The paper presents a model for cascading failures in complex networks, focusing on the dynamical redistribution of flow following the initial breakdown of a single node. The authors demonstrate that the failure of a node can lead to a collapse of the entire system if the node is among the most heavily loaded nodes. This is particularly relevant for networks with highly heterogeneous load distributions, such as the Internet and electrical power grids. The model is based on a network's efficiency, which is defined as the average efficiency of the most efficient paths between nodes. The breakdown of a node triggers a redistribution of loads, leading to overloads and eventually a significant drop in network performance. The study uses both artificial and real-world networks, including Erdös-Rényi random graphs and scale-free networks, to illustrate the model's effectiveness. The results show that the network's stability is significantly influenced by the heterogeneity of node loads, with scale-free networks being more vulnerable to cascading failures compared to random graphs. The findings provide insights into the mechanisms behind large-scale failures in complex systems and highlight the importance of considering cascading failure effects in network design.The paper presents a model for cascading failures in complex networks, focusing on the dynamical redistribution of flow following the initial breakdown of a single node. The authors demonstrate that the failure of a node can lead to a collapse of the entire system if the node is among the most heavily loaded nodes. This is particularly relevant for networks with highly heterogeneous load distributions, such as the Internet and electrical power grids. The model is based on a network's efficiency, which is defined as the average efficiency of the most efficient paths between nodes. The breakdown of a node triggers a redistribution of loads, leading to overloads and eventually a significant drop in network performance. The study uses both artificial and real-world networks, including Erdös-Rényi random graphs and scale-free networks, to illustrate the model's effectiveness. The results show that the network's stability is significantly influenced by the heterogeneity of node loads, with scale-free networks being more vulnerable to cascading failures compared to random graphs. The findings provide insights into the mechanisms behind large-scale failures in complex systems and highlight the importance of considering cascading failure effects in network design.