Olivia Yu's ARIA project: Tackling COVID-19 Vaccine Hesitancy in Ontario: Does Language Matter?
This project aims to evaluate the degree of COVID-19 vaccine hesitancy among francophones in Ontario and to assess the policy implications for circumventing language barriers to health care in a broader context. The first step was to conduct a systematic review of the health economics, health policy, and epidemiology literature on vaccine uptake, which guided the methodology for collecting and interpreting the empirical data.
Vaccine hesitancy was estimated using first-dose vaccination data in Ontario from the ICES COVID-19 dataset, which reports the share of the total population (and of age by decade subpopulations) having received at least one dose of an approved COVID-19 vaccine by Forward Sortation Area (FSA). Taking an FSA as a proxy for a “community”, francophone communities can be identified using the 2016 Canadian Census profile.
From the existing health economics and public health literature, the expectation is that FSA with large francophone populations (as a share of the total population) would tend to have lower and vaccine uptake rates if strong language barriers to healthcare exist. Official language minorities may experience communication barriers during the vaccine registration process at the immunization clinic. Moreover, they may have incomplete, delayed, or misleading information about vaccine safety, administration, or eligibility—any of which might give rise to delays in or outright rejections of vaccination. In practice, vaccine hesitancy can be measured by the gap between the vaccine schedule (the jurisdiction’s capacity to deliver vaccines) and actual vaccine uptake.
Preliminary regression results supported this hypothesis. However, more rigorous analysis reveals that vaccine uptake outcomes are no worse in FSA with a large population share of Francophones; if anything, francophone FSA tends to have higher vaccine uptake levels over the sample period when controlling for supply-side factors that affect uptake levels.
A possible explanation for this result may lie in the fact that the provincial government experienced some pushback early on in the pandemic when COVID-19 updates were not readily available in French. The response was targeted health policy that involved clear and timely messaging in the form of press releases from the provincial government, most likely accompanied by mandated outreach initiatives to promote vaccine uptake in francophone communities.
The positive relationship between the francophone population share and vaccine uptake is somewhat surprising, given the observations in the existing literature that health care underutilization among the francophone population is significant and comparable across the province. This finding has important policy implications: firstly, it is within the Ontario government’s means to overcome language barriers to bolster health care utilization. On top of it, the status and trajectory of the COVID-19 epidemic in Ontario is at least partly contingent on public health authorities’ ability to manage an infodemic.
I was interested in completing an ARIA project this summer because I wanted to understand the extent to which routinely marginalized and underserved communities in the health care context have experienced barriers to care during the COVID-19 pandemic. I saw the ARIA internship as an opportunity not only to explore this topic while applying the concepts and methods I have been studying, but also to develop my inquiry skills.
To this end, I had the following learning objectives. First, I wanted to apply the theoretical concepts and skills learned in my courses to contemporary economic problems. Also, I hoped to frame the empirical problem in terms of the relevant theoretical framework, to bring the theoretical predictions to the data, and to identify the most appropriate sources of data and overcome the challenges that are typical of empirical work. Lastly, I wished to determine the data that are publicly available, obtain access, clean, and analyze them.
This project came with its challenges. The main problems I encountered occurred while I was building my dataset, interpreting the data, or establishing the direction of my project. I also felt pressed for time towards the end of my ARIA project—I had accumulated so much information within such a short period of time that piecing all the elements together in writing was no easy task.
At the same time, working these problems and conferring with Professor Lasio were the highlights of my internship. I appreciated how I was always given the chance to figure out why something was not working on my own, before receiving plenty of guidance from Prof. Lasio over our calls. I have always felt comfortable approaching her with my questions and she taught me so much.
Finally, I would like to thank Mr. Harry Samuel, whose generosity is what made this internship experience possible. I am truly grateful for his financial support that allowed me to work without worries and achieve my dreams in academics.
