Arms Trade Treaty examinedExplosive violence by the Saudi-led coalitionManufactured explosive weaponsAOAV: all our reportsExplosive violence in YemenExplosive violence research

Case studies: Saudi Arabia in Yemen – Assessing the effectiveness of the Arms Trade Treaty, Part 8

In March 2015, in response to the uprisings in Yemen and the Houthi rebel takeover of Sana’a, the Saudi-led coalition[i] began military operations against the Houthi rebels. These operations, including countless airstrikes and a blockade of strategic ports, have culminated in a humanitarian crisis across Yemen. While there has been widespread condemnation of the actions of the coalition forces, many states, including those party to the ATT, have continued to supply arms, including weaponry used to carry out attacks on civilians and civilian infrastructure in violation of international humanitarian law in Yemen.[ii]

This case study will examine Saudi Arabia, as the key perpetrator of explosive weapon use in Yemen, through the lens of explosive weapon harm and ATT effectiveness.

Read the full report here

Civilian deaths and injuries from Saudi Arabia’s explosive weapon use

The below table indicates the harm seen each year from the use of manufactured explosive weapons by Saudi Arabia and Saudi-led coalition forces in Yemen. All 10,408 civilian casualties recorded occurred since 2015 and of these 5,039 were killed and 5,369 were injured.[iii]

Among these casualties, there were at least 367 women, though only 18% of incidents mention one or more female casualties. At least 833 children were also among these casualties – 25% of incidents reported one or more child casualties. The UN Secretary-General’s Report on ‘Children and armed conflict in Yemen’, recorded 341 child casualties as a result of Coalition airstrikes between 2019 and 2020,[iv] while there were 3,497 between 1 April 2013 and 31 December 2018.[v]

Air strikes have been, by far, the most frequently used weapon type and the cause of the vast majority of casualties. In total, when all types of air-launched weapons are considered, they were responsible for at least 95% (9,855) of the total civilian casualties.

The use of explosive weapons in populated areas is responsible for 8,912 of the civilian casualties. The locations which saw the most civilian casualties from Saudi Arabia’s use of explosive weapons were urban residential areas (with 2,540 civilian casualties in 215 incidents), markets (1,204 civilian casualties in 25 incidents), and public buildings (with 890 civilian casualties in 12 incidents).

Much of the below sections relate to further causes of death, injury and illness as a result of the consequences of explosive weapon use, such as the destruction of healthcare facilities and other essential services.

Impact on infrastructure and communities

Damage and destruction

Saudi-led coalition forces in Yemen, 14% were reported to have also resulted in damage to infrastructure. This mostly appears to be damage to residential areas but has also included damage to humanitarian infrastructure, schools and hospitals.

While the type of infrastructure destroyed has specific impacts, each destruction is a loss for families, communities, and society. When an electrical substation was hit and destroyed in Yemen in 2015, the connected water supply installations were immediately shut down, cutting off the water supply for more than 400,000 people and the local hospital. Such impacts can be long-lasting depending on the scale of the damage and intensity of the conflict.[vi]

The damage to Hudaydah port, Yemen’s largest, from the airstrikes in 2015 was linked to a rise in food insecurity and a decrease in fuel imports – essential to the functioning of Yemen’s water systems, health facilities and transportation.[vii] Hudaydah’s port is responsible for 70-80% of Yemen’s commercial imports.[viii]

Road damage has also been significant across Yemen, with a damage assessment by the Road Maintenance Fund, a unit of Yemen’s Ministry of Public Works and Highways, finding that 1,241km of road is thought to be heavily damaged in conflict-impacted governorates and 62 bridges damaged or destroyed by airstrikes.[ix]

Damage to infrastructure in Yemen has also been recorded by the Civilian Impact Monitoring Project (CIMP), a service under the UN Protection Cluster in Yemen. The CIMP recorded 91 airstrike incidents affecting main roads from January 1st 2018 to March 25th 2020 – this damage is estimated to ‘have hindered access or restricted freedom of movement for over 1 million households’.[x] They also recorded 13 airstrike incidents affecting bridges, impacting over 100,000 households. While CIMP’s data covers a smaller time frame than the Road Maintenance Fund, it gives a better scale of the level of impact.

