Chapter 2. Aligning Governing Law and Forum Selection in EPC Disputes: A Practitioner’s Roadmap to Cost-Efficient Infrastructure Arbitration

The choice of applicable law and forum has become a cornerstone of any major cross-border infrastructure project today. Without clear agreement on these parameters, construction projects face delays, increased costs, and legal uncertainty. Procedural conflicts multiply as the global construction industry relies on distributed supply chains and investors increasingly invoke multilateral investment protection mechanisms to safeguard their capital (Alferova, 2016). Modern private international law doctrine emphasizes that only a proper choice of law and forum can minimize transaction costs. This study hypothesized that aligning the key choice-of-law connecting factors with the criteria for jurisdictional admissibility maximizes procedural economy, regardless of whether a dispute is heard by the International Centre for Settlement of Investment Disputes (ICSID), an ad hoc UNCITRAL tribunal, or a national court.

The study’s objective was to develop scientifically grounded guidance on harmonizing conflict-of-law and procedural norms in the Russian practice of Engineering, Procurement and Construction (EPC) contracts – including those based on the standard forms developed by the International Federation of Consulting Engineers (FIDIC). To that end, the research analyzes legal doctrine, arbitral precedents, and FIDIC model contract conditions, identifying problem points at the stages of contract formation, performance, and dispute resolution. The object of the research encompasses the network of social relations arising from cross-border construction projects; the subject is the body of private international law norms and international procedural norms that govern the determination of applicable law and competent forum.

The methodology integrates a formal legal analysis of Article 42(1) of the 1965 Washington Convention (ICSID Convention) on the Settlement of Investment Disputes between States and Nationals of Other States, interpretation of Article 35 (formerly 33) of the UNCITRAL Arbitration Rules, a comparative assessment of Anglo-American and continental risk allocation approaches, as well as a review of arbitral precedents in the field of international construction contract law. This combination of methods provided a multifaceted perspective on the problem.

The relevance of this topic is underscored by recent scholarship. For example, Born and Kalelioglu (2021) demonstrate that even amid universal acceptance of party autonomy, states delineate the bounds of public policy differently. Huang (2023) shows how China’s new Civil Procedure Law (2024) dramatically changes rules on international jurisdiction, forcing contractors to rethink their forum selection strategies. Meanwhile, Yakovleva (2022) finds that even a formally impeccable arbitration clause may be ineffective if the dispute falls outside ICSID Convention criteria – a particularly acute issue for construction concessions.

The core problem lies in the multi-tiered nature of major international construction and investment agreements: a single dispute often intertwines contractual claims with treaty-based investment claims. Properly disentangling these layers allows a tribunal to apply the correct norms and avoid duplicative proceedings. However, Russian practice – despite formally giving priority to treaties over domestic law (Article 15(4) of the Russian Constitution) – still relies on a limited set of “special conditions” and does not always integrate FIDIC standards seamlessly into international projects. This creates a gap between contractual practice and public-law guarantees of investment security. As academician A. G. Lisitsyn-Svetlanov observes, the task for developing the legal system lies “not in creating special conditions for foreign investors, including special dispute resolution arrangements, but in ensuring a general investment regime based on fair competition” (Lisitsyn-Svetlanov, 2021). In other words, overly preferential regimes for foreign contractors, absent a robust general framework for investor rights, risk undermining sovereignty and legal coherence.

International construction projects – for instance, those carried out under FIDIC contract conditions – frequently give rise to complex disputes involving foreign contractors and states. Such disputes invariably raise questions of both the applicable law and the jurisdiction of international arbitral forums. Over the past decades, a substantial body of case law has developed under the three main mechanisms for international dispute resolution: the ICSID system, ad hoc arbitrations under the UNCITRAL Rules, and arbitrations administered by the Permanent Court of Arbitration (PCA). A notable commonality across these fora is how arbitral tribunals determine and apply (1) the applicable law, i.e. the set of legal rules governing the merits of the dispute; and (2) jurisdiction, i.e. whether and to what extent the tribunal has authority to hear the dispute. An analysis of ICSID, UNCITRAL, and PCA decisions reveals both similarities and differences in their approaches to conflict-of-law issues and to the interplay between contractual obligations and investment treaty norms. In what follows, key precedents – leading cases concerning large international construction contracts and infrastructure projects – are examined to illuminate prevailing trends and distinct features.

