Study On The Voluntary Contribution Behavior Of Public Goods Based On Social Networks

Since the 1970s, because of the requirements of social economic development, economists became to focus on the private provision of public goods, and the voluntary contribution mechanism has become the third force in public good provision besides the government and the market. Therefore, this dissertation takes the voluntary contribution behavior of individual as the research subject, and focus attention on the effect of social networks. Based on the current research literature, this dissertation firstly collects the primilary data by laboratory experiments, which is used to directly test the effects of social network structures on individual contributions, and analyze how the social networks affect individuals'behavior. After that, we modify the assumptions of the theoretical model and finds out the strict Nash equilibrium when the cost of link is different.Firstly, individual behavior in the process of public good supply is very sensitive to the environment with much that can affect outcomes but are difficult to control, so we use the laboratory experiments to test the effects of social networks on individual behavior. Before designing the experiment, we review the theoretical literature and get the main reasons to suspect that cooperation should depend on network structure, and simplify the network model of public goods in Bramoulle and Kranton (2007). After this, we design 4 treatments to investigate directly the relationship between network structure and individual behavior, the only difference between treatments is the network structure on which the game played. We find evidence that individuals'strategic decision is bounded rationality, and individual behavior is significant influenced by the network structure, such as the network density and regularity.Secondly, the dissertation tries to find out how the social networks affect the individual behavior, to learn from psychological game theories which are based on belief-dependent motivations, we take elicited belief as the intermediate variable between a series of social networks and individual behavior. The results reveal that network structure exhibits a significant effect on individual elicited belief, which results in the change of individual behavior model. Specially, on the non-regular networks, the inequality between the players leads individuals who are the central nodes to act as a "leader" in the repeated games, which is significantly different with the result of some literature that most people seems to exhibit a selfish bias, i.e. their belief of others'contributions exceed their own contributions, and the behavior of individuals who are the peripheral nodes is as the same with the selfish bias individuals.Finally. this dissertation extends the network formation model of public goods developed by Galeotti and Goyal (2010), and makes it more applicable to the general case. Taking the non-competitive information for example, considering the existence of frictions in communication through social networks, we extend the model to allow for the benefits decay with network distance between players, and analyze the impact of decay on the strict Nash equilibrium strategies among identical individuals when the cost of linking with other players is different. We find out the existence of benefits decay leads to a change in the strict Nash equilibrium strategies of individuals, when the cost of link is moderate, the players prefer to have a short distance or directly relationship between them, which leads the equilibrium network to have small diameter; when the cost of link is in the extreme case, the existence of benefits decay has no effect on the strict Nash equilibrium strategies of individuals, the equilibrium networks are as same as before.