The damage can make essential services harder to access and impede the movement of goods and humanitarian aid. This is hugely important to a country with more than 3 million children facing acute food insecurity.[xi]


According to UNHCR, Yemen has the fourth-largest IDP population due to conflict in the world, with 4.3 million IDPs.[xii] Of these, 40% live in informal displacement sites with inadequate access to basic services. Others often rely on informal rental agreements and face the pervasive threat of eviction.[xiii] The UNHCR estimates that approximately 66% of Yemen’s IDPs live in dangerous locations such as places with food and water insecurity or without healthcare and sanitation services.[xiv]The areas seeing the highest levels of displacement reflect the areas seeing the highest levels of conflict. This is illustrated below, utilising ACLED incident data and UNHCR data for Yemen.[xv]

While it is difficult to determine the levels of displacement due to Saudi Arabia’s use of explosive weapons, specifically, the airstrikes and the destruction drive displacement and prevent return. In Sanaa city, for example, the IOM estimated that over 270,000 civilians were displaced between March 2015 and October 2017, primarily as a result of airstrikes.[xvi]

Impact on health and well-being

There are quite a few organisations reporting data on healthcare damage in Yemen, including Mwatana for Human Rights, Safeguarding Health in Conflict Coalition (SHCC), the UN Secretary-General reports, and the WHO. Some of the most relevant data is included below.

The organisation, Mwatana for Human Rights, documented 35 coalition airstrikes on 32 health facilities between 2015 and 2018 – killing 31 and injuring 56, including healthcare workers. Most of these attacks caused significant damage to the facilities they hit.[xvii]

In 2021, the SHCC recorded six incidents of Saudi-led coalition airstrikes which impacted healthcare directly.[xviii] One incident resulted in the death of a health worker as well as two injuries – this incident also resulted in damage to the health facilities. A further three of the incidents in 2021 also resulted in damage to healthcare facilities; another resulted in destruction to facilities and one resulted in damage to health transportation.

The data from 2020, also indicates six attacks from state use of explosive weapons, resulting in one health worker killed, two health facilities destroyed and three damaged.[xix] Though this data doesn’t identify the specific state perpetrator, it is likely these were caused by the Saudi-led coalition forces.The WHO recorded 175 attacks on healthcare between March 2015 and 2021, and 67 were the result of ‘heavy weapons’.[xx] The vast majority of these attacks occurred in 2015, the worst period for Saudi-coalition airstrikes.

The impact on healthcare

Yemen’s health systems were already limited before the conflict, with more than half the population lacking access to healthcare services.[xxi] With a healthcare system already struggling, damage to infrastructure and loss of personnel have an even greater impact. A loss of a hospital, for example, can leave thousands without care. The ICRC report that the destruction of Hayden Hospital in Saada by airstrikes in 2015 left 200,000 people without medical care.[xxii]

In OCHA’s ‘Humanitarian Needs Overview: Yemen’, published in April 2022, it was reported that only half of Yemen’s health facilities were operational, with 11% fully or partially damaged.[xxiii] And, even those operational, struggle with ‘staff shortages, inadequate supplies and equipment, inability to meet operational costs and power outages caused by lack of fuel’.[xxiv] This has meant that more than 42% of Yemen’s population are estimated to live more than one hour from the nearest functioning hospital and districts which see the highest levels of conflict face the highest access deprivation.[xxv]

2.2 million children are acutely malnourished, with 190,000 and 316,000 of those severely malnourished at risk of death if timely care is not provided. Immunisation has also declined, leading to outbreaks of diseases such as polio, cholera and measles.[xxvi] Maternal and child mortality has also increased.[xxvii]

After over one thousand days of war in Yemen, the first large protest takes place in London of a few hundred people outside Downing Street, 7 March 2018. Source: (Alisdare Hickson)

Impact on education

Education facilities damaged and destroyed

The extent of the damage to schools in Yemen is unclear, with little data covering specifically damage from Saudi coalition airstrikes. Data from Insecurity Insight and the GCPEA have some data available which indicates such harm from Saudi-led coalition airstrikes. In 2021, for example, five schools were damaged and four were destroyed by attacks using explosive weapons. While the perpetrators of these attacks aren’t listed in the data, at least one is likely to have been caused by the Saudi-led coalition, the air-launched explosive attack listed below on May 27th 2021, which damaged a school.