Applicable Law under ICSID Practice. In ICSID arbitration, choice-of-law is governed by Article 42(1) of the Washington Convention 1965. If the parties have not explicitly agreed on the applicable law, an ICSID tribunal “shall apply the law of the Contracting State party to the dispute (including its conflict-of-laws rules) and such rules of international law as may be applicable.” In practical terms, for an investor–state dispute arising from a construction contract (for example, a contract with a state agency), an ICSID tribunal will primarily apply the law chosen by the parties in the contract (e.g. the law of a specified country). In the absence of such choice, the law of the respondent state applies, supplemented by applicable international law.

In practice, applying Article 42 of the ICSID Convention means that the first sources of regulation are the provisions of the bilateral investment treaty (BIT) and the individual contract between the investor and the host state, as well as the domestic law of the host state. International law plays a secondary role in this scheme: it fills gaps in the national law or is used to correct national law if its application would violate peremptory norms of international public policy. For example, in World Duty Free Company v. Kenya (ICSID Award, 2006) (World Duty Free v. Kenya, 2006), the contract contained two conflicting choice-of-law clauses – one selecting Kenyan law (Article 10(A)) and another selecting English law (Article 9(2)(c)). The tribunal had to reconcile these clauses, find their common core regarding the prohibition of corruption, and apply both cumulatively. Guided by Article 42(1) of the ICSID Convention and taking into account international anti-corruption standards, the tribunal held the contract voidable for corruption and – given Kenya’s timely repudiation of the agreement – denied the investor’s claims. This case highlights that even where a national law is expressly chosen by the parties, ICSID tribunals must consider international legal constraints (such as the prohibition of corruption as a jus cogens norm) and adhere to international public policy principles to ensure a fair and lawful resolution.

Frequently, an investment dispute stems from a BIT, and the construction contract itself includes an applicable-law clause. Most BITs stipulate that disputes shall be resolved with regard to the treaty’s provisions and “applicable rules of international law.” When submitted to ICSID, such a clause is treated as the parties’ agreement on applicable law under Article 42(1) of the ICSID Convention. This was the approach in Bayindir v. Pakistan (ICSID Award on Jurisdiction, 2005) (Bayindir v. Pakistan, 2009), where the tribunal deemed the Turkey–Pakistan BIT of 1995 and general principles of international law to be the governing law, since the investment agreement itself did not reference any national law. At the same time, obligations arising from the construction contract with Pakistan’s National Highway Authority were governed by Pakistani law as the law of the contract; the tribunal considered those domestic norms only insofar as needed to determine whether the state breached the BIT (including evaluating the legality of the contract’s termination under the fair and equitable treatment standard). In the landmark case Salini Costruttori S.p.A. and Italstrade S.p.A. v. Morocco (ICSID Award, 2001) (Salini v. Morocco, 2001) concerning a FIDIC highway construction contract (Rabat–Fès), the ICSID tribunal noted that neither the contract nor the Italy–Morocco BIT contained an autonomous choice-of-law clause. Therefore, under the second sentence of Article 42(1) of the Convention, the law of the host state, combined with applicable international investment law norms (primarily the BIT itself), was applied. In practice, Moroccan civil and administrative law governed the assessment of the contractor’s performance of contractual obligations, whereas the question of the state’s violation of standards like “fair and equitable treatment” or “full protection and security” was decided by international law criteria, with the BIT considered a specialized source of such criteria.

