Ethiopia’s Negotiations on Access to the Sea: Can International Law Help Find a Solution?

Written by Angelina Shchukina

On 1 January 2024, Ethiopia signed a Memorandum of Understanding (MoU) with the self-proclaimed Republic of Somaliland, reviving tensions in the Horn of Africa region. According to the signatories, whose statements do not precisely coincide, the MoU enables Ethiopia to lease 20 km of coastland for commercial and military purposes for 50 years. Somaliland announced that in return Ethiopia would become the first country to formally recognise it as a sovereign state. Several states and international organisations expressed their support for Somalia’s territorial integrity (e.g. China, the United States, the African Union and the European Union). Somalia, which considers Somaliland a constituent part of its territory, denounced this step as “aggression” and referred the dispute to the attention of the UN Security Council.

A few months before signing the MoU, Ethiopia’s Prime Minister Abiy Ahmed called for negotiations to secure better access to the Red Sea for land-locked Ethiopia. In a televised speech, he suggested several options for discussion with the neighbouring states (Djibouti, Eritrea and Somalia), including leasing land, investing in a port and a land swap. However, the address caused more consternation than cooperation. At the time, Somalia and other neighbours refused to negotiate on the matter. Now, the position of Somalia seems to have evolved. In the recent interview, the Somali President has stated that two countries can negotiate on access to the sea through any part of Somali territory provided that Ethiopia respects Somalia’s territorial integrity. Hopefully, this can turn the tide and present another opportunity for all the Horn states to consider alternative scenarios in the light of their international obligations. While it has been reported that Somalia seeks to take its grievances to the International Court of Justice (ICJ/Court), Ethiopia may also have access to judicial fora.
What Does International Law Have to Offer?

According to journalists, Abiy Ahmed said that “Ethiopia should assert what he called the landlocked nation’s right to access the Red Sea”, making the law of the sea a natural starting point for discussion.

Ethiopia signed the United Nations Convention on the Law of the Sea (UNCLOS/Convention) in 1982 and ratified it 30 years later in 2012. Nevertheless, it seems that Ethiopia has not deposited its instrument of ratification with the UN Secretary-General. The Convention therefore has not entered into force for it (Article 308(2) of UNCLOS).

One may argue that, given the Convention’s wording, the rights of land-locked states enshrined in Part X of UNCLOS may cover states that are not parties to the Convention. The definition of a land-locked state in Article 124 does not refer to a “State Party” and instead uses the word “State”, which does not have a special meaning in UNCLOS: “ ‘land-locked State’ means a State which has no sea-coast.” The rights of land-locked states are formulated in a similar manner, thus their extension to non-parties cannot be definitively excluded on the basis of textual interpretation alone.

On this reading, Ethiopia, as a land-locked state, may benefit from the right of access to and from the sea, freedom of transit through the territory of transit states (Article 125(1)), transit traffic free from customs duties, taxes or other charges except those levied for specific services (Article 127(1)), cooperation on the part of transit states to expeditiously eliminate delays or other difficulties of a technical nature (Article 130). However, for this understanding to be correct, one must demonstrate that UNCLOS parties had the common intention to grant these rights to third states (Free Zones, pp. 147-148; Article 36(1) of the Vienna Convention on the Law of Treaties).

In the case of Ethiopia, this could be especially problematic. During the negotiations that led to the adoption of UNCLOS, a representative of the Group of 77, to which Ethiopia had belonged since its creation (UN doc. A/5587), expressed the common position of its members as follows: “no State can claim that the … rights established by the Convention apply to that State if it is not a party” thereto.