In 2020, 12 schools were recorded as damaged and destroyed according to Insecurity Insight and nine in 2019, though the data doesn’t indicate the nature of the attacks. The 2018 data that relates to Saudi-led coalition airstrikes can be seen below.

This 2018 data includes one of the most well-known airstrikes by the Saudi-led coalition; when an airstrike hit a school bus in Saada, killing at least 40 children.

The Education Cluster reported that between March 2015 and December 2018, 1,800 schools were either damaged or destroyed due to conflict – a larger estimate than that suggested by the above data.[xxviii]

The report by the UN Secretary-General on ‘Children and armed conflict in Yemen’ in 2019, identified 244 attacks on schools, of which 140 were attributed to the Coalition.[xxix] The report includes examples of airstrikes repeatedly targeting schools and causing damage and destruction. The 2021 publication of this report also identifies a further four attacks on schools by the Coalition forces.[xxx]

It is estimated that 2 million children in Yemen are no longer in school and another 3.7million are at risk of losing access to education.[xxxi] While it is not just the actions of Saudi Arabia and the Saudi-led coalition that have resulted in this harm, the airstrikes on schools further devastate the state of education in Yemen.

Findings related to the ATT

Civilians, civilian infrastructure and populated areas have frequently been a target of Saudi-led coalition airstrikes. The violence discussed above indicates arms sales to Saudi Arabia are likely to breach articles of the ATT and would therefore violate the commitments of states parties to the ATT. Yet countries, including those party to the ATT, have continued to supply arms.

These arms undermine peace and security and have a considerable chance of being used to violate international humanitarian law and international human rights law, namely attacks on civilians and civilian infrastructure. The UN Under-Secretary-General for Humanitarian Affairs highlighted that the Saudi-led coalition airstrikes contravened international humanitarian law.[xxxii]

Arms exports to Saudi Arabia

Arms to Saudi Arabia have been highly criticised since 2015. This is because many of those supplying arms to Saudi Arabia are party to the ATT and seemingly violating their commitments under the Treaty. As can be seen from the table below, and from Appendix 5, of the 24 states recorded in the SIPRI data as having transferred arms to Saudi Arabia since 2012, 20 are party to the ATT, three are signatories and one is not party to the ATT.[xxxiii] Only two of these states have not provided arms post-2015: Finland and Netherlands, though they supplied arms in 2015 itself.

The biggest suppliers of arms, including explosive weapons and systems that launch explosives, in this period have been France, the UK and the US: the US has been by far the biggest supplier. Of these, France and the UK, have ratified the ATT, while the US has not ratified the ATT, despite signing the ATT in 2013. Typically, signatory states are bound to act in a way that is not contrary to the object and purpose of the treaty, despite not ratifying the treaty. However, Trump in April 2019 announced he would be “unsigning” the Treaty. While it is not possible to “unsign” a treaty, the Trump administration confirmed to the UN that the US did not intend to ratify the treaty or abide by the object and purpose of the treaty.[xxxiv] Since entering office President Biden has not recommitted the US to be bound by its obligations as a signatory.

This neighbourhood, where more than 100 buildings have been damaged, has brought attention to the plight of Yemeni blacks, with neighbouring communities coming to witness the damage, Sana’a, Oct. 9, 2015. (Almigdad Mojalli/VOA). Source:

Arms provided by the US to Saudi Arabia in recent years have included thousands of Paveway guided-bombs, missile systems, missiles, combat helicopters, and combat aircrafts, among others. Many of these arms have been found to have been used against civilians in Yemen by the Saudi-led coalition forces.[xxxv]

The UK and France have supplied missiles of various kinds and weapons systems. Others such as Belgium, Bulgaria, Canada, China, Germany, Serbia, Slovakia, South Korea, and Spain have also potentially violated their commitments as Parties to the ATT, though on a smaller scale, with their post-2015 arms transfers to Saudi Arabia. While not recorded in the SIPRI data, the UNROCA database (Appendix 6) shows significant transfers of explosive weapons and systems which launch explosives from Bulgaria between 2015 and 2020, including thousands of mortars and grenade launchers, for example.[xxxvi] Bulgaria ratified the ATT in 2014, which may make such transfers a violation of their ATT commitments, depending on their end user.

