The impact of risk propagation on the robustness of R&D networks: A higher-order interaction perspective

This paper presents a new approach to analyzing risk propagation in R&D networks by incorporating higher-order interactions. Risk propagation in these networks greatly influences the success of collaborative R&D efforts. However, previous research on risk propagation has mainly focused on binary interactions, neglecting the role of higher-order interactions and their effects on the propagation of risk. To address this gap, this paper employs Simplex to model the multi-dimensional relationships within R&D networks, constructing higher-order R&D networks and measuring higher-order effects. Based on this, we have, for the first time, built a higher-order network-based load-capacity model to examine cascading risk propagation in R&D networks, addressing risk load, capacity, and propagation mechanisms. Numerical simulations using both theoretical models and real-world R&D networks show that, although higher-order interactions do not fundamentally alter the basic characteristics of risk propagation in R&D networks, they do increase the risk control threshold. Inter-firm higher-order interactions broaden the risk propagation paths, increasing the range of risk spread, while inter-risk higher-order interactions speed up the failure process of individual firms, thereby accelerating risk propagation across the network. Ignoring higher-order degrees results in an expanded risk propagation range, prompting firms to raise their risk control thresholds. Additionally, the study shows that different attack strategies demonstrate that, due to higher-order interactions, risk propagates more quickly through the R&D network, leading to larger-scale failures. This research offers valuable theoretical and practical insights for both academics and business managers, helping them better understand the cascading propagation of risks within R&D neworks.