemic, the Vancouver School of Economics was contacted by Reka Gustafson, the deputy public health officer for British Columbia at the time. She asked if we could help analyze the economic implications of the crisis.
The COVID-19 Risk/Reward Tool assessed various sectors of the economy based on factors such as the risk of workers transmitting the virus, their vulnerability to economic shutdowns and their overall importance to the economy in terms of size and income generation.
We secured special emergency access to confidential data from Statistics Canada to conduct this analysis. By April or May of that year, we provided our findings to all levels of government in Canada, helping policymakers determine when and how to safely reopen the economy.
We had access to long-form census data, which contains demographic and occupational information on people living in Canada. This allowed us to analyze risk factors such as transportation habits—whether workers relied on public transit, which could increase transmission risk, and whether workers lived with elderly family members who were more vulnerable to severe health conditions.
From an economist’s perspective, this analysis was relatively straightforward. The most technically advanced aspect was factor analysis—a standard tool in our field. We’ve got a gazillion factors, and the analysis aims at collapsing all of them into one number that characterizes risk, making the data more interpretable.
Interactive feature of the VSE COVID-19 Risk/Reward Assessment Tool
Were there any specific digital or computational tools that helped with data analysis?
We used common statistical packages like Stata. Stata is a statistical software for data science, frequently used in academia and the private sector. This software is available for free to anybody at UBC. One faculty member, an expert in data visualization, did data visualization using R and R Studio—another statistical package. These tools are also part of our undergraduate economics curriculum; we ensure that students gain hands-on experience with digital scholarship methods.
What emerging trends or innovations do you foresee in digital scholarship, particularly in economics?
Machine learning algorithms has become one of the statistical tools in an economist’s toolkit, just like factor analysis. We use it for forecasting, prediction, and data classification. Generative AI is transformative as it enables researchers, especially undergraduates, to learn code-writing without formal training. Even as a PhD economist, I sometimes encounter coding challenges. I can ask AI to generate a segment of code in Stata or R, then cross-check it with instruction manuals to understand the logic. This accessibility lowers the barrier for students and researchers, making digital scholarship more inclusive.
Are there any upcoming projects or ideas you’re particularly excited about?
I’m currently working on the impact of automation and advanced technologies on employment—essentially, whether robots are taking jobs and whose jobs are most affected.
Historically, automation has displaced workers but also created new opportunities. For example, as agricultural automation reduced farm labor, workers transitioned to manufacturing jobs in cities. Using digital scholarship techniques and historical economic data, we’re investigating whether these transitions led to a better life.
This research relies on records from the National Archives, which were previously inaccessible. Now, with machine learning techniques, we can analyze these historical data and information that were once buried in dusty stacks of documents.
From your perspective, is there more that researchers and scholars could explore in terms of digital tools and methods?
Absolutely. There’s always more for everybody to explore. In economics, we provide students with a strong foundation in digital tools. For instance, every economics major at UBC must complete a capstone research project before graduating, which involves using statistical software and analytical techniques that we just talked about. The key is ensuring that students and researchers have the skills and access to leverage these tools effectively.