The UK

The UK has in some instances refused the export licences for military goods to Saudi Arabia, as recorded by CAAT. Many of these covered explosive weapons.

All 15 refusals were for unlimited-value licences – no limited-value licences have been refused. It is unclear why these were refused and it may not be related to Saudi Arabia’s use of explosive weapons in Yemen, particularly considering that 127 unlimited licences have been approved since Saudi Arabia began its bombing campaign on Yemen – many include explosive weapons, components and services for combat aircrafts and related equipment.[xxxvii] It is worth noting that these transfers have been approved even despite the UK itself listing Saudi Arabia as a country of concern regarding human rights.[xxxviii] Though the UK does provide arms to many countries on their own list.

Though about £7 billion worth of arms sales to Saudi Arabia have been recorded by the Department of International Trade, CAAT has found evidence of secret UK arms sales through the use of these unlimited Open Licences.[xxxix] Such licences, as the name suggests, allow an unlimited number of the items specified to be transferred over a pre-set period, or indefinitely. With these licences it is difficult to know the number of items exported. CAAT’s research found that BAE systems had sold £17.5 billion worth of equipment and services using open licences since 2015.[xl] CAAT’s report was released in July 2021, so it is likely the value of these exports and the numbers of exported weapons are even higher than these figures.

Despite a High Court imposing a moratorium on the licensing of arms transfers by the UK to Saudi Arabia, pending a review, the UK continued to grant new export licences. The UK then decided that, despite international humanitarian law violations, arms exports could lawfully continue as there was said to be ‘no clear risk’ of future serious breaches. Since this decision, the UK has licensed more than £1.4 billion in arms exports.[xli] (The UK’s arms transfers are also discussed in a separate section further in this report.)

Other ATT state party exporters

As mentioned, many ATT States Parties have continued to supply arms exports to Saudi Arabia despite their ATT commitments. France has continued to supply arms to Saudi Arabia and is their third largest supplier of arms, with 547 licences approved between 2017 and 2020, worth more than €42 billion and exported arms worth €4.4 billion.[xlii]

However, some of the ATT states mentioned have taken action in response to Saudi Arabia’s use of arms. The ATT Monitor in the 2021 report highlighted that “the number of States Parties that exported weapons to Saudi Arabia fell in recent years from 13 to 7, with Belgium, Germany, the Netherlands, Norway, Switzerland and Italy implementing bans or other restrictive policies on exporting arms to Saudi Arabia.[xliii]  Denmark, Finland and Greece have also passed legislation to impose restrictions on military exports to members of the Saudi-led coalition.[xliv] Though some eventually allowed new exports despite bans, in response to industry pressure.[xlv]

An article by the ATT Expert Group highlights the judicial cases against various ATT states parties who supplied arms to Saudi Arabia and other coalition partners. While the report displayed the use of legal challenges to arms transfers in some ATT states, something that rarely occurred prior to the ATT entering into force, the report also demonstrated the lack of domestic enforcement power.

Other forms of effectiveness

Another way to perceive if the ATT has been effective in its mission of reducing human suffering may be to assess the level of harm caused by those using the arms.[xlvi] From the data it would seem that the number of Saudi Arabia and Saudi-led coalition attacks on populated areas has decreased since between 2015 and 2021, as have the casualties. The average number of civilian casualties per reported incident has also decreased quite significantly, from 31 in 2015, to 6 in 2021.

However, this could be due to a reduction in reporting, a decrease in the intensity of the conflict more generally, or, indeed, because many have fled areas of conflict. Further, despite these decreases, the proportion of attacks occurring in populated areas hasn’t decreased.

Nevertheless, as many of the largest suppliers of arms have continued to supply arms since 2015, it seems unlikely that Saudi Arabia, which has continued to hit populated areas in recent years, has really felt much of a sting from exporters’ ATT commitments.

Navigate the report:

Assessing the effectiveness of the Arms Trade Treaty – Executive Summary

Part 1: Nation-by-nation review analysis

Part 2: Who is causing the most harm?

Part 3: Who is providing arms?

Part 4: Thematic examination

Part 5: Conclusion

Part 6: Recommendations 

Part 7: Case studies – Myanmar’s military

Part 8: Case studies – Saudi Arabia in Yemen

Part 9: Case studies – Non-state armed groups in the Philippines

Part 10: Case studies – the Taliban

Part 11: Case studies – China before and after ATT accession

Part 12: Case studies – the United Kingdom, from key ATT architect to key violator?

