Cascading impacts from climatic shocks to food trade chokepoints – an update

Much of the world's staple agricultural commodities must pass through one or more ‘chokepoints’ along maritime trade routes. In 2022, Chatham House published an article, ‘Exploring the cascading impacts from climate shocks to chokepoints in global food trade’, which examined how events related to climate change could disrupt these ‘chokepoints’, and how the impacts might cascade globally, and into Europe in particular. That article considered the modelled impacts of three plausible and discrete disruptions to the Panama Canal, the Turkish Straits, and the Suez Canal and Strait of Bab al-Mandab under a ‘middle-of-the-road’ scenario for 2030 (as articulated through the Shared Socioeconomic Pathway (SSP) 2 scenario). In late 2023, trade through the Panama Canal has been disrupted due to drought and low water levels in much the way that we envisioned.

In 2023, (prior to the events witnessed in the Panama Canal) we undertook a new analysis using updated modelling, with additional SSPs and climate change scenarios. This article reports on the findings of that extended analysis.

The extended analysis considered the same three climate-related hazards but examined the potential economic consequences of them materialising in both 2030 and 2050, and under both SSP2 and the more challenging circumstances of a fragmented global trading regime, reflecting increased regional rivalries (SSP3). We also considered the impacts of the trade shocks under these pathways when accounting for additional climate-change effects on crop yields and the availability of produce for exports. For this, we included the yield impacts under a climate that heats to between 2.0°C and 3.7°C above pre-industrial levels by the end of the century (as reflected by Representative Concentration Pathway (RCP) 6.0) and, for SSP3, also under a climate that warms between 0.9°C and 2.3°C (RCP2.6).

We also updated a simplifying assumption we made in the prior work, namely that recent (2018) bilateral trade dependencies would persevere to 2030. Although changes to the volume of produce traded between countries were reflected in the previous study, it was assumed that countries’ relative exposures to specific disruptions would remain the same, based on the historical proportions of their imports/exports flowing through different chokepoints. That assumption is now revised by disaggregating the world regions used in the global trade model to better reflect how the routing of export–import flows of agricultural commodities might evolve under the different scenarios and time periods. Nonetheless, the shares of global agricultural commodities that transit each chokepoint under all scenarios changed relatively little from those presented in the primary analysis. Finally, we have expanded the analysis of the dynamics triggered by the hazards in agricultural commodity markets to consider how agricultural price impacts relate to both prices and production patterns in the value-added food production sector.

As with the previous analysis, the simulations are based on a general equilibrium model¹ that reports the annualised macro-economic changes from relatively short-lived shocks to the three critical trade chokepoints.

The more fragmented trading regime of SSP3 is simulated in the model by introducing trade frictions that give preference to consuming domestic commodities over imported ones. This contributes to lower overall trade volumes in the SSP3 scenario.

Because the model itself doesn’t encapsulate the chokepoints, the hazard events occurring in the Turkish Straits and Panama and Suez Canals are simulated as productivity shocks that reduce demand for imports. A second, alternative, approach was also used that instead implemented the chokepoint interruptions as tariffs on exports to constrain supplies.

As a consequence of using an equilibrium model, the acute price spikes and other brief but potentially impactful downstream disruptions that are likely to be experienced in the immediate, disequilibrium, aftermath of the hazards are not explicitly captured. Therefore the effects simulated by the model are likely to underplay the magnitude of some of the ‘lived experience’ impacts of these events, and their plausible risk cascades.

For further details on the analytical approach and detailed results, see Key et al. (forthcoming).

Overall, from the ‘one-year’ equilibrium perspective reflected in the model, the results under the extended scenarios are similar to those from the initial analysis. They suggest that the macro-economic impacts, especially for Europe, are likely to be relatively minor, but that there may, nonetheless, be important consequences at the sectoral level. This remains true when the time horizon is extended from 2030 to 2050 and with the additional consideration of climate impacts on agricultural yields.

Economic losses resulting from the shocks as a proportion of gross domestic product (GDP) are small for all countries, with the largest losses being experienced by countries that are net exporters of agricultural commodities, such as Argentina (affected by the Panama Canal disruption), Canada (affected by the Panama and Suez Canal disruptions) and Ukraine (affected by all chokepoint disruptions).

Disruption to the Panama Canal results in the largest GDP losses, reflecting both the importance of this route to global agricultural commodity trade and the larger magnitude of the shock itself (a sensitivity analysis controlling for the size of the shock across chokepoints confirmed the Panama Canal as the most critical of the three trading nodes).

In general across all chokepoints, the GDP losses from the simulated shocks are larger in 2050 compared to 2030 and the picture between the different climate scenarios is mixed but negligible.

Since global trade is less important under the more fragmented SSP3 scenario than the more open trading regime of SSP2, less global trade is restricted by chokepoint disruptions in the fragmented future. This results in varied macroeconomic effects from the shocks depending on how given countries’ and regions’ bilateral trading relationships are reconfigured under SSP3 and how this alters their exposure to chokepoint risks. Global GDP effects (both in magnitude and direction of change) also vary depending on whether the chokepoint hazard events are simulated as productivity shocks or export tariffs, with the latter approach resulting in more widespread negative effects for both importing and exporting countries.