Applicable Law in UNCITRAL and PCA Arbitrations. In ad hoc arbitrations under the UNCITRAL Rules (including those administered by the PCA), the applicable law is determined by the parties’ arbitration agreement and the UNCITRAL Rules themselves. Article 35 of the UNCITRAL Arbitration Rules provides: “The arbitral tribunal shall apply the rules of law designated by the parties as applicable to the substance of the dispute. Failing such designation, the tribunal shall apply the law which it determines to be appropriate.” Thus, if a dispute arises from an international construction contract governed by a specific national law (as is typical in FIDIC contracts, where the governing law of the project country is usually designated in the Particular Conditions), an UNCITRAL tribunal will apply that chosen national law. Conflict-of-laws analysis becomes relevant only if the parties have not designated any applicable law. In that event, arbitrators typically resort to general conflict principles, often favoring the law of the country with the closest connection to the dispute (for example, the law of the state where the construction is located or where the main obligations are to be performed).

In many investor–state disputes involving construction contracts, however, the applicable law is derived less from the contract and more from the BIT’s provisions. When an investor initiates an UNCITRAL arbitration alleging breaches of a BIT – whether expropriation, unfair treatment, or violation of an umbrella clause – the tribunal generally looks to the BIT text and international law. It treats the BIT itself as the parties’ agreement on applicable law, effectively mirroring the ICSID approach, albeit without the explicit mandate of Article 42 of the ICSID Convention. A telling example is Romak v. Uzbekistan (UNCITRAL Award, 2009) (Romak v. Uzbekistan, 2009). There, a Swiss company’s one-off grain sales under GAFTA standard contracts were claimed to be an “investment” under the Switzerland–Uzbekistan BIT. Applying the interpretative principles of the 1969 Vienna Convention on the Law of Treaties, the tribunal confirmed that a protected investment must involve contribution, duration, and risk. A simple transfer of goods did not meet these criteria. Consequently, in the absence of an “investment,” the law applied was only that relevant to determining jurisdiction – comprised of international law rules for BIT interpretation – and no national law governed the merits. By contrast, in disputes where the investment nature of a transaction is uncontested, purely contractual aspects may be additionally governed by domestic law.

Another example of a multi-tiered choice-of-law approach is the Eurotunnel arbitration (Channel Tunnel Group Ltd. & France Manche S.A. v. UK & France, PCA). The applicable law was defined by a layered conflict clause: pursuant to Article 19(6) of the 1986 Treaty of Canterbury (Fixed Link Treaty) and Clause 40.4 of the Concession Agreement, the tribunal applied, first, the Treaty and the Concession Agreement themselves; second, general principles of international law; and only thereafter would English or French law be considered, and then solely to the extent needed for the performance of specific obligations. Moreover, Clause 41.1 confirmed that national law operated subsidiarily “to the extent applicable.” The tribunal gave priority to the Treaty and Concession terms, characterized the claimants’ demands as contract claims, and only took national laws into account insofar as they fit within the agreed “concession legal regime” and did not contradict it. This demonstrates the PCA tribunal’s flexible blending of international and national elements in disputes where a private contract is embedded in a broader public-law framework.

