Poor economies not only produce less; they typically produce things that involve fewer inputs and fewer intermediate steps. Yet the supply chains of poor countries face more frequent disruptions---delivery failures, faulty parts, delays, power outages, theft, government failures---that systematically thwart the production process. To understand how these disruptions affect economic development, we model an evolving input--output network in which disruptions spread contagiously among optimizing agents. The key finding is that a poverty trap can emerge: agents adapt to frequent disruptions by producing simpler, less valuable goods, yet disruptions persist. Growing out of poverty requires that agents invest in buffers to disruptions. These buffers rise and then fall as the economy produces more complex goods, a prediction consistent with global patterns of input inventories. Large jumps in economic complexity can backfire. This result suggests why "big push" policies can fail, and it underscores the importance of reliability and of gradual increases in technological complexity.Comment: Main text: 10 pages, 6 figures. Supplementary Information: 15 pages, 1 figur

In this paper we study a situation in which agents embedded in a network simultaneously play interrelated bilateral contest games with their neighbors. Spillovers between contests induce complex local and global network effects. We first characterize the equilibrium of the game on a given network. Then we study a network formation model, introducing a novel but intuitive link formation protocol. As links represent negative relationships, link formation is unilateral while link destruction is bilateral. The unique stable network topology is a complete K-partite network with partitions of different sizes. Stable networks exhibit properties that are in line with empirical and theoretical findings from other disciplines.

In this paper, we introduce the Input Rank as a measure to study the organization of global supply networks at the firm level. We model the case of a firm that needs assessing the technological relevance of each direct and indirect supplier on a network-like production function with labor and intermediate inputs. In our framework, an input is technologically more relevant if a shock on that upstream market can hit harder the marginal costs of a downstream buyer, considering the topology of the supply structure. A higher labor intensity at each stage buffers the transmission of upstream shocks in the network. In addition, we provide for the possibility that producers have limited knowledge of inputs in the supply network, hence they can underestimate the relevance of more distant inputs. After applications, the Input Rank returns a matrix of technological centralities that order any direct or indirect input for a representative firm in any output industry. We compute the Input Rank on U.S. and world input-output tables. Finally, we test how it correlates with choices of vertical integration made by 20,489 U.S. parent companies controlling 154,836 affiliates worldwide. We find that a higher Input Rank is positively associated with higher odds that that input is vertically integrated, relatively more when final demand is elastic. A supplier's Input Rank remains a significant predictor of a firm's decision to integrate even after controlling for the relative positions on upstreamness/downstreamness segments.

Network structure has a significant role in determining the outcomes of many socio-economic relationships, including the antagonistic ones. In this paper we study a situation in which agents, embedded in a network, simultaneously play interrelated bilateral contest games with their neighbours. Spillovers between contests induce complex local and global network effects. We first characterize the equilibrium of a game on arbitrary fixed network. Then we study a dynamic network formation model, introducing a novel but intuitive link formation protocol. As links represent antagonistic relationships, link formation is unilateral while link destruction is bilateral. A complete k-partite network is the unique stable network topology. As a result, the model provides a micro-foundation for the structural balance concept in social psychology, and the main results go in line with theoretical and empirical findings from other disciplines, including international relations, sociology and biology.