In common with the results from the initial analysis, European price effects across all permutations of the scenarios are very small. Nonetheless, disruptions to the Suez Canal and the Turkish Straits are more significant for Europe than the Panama Canal, especially for rice and wheat trade. In the wake of a chokepoint shock, restricted import flows and increased prices can stimulate a domestic production response in European regions, which are collectively net importers of agricultural commodities. This occurs for rice and oilseed in the wake of the Suez Canal disruption, whereas wheat and other grain production falls alongside prices. Following disruption of the Turkish Straits, contractions of European grain imports are sufficiently small to result in increasing imports from alternative sources only rather than a domestic production response. In a few cases, if the import contraction/price signal is sufficiently large to stimulate significant additional European production volumes, this can also result in an increase in European exports. Although disruptions to the Panama Canal trigger domestic production increases, they are insufficient to compensate for reduced import volumes.

In reality, increased production in Europe would be dependent on conducive climatic conditions in Europe, which is not a given, and would also require a policy environment that allows for fallow or additional land, including land currently dedicated to biodiversity, to be brought into short-term production. This behaviour is not unprecedented; the EU allowed such derogations from the ‘greening’ measures in the EU Common Agricultural Policy following Russia’s invasion of Ukraine, but this step has been criticised for its limited impacts on improving food security and deleterious impacts on biodiversity, so should not be regarded as a no-cost solution to global food trade disruptions.

Because grains and oilseeds are essential inputs for the food industry, their rising prices in the wake of chokepoint shocks can translate into lower food production volumes and increasing food prices for consumers. Although this only occurs to a moderate degree (at most) in all the scenarios modelled, such effects may be potential sources of stress on low-income households that spend a larger proportion of their income on food.

Overall, the balance of these price effects, including as a result of European production responses, are generally sufficient to induce small declines in European production costs, which result in the European food industry slightly increasing its production in response to lower input costs.

The more fragmented global trading regime reflected in SSP3 scenarios tends to have smaller impacts on European agricultural production than in SSP2; the differences between the climate-forcing scenarios (RCP2.6 and RCP6.0) is minimal and more equivocal (as it is for global GDP impacts).

In common with much of Europe, countries and regions that are predominantly net importers of agricultural commodities often experience price increases in the wake of chokepoint shocks. These translate into higher prices and lower production for domestic food industries which require the affected commodities as inputs. Depending on the magnitude of the shock, they frequently also result in increased domestic agricultural production and imports from alternative sources, although prices still tend to rise. In predominantly exporting regions, converse dynamics are frequently observed – demand falls, commodity prices fall and domestic food industries capitalise on cheaper inputs to increase production. Countries that are both buy and sell agricultural commodities can see their production responses affected by both restricted import and export flows through affected chokepoints.

Despite the considerable developments made to the modelling framework since our original analysis was published – extending the time horizons, augmenting socio-economic scenarios, adding climate-forcing pathways, and improving the temporal representation of bilateral chokepoint dependencies – the results remain broadly similar, and our previous conclusions hold. Irrespective of the uncertainties associated with the different scenarios, risk-based management of maritime and coastal chokepoints and the dependencies upon them will likely become even more important for global food security as demands for imported agricultural commodities are expected to increase alongside the multiplying risks to chokepoint disruption. The drought-related disruptions to the Panama Canal in late 2023 are a case in point.

These uncertainties suggest highly precautionary approaches need to be adopted, and that they need to be informed by multiple methodological and stakeholder perspectives. The relatively small impacts on Europe shown in the results from this exercise does not mean that the EU and its supply chains do not need to worry about the potential for more extreme impacts, whether direct or compounding, from climate-induced or climate-exacerbated disruptions to critical food trade chokepoints. This is especially the case since many of the most severe risks will materialise under disequilibrium conditions in the immediate aftermath of disruptions, and such phenomena are not well reflected in the general equilibrium model used in this exercise.

While general equilibrium models of this type can provide important insights into the potential for economic and production shocks to occur, they do not adequately capture the potential for significant disruptions outside of equilibria that may have potentially more important impacts for policy responses in terms of their direct and cascading effects. This modelling framework also assumes rational responses to trade disruptions whereas, as other work in this project has highlighted, unforeseen or irrational human responses are often seen in times of crisis. Responses based on imperfect information, or skewed or limited perceptions of risk and response effectiveness, can magnify and drive risk in international trade systems. Non-rational, strategic and short-sighted responses – including by actors outside Europe – need to be factored into European risk assessments and risk management strategies.

Equilibrium-based models of this type also struggle to capture structural changes in trade and food systems that might arise during or after shocks – such as the emergence of new trading relationships and land use change – and they are limited in their ability to elaborate the impacts of events beyond the implications for GDP or productive output. This exercise has highlighted the difficulty of translating plausible but complex ‘real-world’ shocks into a format that can be used in modelling frameworks. The very different results obtained when recognising a chokepoint disruption as a productivity shock or a tariff shock is a case in point.

Equally, there is scope – and a pressing need – in models representing complex indirect climate risks to better simulate concurrent ‘perfect-storm’ shocks in which several risk factors interact with one another to compound the risks faced. Finally, disaggregating the plausible impacts of chokepoint (or other trade-related) shocks on different stakeholders within the food system (including producers, traders and other companies, and consumers) would help to inform more targeted and risk-calibrated policy responses that are more cognizant of unintended consequences that well-intended policy responses may have.

The limitations of existing models such as those employed in this extended analysis are not surprising in the context of a nascent research field. They do, however, warrant ongoing attention, both so they can begin to inform better calibrated policy responses to novel, complex and dynamic risks, and so that such policy decisions can be taken in full appreciation of their current limitations.