Across all major arbitration fora, a common logic for determining applicable law emerges: if the parties have an agreement on governing law, it is respected; absent such agreement, arbitrators strive to identify the law most closely connected to the dispute. In investment disputes under ICSID, UNCITRAL or PCA auspices, the nearly invariable priority is given to the relevant investment treaty’s norms and international law, whereas domestic law is chiefly brought in to resolve strictly contractual issues (such as the existence and scope of obligations or the consequences of their breach). Differences between institutions are largely attributable to their constitutive instruments: for example, ICSID formally requires consideration of the host state’s law (which can be critical if the BIT leaves certain issues uncovered), whereas the UNCITRAL and PCA rules grant tribunals more latitude in choosing conflict-of-law methods. Yet, in practice, these differences have been virtually eliminated in recent decades: tribunals in all forums generally reach substantively similar outcomes, either applying the law expressly chosen by the parties or, if none, the host state’s law in combination with general principles of international law. The convergence of approaches is illustrated by cases involving interference by domestic courts in the enforcement of construction arbitration awards. In Saipem v. Bangladesh (ICSID Award, 2009) and ATA Construction v. Jordan (ICSID Award, 2010), ICSID tribunals characterized the annulment or obstruction of arbitral awards by national courts as a breach of the state’s international obligation to honor the BIT, even though the underlying contracts were governed by Bangladeshi and Jordanian law, respectively. In both cases, the tribunals relied on BIT provisions ensuring fair and equitable treatment of investments and on state responsibility norms, thereby demonstrating the primacy of international standards over domestic norms when investor rights are undermined.

Jurisdiction and Arbitrability of Investment Disputes. ICSID jurisdiction is based on the Washington Convention 1965 and the parties’ consent (typically via a clause in an investment treaty or directly in a contract with the state). A key requirement is the existence of an “investment” under Article 25(1) of the ICSID Convention, and the dispute must arise directly from that investment. In the context of construction projects, this means the contractor–investor must make a substantial contribution to the host state’s economic development – the construction of infrastructure usually meets this test. In the aforementioned Salini v. Morocco case, the tribunal famously articulated the Salini criteria for an investment: a contribution of capital or other assets, a certain project duration, an element of risk, and a contribution to the host country’s development. The highway construction contract in Morocco, financed by the investor (an Italian company) and involving technology transfer and risks, was held to qualify as an “investment.” This approach has been reaffirmed in numerous subsequent ICSID cases involving roads, energy facilities, real estate, etc. For example, in the above-discussed Bayindir v. Pakistan, as well as in Jan de Nul v. Egypt and Toto Costruzioni v. Lebanon (Toto v. Lebanon, 2012), the tribunals applied the Salini test. In Toto, an ICSID tribunal in a 2009 jurisdictional decision confirmed that a construction contract can constitute an “investment” under Article 25(1) by recognizing contribution, duration, and risk; simultaneously, it dismissed claims based purely on contractual obligations that did not implicate the exercise of sovereign prerogatives. By its final award in 2012, the ICSID tribunal in Toto dismissed all BIT claims on their merits, emphasizing the fundamental difference between contractual breaches and breaches of international investment law, noting that some delays were due to the contractor’s own changes, and splitting the arbitration costs between the parties.

Another jurisdictional prerequisite for ICSID is the presence of a “foreign investor” – a natural or legal person with nationality in a Contracting State other than the host state. Determining nationality can be legally complex, especially with multi-tier corporate ownership. In classic EPC contract disputes, this was rarely at issue, as contractors were typically incorporated in ICSID member states. However, arbitral practice shows that respondent states may challenge an investor’s foreign status by arguing loss of the required nationality or lack of effective foreign control in multi-layered corporate structures. The case of Autopista Concesionada de Venezuela v. Venezuela (ICSID Award, 2003) (Autopista v. Venezuela, 2003) is illustrative: the highway concessionaire was a Venezuelan company, but after a foreign corporation acquired a controlling stake, the parties agreed in the contract to treat the company as an investor from an ICSID Contracting State. The tribunal upheld jurisdiction, noting that Article 25(2)(b) of the ICSID Convention permits such contractual designation of foreign investor status regardless of formal place of incorporation. The only exception is where an investor cannot prove the required foreign nationality: for example, in Soufraki v. UAE (ICSID Award, 2004) (Soufraki v. UAE, 2004), the claim was dismissed because the claimant had lost his Italian nationality before filing. In all construction-related cases, ICSID tribunals independently verify the investor’s citizenship against Article 25’s requirements; finding that the claimant held the host state’s nationality at relevant dates obliges an ICSID tribunal to decline jurisdiction. However, that does not bar the investor from pursuing arbitration under UNCITRAL Rules or at the PCA, which are not constrained by Article 25 and instead follow the BIT’s text.