In terms of practical consequences, Ethiopia’s situation may not immediately improve if it deposits its instrument of ratification and becomes a party to the Convention. The right of access and freedom of transit under UNCLOS are not self-executing (see e.g. Sep. Op. of Judge Laing in the M/V “Saiga” (No 2), fn 26). Article 125(2) of the Convention requires that the terms and modalities for the exercise of freedom of transit “shall be agreed” between a land-locked state and a transit state. Without any other treaties in force between Ethiopia and transit states, Ethiopia’s options for protecting its rights through compulsory dispute settlement are scarce. At the very least, Ethiopia’s participation in UNCLOS could arguably afford it an opportunity to challenge a transit state’s failure to negotiate Ethiopia’s access to the sea in good faith (for details see Uprety and Maggio, Article 125, MN 40, or Churchill, Lowe and Sander, p. 841). Such framing of a case may remind the reader of a dispute between Bolivia and Chile before the ICJ. In that case, Bolivia argued that Chile had breached its obligation “to negotiate with Bolivia in order to reach an agreement granting Bolivia a fully sovereign access to the Pacific Ocean”. The ICJ rejected Bolivia’s claims in its 2018 judgment. For greater clarity, it is worth emphasising that the dispute between Bolivia and Chile did not concern the interpretation or application of UNCLOS, even though both were parties to it. However, Bolivia most likely considered invoking its rights under the Convention. Perhaps their scope was not sufficiently broad to meet Bolivia’s needs and such a course of action was dismissed. Similarly, Ethiopia’s publicly stated suggestions appear to go beyond what is provided for in UNCLOS.

As anticipated in UNCLOS, Ethiopia concluded bilateral agreements regarding the exercise of the right of access. The Djibouti Port Utilization Agreement between the Transitional Government of Ethiopia and the Government of the Republic of Djibouti granted Ethiopia the permanent right of access to the sea, the rights to use the port, its installations and equipment and allowed Ethiopian goods to transit through Djibouti’s territory (Articles 1 and 2). Article 3 exempted Ethiopian cargo from “customs or any type of taxation except for the port charges.” Article 4 entitled Ethiopian enterprises to “demand and obtain a plot of land in the duty-free zone” of Djibouti’s port. Provided that the treaty is still in force, Ethiopia can settle a potential dispute with Djibouti through arbitration under Article 9 of the agreement. UNCLOS dispute settlement may also be relevant in this regard. The Virginia Commentary presumes that disputes over the terms and modalities stipulated in agreements envisaged in Article 125(2) of UNCLOS can be settled through the procedures set out in Part XV of the Convention. Consequently, it may be possible to invoke the provisions of a treaty governing transit across the territory of Djibouti using the compulsory dispute settlement procedures of UNCLOS.

As regards bilateral treaties with Ethiopia, the situation is less encouraging. The Transit and Port Services Agreement between the Transitional Government of Ethiopia and the Government of the State of Eritrea designated the ports of Assab and Massawa as transit ports for goods originating from and destined to Ethiopia and laid down the parties’ commitment to “take all measures necessary” for the expeditious movement of traffic and for the avoidance of unnecessary delay (Article 1). Under Article 2 of this agreement, Eritrea undertook to allow the transit of goods to and from Ethiopia “free of taxes and custom duties”. The Eritrea-Ethiopia Claims Commission (EECC) found that this agreement fell within “the category of treaties which become ineffective in time of war” and was either automatically terminated or suspended as a result of the 1998-2000 conflict between the countries. Interestingly, this approach seems to contradict the conclusions of the International Law Commission (ILC) on the effects of armed conflicts on treaties. Pursuant to the ILC’s articles, the existence of an armed conflict does not ipso facto terminate or suspend the operation of a treaty. There may be another difference between the views of the EECC and the ILC. According to the latter, an aggressor state, characterised as such by the Security Council, cannot initiate the termination or suspension of a treaty for its own benefit. This question was not discussed in the dispute before the EECC, as appears from its reasoning. Although the Security Council did not characterise Eritrea as an aggressor, the EECC itself made the relevant finding. It held that Eritrea had violated Article 2(4) of the UN Charter by resorting to armed force. Whatever the inconsistencies may be, the decisions and awards of the EECC are final and binding, and the EECC’s finding regarding the operation of the agreement is res judicata. Most probably, the states have not subsequently resumed or revived the treaty. In any event, its dispute settlement provisions are likely to be unhelpful in resolving the impasse between Ethiopia and Eritrea. As Eritrea is not a party to UNCLOS, its dispute resolution procedures are also of no assistance.

