Figure 4.

Country-level effects on moral preferences. (a) In Study 1, we collected data from 10 countries where the impact of the pandemic has been highly dissimilar at the time when the study was performed (May 2020). The map displays the countries from where we have obtained data. Colours show the per capita total number of confirmed COVID-19 deaths by the end of data collection. Country labels are Uruguay (URU), Argentina (ARG), Colombia (COL), Honduras (HOR), Bolivia (BOL), Chile (CHI), Mexico (MEX), Panama (PAN), Peru (PER) and Ecuador (ECU). (b) Projection onto the first two principal components (PCs) for participants proceeding from different countries. Participants were grouped based on the number of COVID-19 deaths per million people at the country from where they proceed. Less than 10: URU. Between 10 and 50: ARG, BOL, COL and HON. Between 50 and 100: CHI and MEX. More than 100: PER and ECU. Bars denote average PC score across participants, and vertical lines depict SEM. (c) We performed a multivariate regression analysis on the scores obtained for each PC. Circles show coefficient estimates for the first PC (Human Life Expectancy) and squares show coefficient estimates for the second PC (Equitable Public Health). Predictor variables displayed are Instrumental Harm (IH), Impartial Beneficence (IB) and the country-level pandemic's severity, quantified as the COVID-19 deaths per million people. Vertical lines show SEM and colours code the t-value associated with each coefficient. For the full list of predictor variables included in the analysis, see electronic supplementary material, tables S2 and S3.