Stocks of information in personal consumption: a network model with non-rival borrowing and content overlap

Abstract

Stocks of information that cumulate in personal consumption and other non-market contexts have unrecognized welfare implications through properties that include non-rival borrowing and relatively low obsolescence rates. Although non-rival borrowing is emergent in agent interaction, it is typically not included in agent heuristics. Properties such as non-rival borrowing are best studied in networks. Computational studies of both regular networks and small world networks indicate the clustering that these networks tend to. In spite of its cited welfare relevance to distributional inequality, clustering in small world networks has only been examined in terms of the remoteness parameter of the network. We extend a network model of the stock of information to more explicitly represent efficiency-reducing effects that clustering can have through content duplication or overlap and demonstrate the significance these effects can have in computational results. These studies also show the efficiency-increasing effects of non-rival borrowing that continue to be evidenced and the overlap reducing effects that increasing network remoteness can have.