This paper examines the importance of relative deprivation in Tanzania, a poor African country, using three waves of the Tanzanian National Panel Survey. We contribute to earlier literature in Africa by controlling for time persistent unobservable individual characteristics (panel data) and, most important, by using two measures of satisfaction (life and financial satisfaction), two definitions of the reference group, and testing different hypothesis. By comparing results between satisfaction measures and across definitions of reference groups we can understand the mechanisms through which comparisons work in a poor setting. In contrast with earlier literature, we find strong evidence of relative deprivation in financial satisfaction of all individuals in Tanzania, and evidence for life satisfaction only for individuals with weaker ties with their community. For those with strong ties (45% of our total sample; 52% of the rural sample), we find evidence of a positive correlation between life satisfaction and the leave-out mean consumption of close neighbors and argue that this can be explained by feelings of empathy. We argue in favor of taking comparisons to others into account when assessing and introducing welfare policies, also in less developed countries. Ada Ferrer-i-Carbonell acknowledges financial support from the Spanish Ministry of Economy and Competitiveness through the Severo Ochoa Programme for Centres of Excellence in R&D [SEV-2015–0563]; the AEI/FEDER, UE [PID2020–114251GB-I00]; and the AGAUR-Generalitat de Catalunya [SGR2017–1359]. We gratefully acknowledge the comments from two anonymous referees, as well as from Scott Kostyshak, Joan Llull, and the participants of the micro seminar at the University of Florida. Peer reviewed