[i] Saudi Arabia, UAE, Sudan, Bahrain, Kuwait, Qatar, Egypt, Jordan, and Morocco, with Senegal sending troops shortly after.

[ii] It is worth noting that the Houthi rebels have also frequently violated international humanitarian law, though this does not justify the actions of the Saudi-led coalition or those who have supplied arms.

[iii] It would be expected for there to be a higher number of injuries than fatalities and this may be linked to reporting fatigue and what is considered newsworthy.

[iv] United Nations, ‘Children and armed conflict in Yemen: Report of the Secretary-General’, 03 June 2019, S/2019/453.

[v] United Nations, ‘Children and armed conflict in Yemen: Report of the Secretary-General’, 27 August 2021, S/2021/761.

[vi] ICRC, Explosive Weapons With Wide Area Effects: A Deadly Choice in Populated Areas, ICRC, Geneva,

January 2022.

[vii] A. Bottomley and L. Salavert, ‘Death Sentence to Civilians: The Long-Term Impact of Explosive Weapons in Populated Areas in Yemen’, Humanity and Inclusion, May 2020.

[viii] OCHA, 2018, Yemen Humanitarian Needs Overview, 8.

[ix] Humanity and Inclusion, ‘Death Sentence to Civilians: The Long-Term Impact of Explosive Weapons in Populated Areas in Yemen’, May 2020, (accessed 07 Jan 2023)

[x] CIMP, Data from 1 January 2018 to 25 March 2020, as recorded in: Humanity and Inclusion, ‘Death Sentence to Civilians: The Long-Term Impact of Explosive Weapons in Populated Areas in Yemen’, May 2020, (accessed 07 Jan 2023)

[xi] OCHA, 2022, ‘Humanitarian Needs Overview: Yemen’.

[xii] Data from UNHCR, ‘Yemen’,,internally%20displaced%20persons%20(IDPs). (accessed 07 Jan 2023)

[xiii] UNHCR, ‘Needs mount as conflict in Yemen rages on’, 01 April 2022, (accessed 07 Jan 2023)

[xiv] USA for UNHCR, ‘Yemen Crisis Explained’, (accessed 07 Jan 2023)

[xv] ACLED, ‘Yemen’s Fractured South: Shabwah and Hadramawt’, 09 May 2019, (accessed 07 Jan 2023), and UNHCR data (accessed 07 Jan 2023)

[xvi] IOM, “UN Migration Agency: Military Attack on Yemen’s Al Hudaydah Port, City Will Endanger Lives, Humanitarian Response”, 5 December 2017: (accessed 07 Jan 2023)

[xvii] Physicians for Human Rights, ‘“I ripped the IV out of my arm and started running”: Attacks on Health Care in Yemen’, 18 March 2020,  (accessed 07 Jan 2023)

[xviii] Data from SHCC’s page on the Humanitarian Data Exchange: (accessed 07 Jan 2023)

[xix] Data from SHCC’s page on the Humanitarian Data Exchange: (accessed 07 Jan 2023)

[xx] Data collected from WHO, ‘SURVEILLANCE SYSTEM FOR ATTACKS ON HEALTH CARE (SSA)’,  (accessed 07 Jan 2023)

[xxi] World Bank, 2017, “Yemen: Immediate Priorities for Post-Conflict Recovery of the Health Sector,”4.  (accessed 07 Jan 2023)

[xxii] ICRC, ‘Explosive weapons: Civilians in populated areas must be protected’, 26 January 2022,  (accessed 07 Jan 2023)

[xxiii] OCHA, 2022, ‘Humanitarian Needs Overview: Yemen’.

[xxiv] OCHA, 2022, ‘Humanitarian Needs Overview: Yemen’, pp.63.

[xxv] Garber K, Fox C, Abdalla M, Tatem A, Qirbi N, Lloyd-Braff L, Al-Shabi K, Ongwae K, Dyson M, Hassen K. Estimating access to health care in Yemen, a complex humanitarian emergency setting: a descriptive applied geospatial analysis. Lancet Glob Health. 2020, 8(11).

