- Armament and disarmament
- Conflict, peace and security
- Peace and development
The SIPRI Arms Transfers Database is a record of how certain military technological capabilities are transferred from state to state. Its fundamental aim is to inform, and ideally stimulate, discussion that might lead to a better understanding of the state of peace and security around the world.
But monitoring these transfers can be challenging. Official transparency about the transfers is patchy at best. Furthermore, deals take many forms. As a result, headline summary figures can throw up some initially surprising results.
The release on 15 March of SIPRI’s new data on international transfers of major arms was no exception. This time, questions revolved around a transfer that—at first glance—appeared to violate a national arms export ban. At their root was an issue of exactly what constitutes an ‘international transfer’ for the purpose of the SIPRI data.
The SIPRI Arms Transfers Database lists officially authorized deliveries of major arms from one state to another. Most of these deliveries entail the physical transfer of finished arms to the military of the recipient state. However, in some cases the ‘supplier’ (sometimes also referred to as the ‘exporter’ in SIPRI publications) instead grants the ‘recipient’ permission to produce the arms under a licensing agreement. Under such an agreement the supplier transfers the design and technology required to produce the weapon to the recipient (although the agreement usually stipulates that some components or other support will be provided by the supplier).
Deliveries of arms produced under licence are still counted as ‘transfers’ in SIPRI’s data. The transfer is dated according to when the finished arms are delivered to the military of the recipient state, regardless of where production (or delivery, for that matter) took place.
The only exception—provided the evidence that the arms in question were delivered is sound enough to include at all—would be if we were confident that production happened without, or in violation of, a valid licensing agreement between the two states (including if the licensing agreement were rescinded). In such a case, we would omit (or remove) the delivery from the database.
Given the low level of official transparency on arms transfers, we have to make judgements like this based on the weight of evidence in our public domain sources. However, to help users, we note significant ambiguity or question marks in the relevant entries in the database.
Treating arms production under licence in this way yields useful and valid information—after all, exactly where the arms are produced is relatively unimportant when the purpose is to track trends in international transfers of military capabilities. However, it can occasionally mean that high-level statistics and summaries raise eyebrows.
That happened this year in the case of a transfer we reported from Ukraine to Russia. In June 2014 Ukraine announced a ban on exports of military goods to Russia. This was in response to Russia’s annexation of Crimea and evidence that Russia was arming separatists in eastern Ukraine—meaning that arms supplied by Ukraine could well end up being used against Ukrainian forces.
After 2014, SIPRI has no reports of any physical transfers of major arms from Ukrainian territory to Russia. Nevertheless, in the fact sheet accompanying the data launch, we listed Russia as the second largest recipient of Ukrainian arms in the five-year period 2016–20.
The materiel in question was 15 An-148 transport aircraft, which were produced in Russia by the Russian company Voronezh Aircraft Manufacturing Company (VASO).
Rights to the An-148 design are held by the Ukrainian company Antonov. In 2013, a year before the export ban, 15 An-148s were ordered from VASO for the Russian military. These were to be produced under a 2005 licensing agreement with Antonov. Several sources confirm that VASO delivered 15 An-148s to the Russian military, the last of them in 2018.
But did VASO produce An-148s after 2014 with Ukraine’s approval and without violating the licensing agreement? Our sources gave a mixed picture. There was the export ban. However, while this clearly meant no new arms exports to Russia would be allowed, it is unclear from publicly available sources whether it covered pre-existing agreements. Also unclear is whether the ban would even apply to the An-148, which was designed as a civilian passenger aircraft. (Indeed, the licensing agreement also covered production for civilian customers in Russia.)
A single 2019 report from a Ukrainian media source quoting Antonov’s parent company, UkrOboronProm, said production of An-148s in Russia had continued under the licensing agreement.
Then there is the question of Ukrainian-produced components. Under the agreement components amounting to around 30 per cent of the Russian-produced An-148s were reportedly to be supplied by Ukraine. Did the ban apply to those components? Were they provided, or did VASO produce some of the An-148s using Russian-sourced substitute components—so violating the licensing agreement?
In 2016 VASO’s parent company, United Aircraft Corporation, announced that it would produce no more An-148s after fulfilling the original order. This suggests that at that point the licensing agreement was still valid and that Russia was not intending to produce the aircraft with substitute components.
So, while it is possible that Russia produced An-148s after 2014 without a valid licensing agreement, the evidence is at best weak and contradictory. Thus, we currently include delivery of 15 aircraft up to 2018 as an international transfer in the database. Following our usual practice, the transfer is dated according to when the last aircraft were delivered to the Russian military: 2018.
However, we do note the uncertainty around this transfer in the database entry, specifying two alternative possibilities: that all of the Ukrainian components had already been delivered before the ban came into force, or that production was completed with substitute components. If we find compelling evidence that Ukraine terminated the agreement or Russia violated its terms, we will remove the relevant An-148s from the database.
When compiling the SIPRI Arms Transfers Database we always strive to be accurate, nuanced and transparent. In many cases, the evidence is unequivocal—or nearly so—of what was transferred, when, and that the transfer was approved by both sides.
But limiting the database to only those cases would risk missing out a significant part of the picture. There are many complex cases, as with the An-148s, where we have to weigh up limited, perhaps contradictory, evidence. Is it at least reasonably probable that a mutually approved transfer took place? What can we say about the who, what, where and when? And what caveats and uncertainties should we include in the database?
We always encourage users to consult the database—and external sources—to understand better the stories behind the headline figures.