This work studies an important departure from the classical networked economy. In the benchmark case, an external decision-making observer has full information about the networked economy. In this work, the observer does not have full information, yet must make decisions impacting the economy's agents. Specifically, the observer does not know how the attributes of the economy sit on the network's nodes. This work develops a complete, closed-form statistical approach that enables the observer to overcome this lack of information and still execute a decision. The observer must consider all possible arrangements of attributes on the network nodes. By exhaustively rearranging attributes, the observer can construct probability distributions that accurately characterize the economy. In this work, we first develop the necessary theoretical tools and we then show how the observer can employ these tools in the following settings: (1) education with peer effects, (2) consumption with network externalities, and (3) crime.
We develop a theoretical framework and an accompanying set of tools for mapping the topologies of networks in the economy to different probability distributions of interest. We apply these tools to analytically show how the topology of an agent interaction network enables non-fundamental fluctuations in aggregate macroeconomic sentiment, thereby providing microfoundations for animal spirits; as the network's topology changes, we can compute how the shape of the corresponding distribution of aggregate sentiment adjusts. We can moreover apply these tools to carry out closed-form analysis of complex economic systems and to construct error bounds about the paths of aggregated networked economies.