Who Should I Share Risk with? Gifts can tell: Theory and Evidence from Rural China

Ruixin Wang

Research output: Working paperDiscussion paperOther research output

577 Downloads (Pure)

Abstract

This paper studies how gift exchange may help to overcome limited commitment problem in risk sharing. When efficient contract enforcement is lacking, people rely on friends (or relatives) to share risk since emotional or moral cost of defaulting between friends can help to prevent moral hazard. The problem is how to distinguish between friends and non-friends? Gift expense serves as a signal of friendship since giving a gift is less costly for a friend than a non-friend due to altruism. The model re-evaluates the role of gift exchange in developing economies, and helps to rationalize the large amount of gift exchange in China (10% of living expenditure). As a signal, gift exchange improves the efficiency in risk sharing and facilitates favor exchange, but I also demonstrate that the welfare gains due to this improvement may be offset by increased inequality. By using a unique data set containing detailed records about gift exchange in rural China, the empirical study suggests gift expenses, as a signal, significantly increase the probability of risk sharing. I also show further empirical evidence to the theory by testing more model predictions.
Original languageEnglish
Place of PublicationTilburg
PublisherCentER, Center for Economic Research
Number of pages45
Volume2016-003
Publication statusPublished - 19 Jan 2016

Publication series

NameCentER Discussion Paper
Volume2016-003

Keywords

  • gift exchange
  • risk sharing
  • emotinal collateral
  • signaling

Fingerprint Dive into the research topics of 'Who Should I Share Risk with? Gifts can tell: Theory and Evidence from Rural China'. Together they form a unique fingerprint.

  • Cite this

    Wang, R. (2016). Who Should I Share Risk with? Gifts can tell: Theory and Evidence from Rural China. (CentER Discussion Paper; Vol. 2016-003). CentER, Center for Economic Research.