The Perfect Enemy | Public support in the United States for global equity in vaccine pricing | Scientific Reports
July 1, 2022

Public support in the United States for global equity in vaccine pricing | Scientific Reports

Public support in the United States for global equity in vaccine pricing | Scientific Reports

Read Time:8 Minute

Study sample and procedure

The cross-sectional survey experiment was conducted in the US in April and May 2021. The experiment was programmed in Qualtrics and conducted via Prolific, a UK-based online research platform. The final sample included 803 adult participants living in the US who represented the adult US population aged 18 and over in terms of age, sex, and race (see Supplementary Table S2).

Ethics declaration

The survey experiment (Project ID 27,264) was approved by the Monash University Human Research Ethics Committee (MUHREC). The survey experiment was performed in accordance with MUHREC’s guidelines and regulations. Informed consent was obtained from all participants.

Design of experiment

Participants were randomly assigned to one of eight treatment groups (two vaccine cost conditions × four information conditions). Each treatment group had roughly 100 participants. The experiment consisted of several parts (see Supplementary information). After providing informed consent, participants answered a pre-experiment survey containing two questions about their attitudes towards inequality, five questions about their experiences during the COVID-19 pandemic, and a total of seven questions about their baseline knowledge of factors relevant to vaccine pricing (e.g., whether lower-middle income countries are eligible for GAVI support to access vaccines). Since the study was conducted during the COVID-19 pandemic, it was important to check whether exposure to COVID-19 and attitudes towards vaccination were balanced across all groups.

Participants then read information about the basic concepts and choice situations in the experiment, completed a quiz designed to assess their understanding of the experimental instructions, read information in their randomly assigned condition, completed the tasks measuring individual and collective-level support for TEP relative to other pricing strategies, and provided basic socio-demographic information. Participants across the eight treatment groups were balanced in terms of exposure to Covid-19 and attitudes towards vaccination as elicited in the pre-experiment survey, as well as their demographic characteristics. Further, baseline knowledge of factors relevant to vaccine pricing was generally poor and balanced across the eight treatment groups (see Supplementary Table S2).

Cost and information conditions

To examine how vaccine cost and information related to equity and profitability considerations interact, we chose two cost conditions and four information conditions. We needed at least two cost conditions for a meaningful study. The low-cost condition was set at USD 10 per dose, while the high-cost was set at USD 50 per dose. These cost levels are based on price data reported by WHO, where the highest price paid by poorer countries was about USD 10 per dose and the highest price paid by richer countries was about USD 50 per dose23. Alternatively, the low-cost conditions may be viewed as reflecting as scenarios where intellectual property rights have been temporarily suspended or permanently expired, or the availability of generic brands at later stages in the product cycle24. For both cost conditions, the production (marginal) and pre-production (fixed) cost were set at 30% and 70%, to reflect the relatively high costs of vaccine R&D25,26,27.

The four information conditions are: Arguments, Facts, Arguments + Facts, and NoInfo. Our choice of the nature and number of information conditions was guided by potential heterogeneity in what it takes to influence people’s opinion: empirical facts, conceptual arguments, or both. The Arguments condition provided arguments (without supporting facts) in favor of ensuring both affordability for poorer countries and profitability for manufacturers (Table 1). The advantage of these arguments is that they are relatively simple and can potentially help individuals reason the implications of different decisions. The Facts condition provided statistics about the four country groups relevant to equity and profitability considerations in global vaccine pricing (Table 1). Knowledge of such facts could be useful in understanding the relevance and importance of importance of equity and profitability considerations in global vaccine pricing. The Arguments + Facts condition combined the Arguments and Facts conditions by using the statistics from the Facts condition to justify the arguments in the Arguments condition. Participants in the NoInfo condition were not provided any arguments or facts. The four information conditions along with the two cost conditions constitute the total eight treatment groups in our study. This led us to select roughly 100 participants per treatment group because in order to generate a demographically representative sample of the adult US population within the study period.

Table 1 Information provided in the Arguments condition and Facts condition.

Choice scenario

Participants were presented with a hypothetical scenario where, in a global pandemic, a firm with the ability to supply the whole world had developed a safe and effective vaccine that had been approved by health authorities. This framing was used to encourage participants to focus on the tension between equity and profitability, and minimize concerns about vaccine safety, effectiveness, or supply. The instructions to the participants did not explicitly mention this vaccine was for COVID-19. However, participants are likely to have been thinking about vaccines for COVID-19 while making their choices given the study was conducted during the COVID-19 pandemic. We therefore took care to ensure participants were randomly assigned to the treatment groups and checked for balance in demographic characteristics and baseline knowledge about factors relevant to global vaccine pricing. Thus, any priming due to COVID-19 is unlikely to impact the validity of our findings.