[xxvi] OCHA, 2022, ‘Humanitarian Needs Overview: Yemen’.

[xxvii] World Bank, ‘Health Sector in Yemen – Policy Note’, September, 2021. Available at: (accessed 07 Jan 2023)

[xxviii] Global Coalition to Protect Education from Attack, 2020, ‘EDUCATION UNDER ATTACK 2020’,  (accessed 07 Jan 2023) 

[xxix] United Nations, ‘Children and armed conflict in Yemen: Report of the Secretary-General’, 03 June 2019, S/2019/453.

[xxx] United Nations, ‘Children and armed conflict in Yemen: Report of the Secretary-General’, 27 August 2021, S/2021/761.

[xxxi] UNICEF, ‘As school year starts in Yemen, 2 million children are out of school and another 3.7 million are at risk of dropping out’, 25 September 2019, (accessed 07 Jan 2023)

[xxxii] Reuters, ‘U.N. says air strikes on Yemen port could worsen crisis’, 19 August 2015, (accessed 07 Jan 2023)

[xxxiii] Stockholm International Peace Research Institute (SIPRI) ‘Trade Registers’ data available here: (accessed 07 Jan 2023)

[xxxiv] Stohl, R. ‘Why is the Biden Administration Still Silent on Arms Trade Treaty?’, Stimson, 27 April 2022, (accessed 07 Jan 2023)

[xxxv] Amnesty International, ‘Yemen: US-made bomb used in deadly air strike on civilians’, 26 September 2019, (accessed 07 Jan 2023); Amnesty International, ‘Yemen: US-made weapon used in air strike that killed scores in escalation of Saudi-led coalition attacks ‘, 26 January 2022, (accessed 07 Jan 2023); and Elbagir, N., S. Abdelaziz, R. Browne, B. Arvanitidis, and L. Smith-Spark, ‘Bomb that killed 40 children in Yemen was supplied by the US’, 17 August 2018, (accessed 07 Jan 2023)

[xxxvi] Data collected from the United Nations Register of Conventional Arms, available at: (accessed 07 Jan 2023)

[xxxvii] Data collected from Campaign Against the Arms Trade, ‘UK Export Licence Data’. All data: Specific data link: (accessed 07 Jan 2023)

[xxxviii] Foreign, Commonwealth & Development Office, ‘Human rights priority countries: ministerial statement for January to June 2021’, 23 November 2021, (accessed 07 Jan 2023)

[xxxix] CAAT, 2021. ‘OPEN? The UK’S secret arms sales’, (accessed 07 Jan 2023)

[xl] CAAT, 2021. ‘OPEN? The UK’S secret arms sales’, (accessed 07 Jan 2023)

[xli] CAAT, ‘£1.4 billion worth of arms exports licensed to Saudi Arabia in aftermath of decision to resume arms sales’, 09 February 2021, (accessed 07 Jan 2023)

[xlii] Gallagher, K. and E. Standfield, ‘COMPARING NATIONAL POSITIONS ON

MILITARY EXPORTS TO SAUDI ARABIA’, Project Ploughshares, October 2021,  (accessed 07 Jan 2023)

[xliii] ATT Monitor, 2021. ‘ATT Monitor Report 2021, (accessed 07 Jan 2023)

[xliv] Uddin, R. ‘Saudi Arabia arms sales: Which countries are still exporting?’, Middle East Eye,  19 February 2021, (accessed 07 Jan 2023)

[xlv] Bogers, M., Beeres, R., and Smetsers, K. 2022. ‘Effectiveness of Arms Control: The Case of Saudi Arabia’, in NL ARMS Netherlands Annual Review of Military Studies 2021: Compliance and Integrity in International Military Trade, eds. R Beeres, R Bertrand, J Klomp, J Timmermans, J Voetelink. Asser Press/Springer, pp 243-262.[xlvi] For more of an exploration into this see: Bogers, M., Beeres, R., and Smetsers, K. 2022. ‘Effectiveness of Arms Control: The Case of Saudi Arabia’, in NL ARMS Netherlands Annual Review of Military Studies 2021: Compliance and Integrity in International Military Trade, eds. R Beeres, R Bertrand, J Klomp, J Timmermans, J Voetelink. Asser Press/Springer, pp 243-262.