Justification of the COVID-19 vaccine distribution: Contested inclusiveness in essential resource allocation

How to cite

Nakabayashi, Masaki; Kawata, Keisuke; Iida, Takashi. 2021. "Justification of the COVID-19 vaccine distribution: Contested inclusiveness in essential resource allocation." AEA RCT Registry. AEARCTR-0008105. https://doi.org/10.1257/rct.8105-1.2000000000000002

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Masaki Nakabayashi* , Keisuke Kawata , Takashi Iida

* Primary investigator

Registry ID
AEARCTR-0008105
DOI
https://doi.org/10.1257/rct.8105-1.2000000000000002
First registered
Last updated
Status
Completed
Trial dates
2021-11-08 – 2021-11-24
Countries
Japan ()
JEL codes
I12; J15; I18
Registry record
http://www.socialscienceregistry.org/trials/8105

Abstract

The COVID-19 pandemic is a challenge to all global communities. To overcome it, successful cooperation is critical. At the same time, essential resources such as safe and effective vaccines against COVID-19 and intensive care units in hospitals are scarce resources. For a nation to survive, such scarce resources should be available to all residents, including foreigners. However, the competition of individuals for scarce resources might lead to deepened social divides. Whether a nation can distribute scarce resources in a manner considered optimal for society as a whole, overcoming the conflicts of individual interests, would be an informative measure of that nation's inclusiveness, an ability that affects its survival.

Our experiment is set amid the natural experimental environment of the COVID-19 pandemic in Japan. Japan's capacity for vaccine production is still tightly limited as emergent variants of COVID-19 might necessitate three or more doses of a vaccine. We investigate whether respondents want to allocate COVID-19 vaccines equally among residents of Japanese and foreigners with the same background characteristics, in accord with standards that the government has applied.

We employ a randomized conjoint design. Respondents are shown three hypothetical COVID-19 vaccine recipients, each with different background characteristics, including nationality and residency status. We then asked the respondents to place three potential recipients in order of priority allocation. We observe a marginal change in the preferred allocation of vaccines when nationality and/or residency status is changed.

Intervention

We adopt a factorial design as a randomized conjoint design. We employ 15,000 respondents in our internet-based experiment and task them with performing a randomized conjoint experiment. In each case, respondents are shown three hypothetical COVID-19 vaccine recipients with different background characteristics, including nationality and residency status. Among those hypothetical recipients, the Japanese nationality is fixed. We then observe whether and how a change in the nationality of a hypothetical recipient affects the respondent's selection of the person in relation to recipients with the same and different background characteristics.

Experimental design

We implement a randomized conjoint experiment using 15,000 randomly selected Japanese adult respondents. We assign each respondent 5 tasks; in each, they are asked to place three hypothetical vaccine recipients (whose occupation, age, residency status, family composition, and nationality are randomly combined except that one of the three choices is always Japanese) in order of priority. Responses to Japanese recipients are our control group, and responses to the other recipients are our treatment group.

We collect information about the background characteristics of the 15,000 respondents such as gender, age, prefecture, work status, political preferences, partisanship, income, household income, education, whether reading a newspaper, Internet news, or books, and self-perception of social class. Using a generalized random forest algorithm, we look at whether the effects of the intervention differed based on the background characteristics of the respondents.

Design details

Randomization method
We implement a randomized conjoint experiment. Randomization is performed by a computer of the survey company with which we contract.
Randomization unit
Individual.
Clusters
15,000 individuals.
Observations
15,000 individuals.
Arms
(15,000 individuals) X (5 choice tasks) X (3 ordered preferences)
Primary outcomes
The outcome of the greatest interest is whether foreign nationality affects respondents' preferences regarding hypothetical vaccine recipients, and, if so, whether the difference in preferences between foreign nationality and own nationality is correlated with geopolitical tension with the home country. If respondents are perfectly rational in their wish to build herd immunity, there should be no discrimination regarding nationality and geopolitical tension with the home country. If there is discrimination on those grounds, this can be a measure of the degree of ethnic and geopolitical distortion shared by Japanese adults facing the pandemic challenges.

Cited by

No citation data available for this registration yet.

Keywords