Despite comprehensive export sanctions from Western allies against Russia, Western components continue to appear in Russian weapons. This column uses transaction-level Russian customs data combined with novel data on EU export bans to separate and quantify three distinct channels through which banned military-relevant goods reached Russia. Trade continued legally for nearly two years, as restrictions were gradually introduced. Furthermore, goods could claim to transit through Russia en route to other destinations or they were rerouted through intermediary countries. It argues that the design and implementation of export bans deserve at least as much attention as their announcement.
Russia’s invasion of Ukraine triggered the most comprehensive export sanctions the EU and its Western allies have ever imposed. Thousands of products – from semiconductors to precision machinery – were banned from export following the invasion of Ukraine. Yet Western components continue to show up in Russian weapons recovered from Ukrainian battlefields. Understanding how export bans operate in practice and why their effects fall short is crucial for designing effective sanction regimes (Hausmann et al. 2023, Felbermayr et al. 2025).
The conventional answer focuses on sanctions-busting. The reality is more complex: trade kept flowing through multiple channels, some perfectly legal under the rules as written. While previous evidence on sanctions evasion exists (Borin et al. 2023, Chupilkin et al. 2025, Scheckenhofer et al. 2025), it was difficult to distinguish between different channels. This column uses transaction-level Russian customs data combined with novel data on EU export bans to separate and quantify three distinct channels through which banned military-relevant goods reached Russia: bans that were too narrowly written, leaving numerous product variants legal; transit loopholes that allowed goods to ‘disappear’ en route; and rerouting through third countries. A more extensive analysis of these channels for military goods appears in Scheckenhofer et al. (2026a).
Bans in name only: The problem of partial coverage
When the EU announced export bans on military-relevant goods, most observers assumed these were comprehensive prohibitions. They were not. The restrictions were introduced gradually and approached full coverage only nearly two years after the invasion.
The core issue lies in how policymakers defined what to ban. Many military-relevant items are dual-use goods with both civilian and military applications, such as semiconductors or industrial machinery. Rather than banning entire product codes, early export bans were text-based, prohibiting exports only of goods meeting specific technical descriptions or destined for particular end uses. A ban might target semiconductors with specifications X, Y, and Z, while leaving other variants unrestricted. The logic seemed sound: minimise disruption to legitimate European exporters while denying Russia critical military capabilities.
In practice, this approach created exploitable gaps. Many product variants remained untargeted for a very long time. Even for products that were restricted, exporters could switch to closely related variants. When variant X was banned but variant Y was not, substitution within product categories is often straightforward, particularly when technical specifications differ only marginally. Text-based restrictions also made enforcement difficult. Verifying detailed technical specifications at the border is nearly impossible, and legal ambiguity made misreporting easier.
We track 42 product categories from the EU’s Common High-Priority (CHP) list, goods repeatedly found in Russian weapons systems, including semiconductors, electronic components, and precision machinery. While all products faced some restrictions shortly after February 2022, most were only partially sanctioned. Full product-code-level bans covering all 42 categories were not in place until January 2024.
These partial bans allowed significant trade to continue legally. Figure 1 shows Russian imports of EU-origin military-relevant goods over time, based on Russian customs data. The light red bars capture this channel: goods that were partially restricted but not fully banned. These flows persisted throughout 2022 and much of 2023, accounting for the largest share of continued EU exports during this period. In 2022, imports of partially sanctioned goods averaged around $36 million monthly, corresponding to roughly one-fifth of pre-war levels. During September through December 2022, these flows spiked to 30-40% of pre-war levels. Importantly, these flows also appear in official EU trade statistics, confirming this is not an artefact of Russian data (see Scheckenhofer et al. 2026a for military goods, and Scheckenhofer et al. 2026b for all goods).
Figure 1 Monthly Russian imports of military goods with EU origin (millions of US dollars)
Note: The figure shows monthly Russian imports of 42 military-relevant product categories from the EU Common High-Priority list, total and decomposed by channel, using Russian customs transaction data and authors’ coding of EU export bans. Dark red: products not yet subject to any ban. Light red: partially banned products (some product variants are banned but others remain legal to export). Dark grey: in-transit flows with EU as both origin and dispatch country. Light grey: fully banned products rerouted via third countries (non-EU dispatch country).
The transit loophole
As direct export bans tightened and covered entire product codes, a second channel remained open. Goods could claim to transit through Russia en route to other destinations. For example, an exporter ships goods from the EU declaring a third country, such as Kazakhstan or Belarus, as the final destination. The goods enter Russia, supposedly to pass through and continue onward. Instead, they remain in Russia.