ICSID also requires that the dispute arise directly out of investment activities. In construction contracts, there are often simultaneously “contractual disputes” (e.g. over unpaid work or unlawful termination by the state purchaser) and potential “treaty claims” (e.g. expropriation if the state confiscates equipment or materials, or a breach of fair treatment due to arbitrary government interference). In such cases, a fundamental issue is delineating whether the tribunal is addressing only BIT violations or may also adjudicate claims based solely on the civil contract, which do not implicate international law norms.

ICSID practice has established a clear stance: a contractual forum selection clause (for instance, requiring submission of contract disputes to domestic courts or commercial arbitration) does not in itself divest ICSID of jurisdiction over BIT claims. An investor may bring claims for breach of an investment treaty even if the same facts also constitute a breach of the state’s contractual obligations. The seminal statement of this principle came from the ICSID ad hoc Annulment Committee in Vivendi (Compañía de Aguas del Aconquija) v. Argentina (Decision on Annulment, 2002), which held that a state cannot invoke an exclusive jurisdiction clause in a contract to avoid the characterization of its conduct as a breach of international law; this was reaffirmed in the resubmitted Vivendi case in 2007 (Vivendi v. Argentina, 2007). Therefore, inclusion of a clause in a construction or other investment contract conferring competence on domestic courts affects only the resolution of pure contract claims and does not exempt the host state from liability for breaches of its BIT obligations. Subsequent tribunals, including those in construction disputes, have adhered to this principle. For instance, in the previously discussed Bayindir v. Pakistan (ICSID Award, 2009), the state argued that the core dispute – termination of a road contract – should be heard in the local arbitration specified by the contract (under Pakistani law), and that there was “no independent BIT breach.” The ICSID tribunal rejected these arguments, noting that the claimant’s claims were based on the investment treaty (expropriation, fair treatment, etc.), and even if they overlapped with a contract dispute, that did not deprive the ICSID tribunal of competence. The tribunal also considered the issue of parallel proceedings: since Bayindir had not completed the contractually mandated arbitration but went straight to ICSID, no true “double” proceeding occurred. Similarly, in Toto Costruzioni v. Lebanon (ICSID Decision on Jurisdiction, 2009), a road contractor faced a contractual forum clause requiring disputes to be heard by Lebanon’s administrative courts. The contractor nonetheless brought claims under the Italy–Lebanon BIT alleging delays and government actions. The ICSID tribunal asserted jurisdiction over the BIT claims, finding that they were distinct from the purely contractual claims and were grounded in international law. In that decision, the tribunal also examined the applicability of the “fork-in-the-road” doctrine (under which an investor’s choice to litigate in national courts precludes later international arbitration if the BIT contains an exclusive choice provision). Because the Italy–Lebanon BIT contained no explicit fork-in-the-road bar (only a requirement of a 12-month local litigation period), the tribunal analyzed the investor’s conduct. Claims not previously litigated domestically were deemed admissible, whereas those identical to issues already decided by Lebanese courts were dismissed as inadmissible under the principle of ne bis in idem (no double relief for the same cause of action).

