Since 2018, ‘de-risking’ – a policy of limiting trade to reduce economic coercion – has become a defining mantra for global economic security. Fearing that economic dependencies can be weaponised and critical supply chains disrupted, many countries have endorsed de-risking, diversification, investment screening, and export-control tools. Geopolitics is already reshaping the geography of global trade flows (Aiyar and Ilyina 2023, Alfaro and Chor 2023, Bosone et al. 2024, Doan et al. 2026).

Mitigating strategic vulnerability is a legitimate goal. However, the current policy debate has potentially tilted too far towards precaution. By focusing almost exclusively on how trade exposes nations to coercion, the de-risking narrative largely overlooks a powerful countervailing force: trade can lower the likelihood of war by making military confrontation economically costly.

This idea, often referred to as the ‘commercial peace’, is a classic insight in economics and political economy (Polachek 1980, Martin et al. 2008). Nevertheless, empiricists have never identified a causal effect of trade on conflict. Trade and conflict are jointly determined leading to reverse causality. For example, trade may foster peaceful relationships, while conflict-prone country pairs may trade less, biasing the effect of trade upwards. Alternatively, communication technologies may improve trade but also fan the flames of nationalism and ‘civilisational conflict’ (Huntington 1996). 

In our new research, we address this empirical challenge by exploiting the way advances in aviation technology reshaped the effective geography of global trade, allowing us to isolate the pacifying effect of trade (Feng et al. 2026).

The geography of the aviation revolution

To identify the causal effect of trade on peace, we exploit improvements in aviation technology. As air transport became cheaper and faster in the late 20th century, it did not reduce transport costs uniformly across country pairs. Instead, it induced a systematic reallocation of global trade. Country pairs where air routes offered a large geographic shortcut over circuitous sea routes experienced a much larger trade boom than pairs where sea routes were already direct.

Table 1 illustrates the extent of geographic variation and its overlap with historical conflict. The sea-to-air ratio measures the detour required by ocean shipping relative to a direct flight. As the table shows, militarised conflicts are heavily concentrated among geographically proximate country pairs. Over 76% of disputes occur between countries located within 5,000 kilometres of each other. Crucially, these same short- and medium-distance pairs also display the largest average sea-to-air distance ratios, ranging from 1.69 to 1.91, as well as the greatest variation in those ratios.

Table 1 Conflict and aviation geography by direct air distance, 1962–2014

Notes: Table reports the distribution of conflicts by direct air distance, for the period 1962–2014. For country pairs falling within each air-distance range, the table reports (1) number of pairs, (2) number of conflicts occurring within these pairs, (3) average sea-transport distance, (4) average sea-to-air distance ratio, and (5) the standard deviation of this ratio. 
Source: Feng et al. (2026), Table 2.

This alignment between high baseline conflict risk and large variation in aviation-induced geographic shortcuts provides a suitable setting for identifying the causal effect of international trade on conflict. Because these geographic structures are largely fixed by physical geography, rather than chosen by governments in response to political alliances or security concerns, they can be used to predict exogenous trade reallocations and better isolate the pacifying effect of economic integration.

Defusing crises: Less conflict, softer rivalry

Our empirical analysis confirms a robust and economically meaningful ‘peace dividend’ of international trade. Our preferred instrumental-variable estimates show that a doubling of bilateral trade reduces the likelihood of militarised conflict by roughly 30%.

Trade also reduces the severity of conflict. Our estimates show that greater trade integration lowers the probability that a dispute escalates to the highest levels of hostility, which involve use of military force and high-casualty wars.

Further, this pacifying effect takes hold before conflict becomes militarised. Using data on strategic rivalry, we find that trade reduces the probability that two states view each other as strategic rivals. The effect is particularly pronounced for positional rivalries – competition over status, influence, and regional standing. Trade therefore appears to soften antagonistic relationships before they escalate into open crises.

The geography of trade’s peace dividend

To assess the importance of trade, we use our baseline instrumental-variable estimates to calculate, for each country, the predicted reduction in conflict probability generated by trade integration.

