A large number of observational and experimental studies have explored the determinants of individual preferences for redistribution. In general, inequalities are more likely to be accepted by people of higher socioeconomic status, in richer societies and when inequalities are perceived as justifiable owing to differences in productivity. Almås et al. (2020) show that in a relatively unequal society (the United States), the highly educated accept inequality significantly more than the less educated, whereas, in a relatively equal society (Norway), the less educated accept inequality more, but not significantly more, than the highly educated. Here, we replicate this finding using data from experiments conducted in four locations across three countries all distinct from the ones studied by Almås et al. However, a closer look at the data indicates that the origin of the interaction effect varies depending on which societies one compares. Data for Norway and the United States indicate that meritocratic values among the highly educated are less prevalent in more equal societies and that this is the driver of the triple interaction effect. In contrast, in our data the interaction effects have multiple drivers.

大量的观察和实验研究已经探索了个人偏好重新分配的决定因素。通常，在较富裕的社会中，由于生产力的差异，不平等被认为是合理的，不平等更容易被较高社会经济地位的人们接受。Almås等。（2020年）表明，在一个相对不平等的社会（美国）中，受过高等教育的人接受不平等的程度明显高于受教育程度较低的人，而在一个相对平等的社会（挪威）中，受教育程度较低的人接受不平等的现象更多，但不更多，比受过高等教育的人高。在这里，我们使用在三个国家（与阿尔马斯大学研究的地区）不同的三个地区的四个地区进行的实验数据来重复这一发现等。然而，仔细观察数据表明，互动效应的根源取决于一个人所比较的社会。挪威和美国的数据表明，受过良好教育的精英阶层价值观在更平等的社会中不那么普遍，这是三重互动效应的驱动力。相反，在我们的数据中，交互作用具有多个驱动因素。