A fundamental question for economic analysis is how successful an economy is at delivering welfare for its members. According to virtually all cultural traditions, non-instrumental social relations, which we refer to as relational goods, are indeed necessary for a good life. However, relational goods are quintessentially gratuitous and cannot therefore be exchanged through markets or produced by the State. For these reasons their social value cannot enter GDP calculations and risks being ignored in our public discourse. In the past the quantitative evaluation of the impact of relational goods on welfare was just impossible. This is no longer true thanks to the increasing availability of data on subjective well-being (and on social activities themselves). Findings from these data strongly confirm the importance of social ties for our welfare and urge us to give up the anthropological reductionism still characterizing much of economic research.