Regardless of whether the states in question are parties to UNCLOS or other relevant treaties in force, Ethiopia can still rely on customary international law as a source of its rights. However, there is little clarity about their existence and scope. Proponents of the customary status of the right of transit rely on the treaty practice of states and repeated statements affirming the existence of this right (see Churchill, Lowe and Sander referring to Uprety and Einhorn, pp. 848-9). On the other hand, it could be argued that the right of transit is not sufficiently of a norm-creating character, and there is not enough evidence of a belief that granting this right is a legal obligation (ibid., referring to North Sea Continental Shelf).

This uncertainty can be resolved by bringing the matter before the ICJ. In theory, land-locked states can try following the path taken by small island states which have convinced the UN General Assembly to request the Court’s opinion on obligations in respect of climate change. An option to choose contentious proceedings is also potentially available. While Ethiopia has not yet made a declaration under Article 36(2) of the ICJ Statute, the other states have. Somalia has accepted the compulsory jurisdiction of the Court for an indefinite period, though subject to several conditions. In 2005, four months before submitting its application in the case against France, Djibouti deposited a declaration recognising the Court’s jurisdiction as compulsory for five years. There is some ambiguity as to whether this declaration has been renewed. The UN Treaty Collection website does not explicitly confirm the renewal. The ICJ website and Djibouti’s statement in 2020 (see UN doc. S/PV.8699 (Resumption 2), p. 4) both seem to suggest that the declaration is in force today. As the case of Iran has recently shown, the existence of a suitable legal dispute may prompt a government to accept the compulsory jurisdiction of the Court, although in such cases the scope of consent may be limited (see the discussion of the Iranian declaration in the blogpost here).

The lack of strong legal protection of the land-locked states’ interests by the law of the sea is somewhat remedied by guarantees stemming from regional economic integration.

Ethiopia and its coastal neighbours are among the parties to the Treaty establishing the Common Market for Eastern and Southern Africa (COMESA Treaty). Its Annex I contains the Protocol on Transit Trade and Transit Facilities (Protocol) which imposes the obligation to grant member states’ transit traffic freedom to cross the territory of the respective states and details its scope (Articles 2 and 3). Article 46 of the COMESA Treaty and Article 2(3) of the Protocol aim to eliminate customs duties. Article 2(4) of the Protocol prohibits discrimination and prescribes that rates and tariffs for the use of facilities by other member states “shall not be less favourable than those accorded to their own traffic.” Notably, the coastal parties to the COMESA Treaty undertook to “co-operate with … land-locked Member States in maritime transport so as to facilitate [their] trade” (Article 88(d) of the COMESA Treaty), and to improve transport infrastructure, while granting “special treatment to land-locked and island Member States” (Article 84(d) of the COMESA Treaty). Article 24(1) of the COMESA Treaty provides for the possibility of referring a member state’s failure to fulfil its obligation under the treaty or any breach of the treaty to the COMESA Court of Justice. Given the customary international law obligation to perform a treaty in good faith (Pulp Mills, para. 145), the absence of genuine cooperation, for example, may result in a violation of the COMESA Treaty and provide a basis for subsequent proceedings before the COMESA judicial body.


This brief and non-exhaustive overview highlights several options that may help overcome deadlocks in potential negotiations provided that a decision of a competent judicial organ is respected by the parties concerned. Although the scope of the rights that Ethiopia can invoke may not be sufficient to achieve all its goals, there are ways of addressing the challenges that Ethiopia faces as a land-locked state.