Multinomial logit model of scheme choice.

To assess the statistical significance of the observation in Table 2 while controlling for session effects and socio-economic characteristics, we estimate the multinomial logit model outlined in Equation 1, adding the variable Envious as control, in order to control for a general tendency to prefer tournaments. We observe that EPs do not chose the tournament more often when perceiving the Thuringian accent, but there is a significant increase in tournament take up over all tasks when EPs perceive the Bavarian accent. Standard errors in parentheses (clustered on the individual level in model (1)). Multinomial logit model, dependent variable Choice of payment scheme;*p **p ***p Multinomial logit model of scheme choice.

为评估表2中观测结果的统计显著性，同时控制实验环节效应与社会经济特征，我们基于公式1所概述的多项logit模型（multinomial logit model）进行估计，并引入变量Envious作为控制变量，以控制个体普遍偏好锦标赛的行为倾向。我们观察到，当EPs感知到图林根口音时，并未更频繁地选择锦标赛方案；但当EPs感知到巴伐利亚口音时，所有任务中的锦标赛方案选择率均出现显著提升。括号内为标准误（模型(1)的标准误采用个体层面聚类调整）。本模型为多项logit模型，因变量为支付方案选择；*p、**p、***p，即方案选择的多项logit模型。