COVID-19 Governance Series: Robert Blair, Travis Curtice, David Dow, Guy Grossman
Author: Matthew Lisiecki
In today’s COVID-19 Governance Series we feature EGAP members Robert Blair (Brown University) and Guy Grossman (University of Pennsylvania), along with their co-authors Travis Curtice (Dartmouth College) and David Dow (Duke University). We asked them about their recent working paper “Public Trust, Policing, and the COVID-19 Pandemic: Evidence from an Electoral Authoritarian Regime” and the connections between trust in institutions, public messaging, and health measures.
This study looks at the role of public trust in creating support for health measures during the COVID-19 pandemic, along with the relationship between compliance and trust in sources of public messaging. How does this project connect to your broader research agenda? Have you had to adapt your research agenda at all to fit the constraints of conducting studies in the context of COVID?
Robert Blair, Travis Curtice, David Dow, and Guy Grossman: We have all previously worked on understanding policing in the Global South, including in Uganda, examining how the police interact with communities and how citizens’ perceptions of the police vary. Several of us have also previously done research on the role of trust in sustaining compliance with public health restrictions (Rob) and the role leaders play in overcoming collective action problems (Guy). This project offered an opportunity to explore the central (and very controversial) role that the police play in enforcing public health policies in Uganda, and in many other low-income countries. In this way, the project integrated various strands of all our research agendas.
The pandemic did also change the way we approached fieldwork. Several of us postponed projects. In many other projects, we’ve adapted by pivoting to phone surveys and other virtual formats rather than face-to-face surveys to mitigate health risks. For this study on COVID-19 in Uganda, we benefited from having a large sample of recent respondents where following up via phone was not too difficult and attrition could be minimized.
What were the primary findings from this project, along with the implications of those findings on policies related to public health messaging?
RB / TC / DD / GG: We would highlight two key findings. First, trust in leaders — be they religious, tribal, communal, or political — plays an important role in overcoming collective action problems associated with behavior that has public health implications. Second, political leaders and the state in particular have an outsized effect when it comes to encouraging individuals to bear personal costs for the greater societal good. During public health emergencies in low-income countries, it can sometimes be tempting for foreign donors or international NGOs to deliver care directly to citizens without engaging the governments of those countries, except in rather superficial ways. This is understandable but shortsighted. Ultimately, citizens need to be able to trust the messages coming from their own political leaders and local public health officials. Donors and other third parties should be careful not to marginalize or undermine the government’s credibility in the eyes of citizens. As we’ve seen in the US as well, credibility is really key to curbing the spread of infectious diseases.
Ugandan citizens interact with a number of institutions, including the Ministry of Health, local governments, traditional leaders, religious authorities, and the police. For which of these institutions was the level of trust most important in determining compliance with COVID-19 policies, and why?
RB / TC / DD / GG: As you hinted in your question, “the state” is a complex organization made up of many actors, agencies, and ministries. In our study, we sought to explore what parts of the state are most consequential for public health. We find — somewhat surprisingly — that trust in the police is especially important for convincing people to adhere to Covid-19 regulations, arguably because the police have been tasked with enforcing those rules. If you believe that enforcement of rules is equitable and fair, then you are more likely to support and comply with them, even when compliance is costly. Yet, we know that institutions like the police are often highly politicized and enforcement is unequal. These institutional inequalities can therefore be a large barrier to promoting broader compliance.
How do these findings complement the existing literature on public health messaging and public trust?
RB / TC / DD / GG: Existing studies (including our own prior research) show that trust in the state — and in leaders more generally — is key for encouraging people to adhere to public health guidance during public health crises. However, past work treated “trust” in a rather coarse way, generally paying insufficient attention to the fact that there may be significant variation in the types of agencies and authorities that citizens do and do not trust. Our study complements past research by using experimental methods to juxtapose different types of trusted leaders, and testing which is most consequential for compliance with public health directives.
What elements of this study are unique to the context of Uganda? As an electoral authoritarian regime, what do those who work in public health contend with that would not be the case in different contexts?
RB / TC / DD / GG: Some aspects of the study are widely applicable: trust matters for public health because it increases the likelihood that citizens will take actions that are costly and disruptive for them as individuals, but necessary for containing the spread of a deadly disease like COVID-19. It is also the case that in many if not all countries, public officials can play a crucial role in encouraging individuals to adhere to public health regulations. What is more context specific, is the identity of those public officials who are most consequential. We hope that our study will encourage other researchers to explore which actors and agencies resonate most with individuals in different contexts. This is important not only for theory, but also for policy, as we may wish to design interventions with an eye toward employing the most effective “messenger” for a given setting.