The Russia-Ukraine conflict significantly impacted global food security, particularly for low-income countries reliant on grain imports. Russia and Ukraine are major global grain exporters, with their 2021 exports of wheat, barley, and maize accounting for 28%, 26%, and 15% of the global share, respectively. These grains provide a substantial portion of global caloric and protein intake. The conflict disrupted exports from both countries, leading to a global food supply crunch and increased risk of poverty and hunger. This study examines the cascading effects of this disruption on the global grain trade network, considering the interconnectedness of countries and the role of trade in both mitigating and propagating shocks. The integrated nature of the global market should ideally facilitate risk diversification, yet high-intensity shocks can lead to severe vulnerabilities. Exporting countries may impose trade restrictions to stabilize domestic supplies, further impacting global supply. This necessitates a model that captures the propagation of shocks along trade links to understand both direct and indirect dependencies between countries and quantify the overall supply gap.

Existing economic literature emphasizes the complexity of system responses to external disturbances. Studies on financial networks highlight the role of network connections in crisis contagion. Research on grain supply systems recognizes the complexity of international trade and its impact on vulnerability to shocks. Previous work has explored the adverse effects of natural (weather-related) and human factors (conflicts, policies) generating grain trade shocks, leading to global supply shortages and food insecurity. These studies often focus on linear stability or reactivity analysis, quantifying immediate impacts or using unweighted networks, thereby overlooking the cascading effects of higher-order dependence and the diverse national responses to shocks. The limitations of existing research include the infrequent consideration of compensatory import demand, a singular focus on national responses, limited inclusion of a country's level of economic development and its impact on shock resilience, and the use of short-term data sets that don't reflect long-term development trends.

This study constructs three weighted and directed global trade networks for wheat, barley, and maize using FAO trade data from 1995 to 2021. The networks consist of nodes representing countries and directed edges representing bilateral trade flows, weighted by export quantities. The study employs a cascading failure network model to simulate the propagation of shocks resulting from the disruption of Russian and Ukrainian grain exports. The model accounts for annual production, import and export quantities, and assumes that countries maintain their trading partner preferences. When a shock occurs (reduction in imports), importing countries compensate by shifting demand to other trading partners. Exporting countries receiving increased demand may absorb the shock by increasing exports or may fail if their domestic supply space is exhausted. This failure then triggers further cascading failures in their importing partners. The model incorporates the effect of GDP per capita, reflecting different country capacities to absorb shocks and respond to price fluctuations. Import demand shocks are distributed among trading partners proportionally to their trade volume and GDP per capita. The model simulates the cascading process over time steps within a year, capturing the diffusion of shocks and revealing indirect dependencies. To assess national-level vulnerability, the study calculates supply reduction ratios (SR) for each country, representing the percentage of unmet import demand relative to domestic supply. The number of poor people affected (PA) is estimated by multiplying SR by the country's poverty rate and population. Regression analysis is performed using a panel Tobit model and fixed effects model to determine the relationship between network characteristics (in-degree, out-degree, betweenness centrality, eigenvector centrality, clustering coefficient, import concentration) and country attributes (GDP, population, number of trade agreements, regional location) and vulnerability to supply shocks.

The analysis reveals several key findings: 1. **Russia and Ukraine's Central Role:** The backbone networks of the three grains show a significant improvement in Russia and Ukraine's trade status. The concentration of import trade increased during the COVID-19 pandemic. Russia and Ukraine became major wheat and barley suppliers, directly impacting 77 countries in Africa and Asia and many others indirectly. Lower middle-income countries in West Asia, North Africa, and Southeast Asia had the highest dependence on direct imports from Russia and Ukraine. 2. **Network Topology Disruption:** The simulation of cascading failures illustrates how the failure of nodes due to trade shocks disrupts network topology. Connectivity decreased significantly in post-shock networks, particularly in wheat, barley, and maize trade networks after 2011, 2013, and 2009, respectively. The COVID-19 pandemic impacted connectivity differently across grains, illustrating varying resilience. 3. **Import Demand Shifting and Supplier Pressures:** The analysis of positive shocks from increased import demand reveals the major grain suppliers such as the US, Canada, France, Argentina, and Brazil absorbing much of the demand. Post-COVID-19, Australia saw a significant increase in demand, particularly for wheat and barley. Eastern and Southern European countries often acted as intermediaries, propagating shocks. 4. **Global Decline in Grain Supplies:** The simulation of cascading failures reveals the extent of supply shortages across importing countries. North Africa, West Asia, and Southeast Asia faced the most significant wheat shortages. The COVID-19 pandemic further exacerbated these shortages. Similar patterns are seen for barley and maize shortages, with geographical variations in the most affected regions. 5. **Regression Analysis Results:** Regression analysis demonstrates that import channel diversity is effective in reducing vulnerability to wheat and barley shocks but less so for maize. Betweenness centrality played a crucial role in shock propagation, suggesting that countries serving as trade intermediaries face higher vulnerability. Import concentration negatively correlates with vulnerability, emphasizing the importance of diversification of grain import sources. The population size of a country has a positive correlation with its vulnerability to wheat supply shocks. Countries in Asia and Africa, especially, are vulnerable to supply shocks in wheat and barley, while European countries show the most vulnerability in the maize trade.

The findings highlight the cascading nature of grain trade shocks from the Russia-Ukraine conflict, impacting countries both directly and indirectly. The study's results directly address the research question by quantifying the impact of the conflict on global food security, showing that lower middle-income countries are disproportionately affected due to their limited ability to leverage trade networks for supply diversification. The study’s significance lies in its novel methodology, which integrates network science with economic analysis to comprehensively assess the interconnected nature of global food systems. It contributes to a better understanding of systemic risks and offers policy recommendations for strengthening resilience against such shocks. The results underscore the need for proactive measures to prevent future crises.

This study contributes significantly to our understanding of global food insecurity by quantifying the cascading effects of the Russia-Ukraine conflict on grain trade networks. It highlights the differentiated vulnerabilities of countries based on their network positions and attributes. The policy recommendations, emphasizing import diversification, regional cooperation, strategic grain reserves, and production capacity enhancement, are crucial for building more resilient food systems. Future research could incorporate the effects of price dynamics and substitution between grain types for a more comprehensive assessment.

The model simplifies several aspects of the complex global grain trade system. It does not explicitly model price fluctuations, substitution effects between different grain types, or the impact of other exogenous factors beyond export disruptions. The model uses annual aggregated trade data, potentially overlooking shorter-term fluctuations. Additionally, the model assumes a stable preference for trading partners, which may not fully capture dynamic adjustments in trade relationships. Future research should aim to incorporate these factors for a more nuanced understanding of cascading failures in the global grain trade network.