We find that the largest estimated peace dividends are concentrated in East and Southeast Asia. Economies that experienced rapid industrialisation and deep integration into global value chains, such as China and South Korea, fall into the top deciles of conflict reduction. The same is true for countries such as Myanmar, the Philippines, and Thailand.

Figure 1 plots our model’s predicted peace gains (i.e. reductions in the likelihood of war) against the actual historical declines in militarised disputes that countries experienced. The correlation is positive and statistically significant. Countries that our model identifies as the biggest beneficiaries of the aviation-driven trade boom are also those that experienced the largest actual drops in conflict occurrence. Many economies in East and Southeast Asia appear to be significant beneficiaries of increased trade.

Figure 1 Actual and fitted reductions in the probability of conflict by country

Notes: The horizontal axis reports the fitted reduction in conflict probability in 2012 from the baseline model relative to a counterfactual in which bilateral trade is reset to 1970 levels. The vertical axis reports the actual country-level reduction. Each dot represents a country.
Source: Feng et al. (2026), Figure 6b.

Unpacking the mechanisms: Where does trade matter most?

Beneath these aggregate peace dividends lies a distinct geographic logic. We find that the pacifying effect of trade is concentrated precisely where (and when) the aviation revolution produced the deepest structural changes in trade.

First, trade significantly restrains conflict among country pairs with high sea-to-air distance ratios. For country pairs where air transport did not provide a substantial geographic shortcut relative to maritime shipping, the pacifying effect of trade is small and statistically insignificant.

Second, the effect is driven primarily by non-island nations. For country pairs involving at least one island country, which naturally remain heavily reliant on maritime routes, the impact of trade on conflict is weak and imprecisely estimated. By contrast, among continental pairs, the aviation-driven trade boom softened political tensions.

Finally, the timing of the peace dividend aligns with the acceleration of modern globalisation. When we split the sample chronologically, we find that before 1980, the estimated effect of trade on conflict was negligible. However, after 1980, when long-range aircraft and modern logistics became increasingly important to global production networks, the conflict-reducing impact of trade becomes large and precisely estimated.

Do not evaluate benefits and costs of production-network fragmentation with one security margin

By focusing almost exclusively on how interdependence exposes nations to shocks or coercion, policymakers may underweight the benefits of international trade. Trade can raise the opportunity cost of conflict and strengthen incentives for restraint (Martin et al. 2008, Mayer et al. 2025). An economy less engaged with the global economy may be safer from coercion, but the evidence also suggests that such reductions in integration can strengthen international rivalries and heighten the chances of outright war.

The relevant policy question for the 21st century is therefore not a binary choice between ‘openness’ and ‘security’. Instead, nations should attempt to measure the benefits of de-risking against both economic and security costs.

Ultimately, the relevant policy distinction in the years to come is not a binary choice between openness and security. It is the balance between dangerous dependence and stabilising integration.

References

Aiyar, S, and A Ilyina (2023), “Geo-economic fragmentation and the world economy”, VoxEU.org, 27 March.

Alfaro, L, and D Chor (2023), “A perspective on the great reallocation of global supply chains”, VoxEU.org, 28 September.

Bosone, C, E Dautović, M Fidora, and G Stamato (2024), “How geopolitics is changing trade”, VoxEU.org, 14 May.

Doan, T T H, A Ito, C Luo, and H Zhang (2026), “Geopolitical risk and supply chain diversification”, VoxEU.org, 23 February.

Feng, L, Q Huang, Z Li, and C M Meissner (2026), “The ‘peace dividend’ of international trade: A new empirical approach”, NBER Working Paper No. 35078.

G7 (2023), “G7 Hiroshima leaders’ communiqué”, 20 May.

Huntington, S P (1996), The clash of civilizations and the remaking of world order, New York: Simon and Schuster.

Martin, P, T Mayer, and M Thoenig (2008), “Make trade not war?”, Review of Economic Studies 75(3): 865–900.

Mayer, T, I Méjean, and M Thoenig (2025), “The fragmentation paradox: De-risking trade and global safety”, CEPR Discussion Paper 20564.

Polachek, S W (1980), “Conflict and trade”, Journal of Conflict Resolution 24(1): 55–78.