We identify these flows in Russian customs data by looking for shipments with both EU origin and EU dispatch country for goods fully covered by export bans. Since official EU export statistics confirm that no direct exports to Russia occur once products are fully banned (Scheckenhofer et al. 2026a, 2026b), this pattern indicates goods shipped from the EU and transiting through Russia to third destinations. The dark grey bars in Figure 1 show the magnitude of this channel. In the year following the invasion, in-transit flows averaged around 4% of pre-war import levels. For much of 2022 and 2023, EU sanctions did not prohibit this. Sanctions banned exports to Russia but not transit through Russia. The EU began closing this loophole in mid-2023, prohibiting transit of certain military-relevant goods through Russia. By 2024, stricter enforcement had reduced in-transit flows to roughly 1% of pre-war levels.
Rerouting through third countries
As direct routes closed and transit loopholes tightened, a third channel grew in importance: rerouting through intermediary countries. The light grey bars in Figure 1 show these flows, representing fully sanctioned EU-origin goods that reached Russia via third countries rather than being exported directly from the EU. Intermediary firms operating outside EU jurisdiction can purchase goods in the EU, re-export them to Russia for a profit, and remain beyond the reach of European enforcement. While EU export bans prohibit both direct and indirect exports of restricted goods to Russia, and exporters must conduct due diligence on end users, proving intent to circumvent sanctions is difficult, and intermediary firms face little legal risk.
Rerouting rose sharply from late 2022 onward. During 2023, rerouted flows averaged around $25 million monthly, corresponding to 15% of pre-war levels. Between August 2023 and January 2024, flows peaked at around $36 million monthly (see Figure 1). These flows declined significantly only in 2024, following major EU sanctions tightening that expanded exporter liability, strengthened due-diligence requirements, and increased secondary sanctions against intermediary firms. By late 2024, rerouted flows had fallen to around 8% of pre-war levels (averaging $12 million monthly). This demonstrates that enforcement can work when it targets rerouting channels directly.
Which are the rerouting hubs?
Russian customs data record both the country of origin and the country of dispatch, allowing us to identify the intermediary hubs used for rerouting. The pattern is highly concentrated: more than one-third of EU-origin military goods reaching Russia indirectly pass through Turkey, making it by far the most important hub. China, Hong Kong, and the United Arab Emirates account for most remaining flows.
This concentration has important policy implications. Sanction evasion does not rely on a diffuse network of minor intermediaries but on a handful of key transit routes. Targeting these hubs through diplomatic pressure, information sharing, and secondary sanctions on intermediary firms coordinated across the sanctioning coalition can deliver substantial effects, as the sharp decline in rerouting during 2024 demonstrates. Alternatively, bringing these countries into the sanctions coalition would limit evasion while also expanding the direct economic pressure on Russia (Chowdhry et al. 2022). This could be incentivised through compensatory transfers from coalition members.
How much did the export ban raise trade costs?
The persistence of both direct and indirect exports raises a natural question: how restrictive were the export bans in economic terms? Following the approach in Scheckenhofer et al. (2026a, 2026b), we estimate the implied increase in trade costs associated with EU export bans on military goods. We compare changes in Russian imports of sanctioned EU goods to both non-sanctioned goods and imports from non-sanctioning countries. The export bans raised effective trade costs by around 19% on average over the post-invasion period.
This is a meaningful increase but far from prohibitive. A fully binding export ban would raise trade costs so high that trade collapses entirely (an infinitely high increase). Here, partial coverage and evasion kept trade costs finite, allowing significant flows to persist. Moreover, this 19% estimate is likely an upper bound as our data capture only observable rerouting, not smuggling, misclassification, or falsified documentation. Actual EU exports are likely larger, implying the true trade cost increase is smaller.
Implications for sanctions policy
Three lessons emerge from our analysis.
- First, partial sanctions allow legal trade to continue. Political constraints may explain gradual tightening, but the cost is a delayed and diluted impact.
- Second, blocking direct exports is not enough. Trade simply shifts to indirect routes. Stopping rerouting requires targeting third-country intermediaries through stricter due-diligence requirements and credible penalties.
- Third, targeting intermediary hubs works. The experience of 2024 shows that stricter enforcement and targeted measures against intermediaries can substantially reduce rerouted trade when they focus on key transit countries.
Export bans are powerful, but only when comprehensive, immediate, and enforced. As the EU continues to rely on sanctions as a central tool of foreign policy, the design and implementation of export bans deserve at least as much attention as their announcement.
Source : VOXeu