Many BITs – especially older ones – include a fork-in-the-road clause providing that if an investor submits a dispute to national courts or another chosen forum, it cannot later go to international arbitration over the same subject matter. This condition is highly significant for international construction disputes: a contractor who litigates in domestic courts or commercial arbitration on contract issues risks losing the right to ICSID or other treaty arbitration on the same dispute. A strict application of fork-in-the-road is exemplified by Pantechniki S.A. v. Albania (ICSID Award, 2009) (Pantechniki v. Albania, 2009). After riots in 1997 led to the looting of a Greek road contractor’s equipment in Albania, the investor first sued in Albanian courts for contract damages (invoking a contract clause on civil disturbance risks). The Albanian appellate court voided that clause under local law. The contractor then initiated an ICSID case, alleging breaches of the Greece–Albania BIT, including failure to protect investments and denial of justice. The sole arbitrator (Y. Paulsson) carefully examined the fork-in-the-road clause and the subject matter of the domestic vs. arbitral proceedings, concluding that both were based on the same facts and sought the same relief – compensation for the same harm – despite formally different legal bases (contract vs. treaty). The tribunal held that the investor had exhausted its recourse in national courts and thus lost the right to arbitrate under the BIT. The investor’s claims were dismissed not for lack of substantive merit, but because after pursuing national court litigation, the BIT could no longer be invoked.

Multi-Tier Dispute Resolution Mechanisms. International construction contracts (especially FIDIC contracts) almost always provide for a multi-step dispute resolution process: first, referral to the Engineer (the project’s technical administrator); then to a Dispute Adjudication Board (DAB) – in pre-2017 FIDIC editions – or a Dispute Avoidance/Adjudication Board (DAAB) in the 2017 FIDIC update; and only thereafter to arbitration. Likewise, BITs often require an attempt at amicable settlement or litigation in local courts for a period (e.g. 6 months or 1 year) before international arbitration. Non-compliance with these pre-arbitration steps is usually treated as a question of admissibility of the claim, rather than a bar to jurisdiction (ipso jure). In practice, both ICSID and UNCITRAL tribunals are reluctant to dismiss claims purely on procedural formalities if the pre-arbitration step has lost its purpose. For example, a 6-month negotiation period in a BIT is seen not as a jurisdictional barrier but as a procedural requirement that can be “implicitly waived” if the state participated in the dispute on the merits without objection or if negotiations would clearly be futile. In the earlier Bayindir v. Pakistan case, the claimant had notified the dispute and by the time the claim was registered, 6 months had passed – the condition was deemed satisfied. In SGS v. Pakistan (ICSID Decision on Jurisdiction, 2003) (SGS v. Pakistan, 2003), the tribunal held that the formal failure to observe a 12-month waiting period for negotiations (per Article 9 of the BIT) did not deprive it of jurisdiction, as the period had expired by the time jurisdiction was considered.

Regarding the FIDIC contract requirement to refer disputes to a DAB/DAAB first: this issue arose, for instance, in the Pantechniki case discussed above. There, the investor attempted to characterize the contract dispute as an expropriation without going through the DAB. The ICSID tribunal focused on other grounds (notably fork-in-the-road) and did not rule on the effect of bypassing the DAB. In commercial disputes, national courts have confronted similar scenarios. Notably, the Supreme Court of the Philippines considered this in Hutama–RSEA Joint Operations, Inc. v. Citra Metro Manila Tollways Corp. (decision dated April 24, 2009) (Hutama-RSEA v. Citra, 2009). The dispute arose from an EPC contract for the Skyway Project between the main contractor (Citra) and a subcontractor (Hutama-RSEA JO), after part of the contract price was not paid. When negotiations failed, the subcontractor filed for arbitration before the Philippine Construction Industry Arbitration Commission (CIAC) without first referring the matter to a DAB as required by Clause 20.4 of the EPC contract. The main contractor objected to CIAC’s jurisdiction due to non-compliance with the DAB precondition. CIAC, in its decision of August 30, 2005, affirmed its own jurisdiction. On appeal, the Philippine Court of Appeals on May 23, 2007, sided with the main contractor and vacated the CIAC award, holding that prior reference to the DAB was mandatory before arbitration. The subcontractor then appealed, and the Philippine Supreme Court reversed on April 24, 2009, reinstating the CIAC award and affirming CIAC’s authority to decide the dispute. The Supreme Court reasoned that the presence of an arbitration clause itself indicates the parties’ consent to arbitration, and non-compliance with the DAB step does not deprive CIAC of jurisdiction – though it may have other consequences, such as affecting cost allocation or warranting a stay of proceedings to allow the DAB process.

