We extend the utility specification in Ballester et al. (2006) to study social interactions when individuals hold altruistic preferences in social networks. We show that more enriched network features can be captured in the best response function derived from the maximization of the extended utility which incorporates altruism, providing microfoundation for studies which investigate how network features mediate peer effects or other important features in social interactions. We demonstrate that often ignored altruism is another serious confounding factor of peer effects. Our results show that the estimates of peer affects are about 36% smaller under altruistic preferences. Furthermore, we are able to separate the two different types of effects generated by peers: the (usually positive) spillover effects from peers and the (negative or positive) externality effects directly generated from peers’ outcomes, which is not possible in conventional social interactions model based on self-interest hypothesis. JEL Classification: I10 I20 C31

中文翻译：

---

社交网络中的社交互动和社交偏好 †

我们扩展了 Ballester 等人的实用程序规范。(2006) 研究当个人在社交网络中持有利他偏好时的社交互动。我们表明，可以在结合利他主义的扩展效用最大化的最佳响应函数中捕获更丰富的网络特征，为研究网络特征如何调节同伴效应或社交互动中的其他重要特征的研究提供微观基础。我们证明，经常被忽视的利他主义是同伴效应的另一个严重混杂因素。我们的结果表明，在利他偏好下，同伴影响的估计值要小 36%。此外，我们能够将同行产生的两种不同类型的影响分开：来自同伴的（通常是积极的）溢出效应和由同伴结果直接产生的（消极或积极的）外部效应，这在基于自利假设的传统社会互动模型中是不可能的。JEL 分类： I10 I20 C31