Participants chose the price per dose of vaccine that the firm should charge for each of the four country groups according to the income classification by World Bank: low-, lower-middle, upper-middle, and high-income countries28. For each country group, participants were asked to choose one out of the following six price categories (from lowest to highest): (1) below marginal cost, (2) equal to marginal cost, (3) above marginal cost but significantly less than total cost, (4) significantly above marginal cost but less than total cost, (5) equal to total cost, (6) above total cost. It was emphasized that prices refer to prices paid by a country to the manufacturer, and that governments may choose to offer vaccines to their citizens for free. After choosing the vaccine prices for all four country groups, participants answered a question about whether the firm should make an “overall profit”. They could respond either “Yes”, “No”, or “Unsure”. Participants answered these questions in an individual task measuring the individual-level (personal) views of participants.

After the individual task, participants answered the same questions in a coordination task designed to reveal the pricing strategy most likely to achieve collective agreement. For each coordination decision, a participant could earn a bonus of GBP 0.10 if their decision matched the most frequent decision among all participants.

In both the individual task and the coordination task, participants were asked to make the same decisions about vaccine pricing in relation to the same hypothetical scenario. However, there is a fundamental difference between the two tasks in terms of what they encourage participants to think about before making their choices. A participant does not need to think about the views of other participants about vaccine pricing while responding in the individual task, as it is designed to elicit an individual’s personal views and preferences. In contrast, the coordination task is designed to encourage participants to think about the views of others. Each participant knows they will receive a bonus payment if their reported pricing choices match with the most frequent pricing choices among all the participants. This feature likely encourages participants to think about the views of others, rather than their own personal preferences. The choices made in the coordination task reveal what people with potentially different personal preferences are most likely to collective agree upon29.

Outcome measures

Our primary outcome was support for TEP as measured by three components: tiering of prices across countries, equitable prices for poorer countries, and profitability for manufacturers. As tiered pricing requires poorer countries to be charged a lower price than richer countries, a participant was deemed to support tiered pricing if two conditions were met: (i) the prices for low- and lower-middle income countries were lower than the price for upper-middle income countries, (ii) the price for upper-middle income countries was less than or equal to the price for high-income countries. Equitable pricing was defined as prices for low- and lower-middle income countries that were close to the production (marginal) cost, i.e., below marginal cost, equal to marginal cost, or above marginal cost but significantly less than total cost12,17,19. Profitable pricing was defined as a participant who responded “Yes” to firms making overall profits.

We coded a participant as supporting TEP only if they supported all three components. Otherwise, we classified choices as falling within three alternative pricing strategies: support for tiered pricing and equitable pricing but not profitable pricing (TEnotP), support for tiered pricing and profitable pricing but not equitable pricing (TPnotE), and a residual category capturing all other decisions (Other).

Data analysis

We first used participants’ responses to seven questions about prior familiarity with GAVI, COVAX, and factors relevant to vaccine pricing to assess whether participants had “sufficient baseline knowledge” about factors that are relevant to TEP (defined as score of at least four out of seven). Only 17% of participants had sufficient baseline knowledge. This low level of baseline knowledge is statistically similar across all groups (see Supplementary Table S2).

We examined the effects of an increase in vaccine cost and provision of information by comparing the level of support for a pricing strategy in each treatment group to the LowCost × NoInfo baseline group. We consider the LowCost × NoInfo baseline group as the control group where the participants in this group face a low-cost vaccine and possess a low level of information relevant for TEP. Given the uniformly low fraction of participants with sufficient baseline knowledge in all cost and information conditions, the treatment differences relative to LowCost × NoInfo control group are interpreted as the impact of “increasing vaccine cost” and/or “increasing information” on public support for TEP.

Treatment effects were estimated using multivariate ordinary least squares regression. We included controls for comprehension of experimental instructions, whether participants paid attention during the experiment. In addition, we controlled for whether a participant had sufficient baseline knowledge of factors relevant to vaccine pricing. T-tests were used to evaluate treatment effects and differences between groups. We considered p values of 0.10 or less to be significant. Statistical analyses were conducted using STATA SE, version 16.