The Role of the PCA. The PCA, as an institution, does not impose additional jurisdictional requirements – they are defined either by the parties’ agreement (consent to arbitration) or the arbitral rules chosen. The PCA even administers inter-state disputes, but in the context of private investments it often acts as a registry for ad hoc arbitrations. In fact, many major UNCITRAL arbitrations involving states in the 2010s (for example, claims against Venezuela after it exited ICSID) were administered by the PCA. In such cases, the jurisdictional conditions are those set out in the applicable BIT. Notably, when resolving disputes under multilateral treaties or special agreements, PCA tribunals also follow the general logic described above. In the earlier Eurotunnel arbitration, besides applicable law issues, the tribunal considered whether certain claims fell outside the scope of the arbitration clause. Ultimately, some claims (e.g. those seeking to hold both governments jointly and severally liable for losses) were found to be outside jurisdiction, as they exceeded the states’ obligations under the Concession. This underlines that in the absence of a universal instrument like the ICSID Convention, a PCA tribunal’s jurisdiction is strictly defined by the parties’ consent.

The results of this study confirm that when choosing the applicable law and determining jurisdiction in international construction contracts, the optimal outcomes occur only by jointly accounting for the functional-sectoral criteria of investment activity formulated in ICSID practice. These criteria, initially systematized in the Salini precedent and detailed by Russian scholars on ICSID subject-matter jurisdiction (Tereshkova & Gadalov, 2022), reliably separate contractual claims from claims based on international investment law. The research demonstrates that this distinction becomes especially persuasive when construction-concession models combine elements of an EPC contract with an investment protection agreement; in such cases, a choice-of-law clause embedded in the FIDIC conditions can function as a “legal bridge” between the public-law regime and the commercial deal.

At the same time, practical findings from UNCITRAL ad hoc arbitrations validate concerns raised about the normative uncertainty of Caspian pipeline projects (Glikman & Mamedov, 2024): the absence of detailed agreed procedures for referring disputes to arbitration still increases transaction costs and encourages parallel proceedings. In the Eurasian context, these risks are exacerbated by the multi-tier treaty network of the Shanghai Cooperation Organisation (SCO), where – as Chinese authors rightly note – collisions have emerged between an overburdened BIT system and rising arbitration costs (Yanyan & Sisi, 2024).

The unifying role of party autonomy should be evaluated through the prism of globalization challenges highlighted in Russian “law of global development” research (Alferova, 2016). These works foreground the balance between state sovereignty and openness of the legal system. The Russian constitutional model already enshrines the primacy of international obligations; the evolution from a dualist to a monist approach has been analyzed comprehensively by O. N. Khlestov (Khlestov, 2021), and I. A. Umnova-Konyukhova has pointed to the need for institutional adaptation of constitutional mechanisms (Umnova-Konyukhova, 2016).

The internationalization of constitutional law and the “constitutionalization” of international agreements – as demonstrated in works by N. V. Varlamova and T. A. Vasilieva (Varlamova & Vasilieva, 2017) – increase national courts’ scrutiny of arbitral clauses. This, in turn, requires practitioners to carefully align choice-of-law anchors with public order mandates – a point repeatedly stressed by Anglo-American authors analyzing the effects of multi-level constitutional governance on cross-border contracts (Weiler & Wind, 2003).

The findings of this study dovetail with Hans Kelsen’s classic concept of the primacy of international law (Kelsen, 2003) and the contemporary doctrine of the “functional supremacy” of international law over fragmented national norms, advanced by James Crawford in Brownlie’s Principles of Public International Law (Crawford & Brownlie, 2019). However, the “twilight” trends of constitutionalism described by D. Grimm, P. Dobner, and M. Loughlin (Gri