¹ For a more complete list of international adjudicative bodies, see the synopsis of the Project on International Courts and Tribunals, archived at https://elaw.org/system/files/intl%20tribunals%20synoptic_chart2.pdf.

² Mnookin, Robert H. & Kornhauser, Lewis, Bargaining in the Shadow of the Law: The Case of Divorce, 88 Yale L.J. 950 (1979)CrossRefGoogle Scholar.

³ See J. Michael Greig, Andrew P. Owsiak & Paul F. Diehl, International Conflict Management (2019).

⁴ If rational litigants can reach a similar (or better) outcome through a less costly process, as opposed to a more costly one, they should use the less costly process. For similar reasoning, see Fearon, James D., Rational Explanations for War, 49 Int'l Org. 379 (1995)CrossRefGoogle Scholar.

⁵ For example, see Marc L. Busch, Overlapping Institutions, Forum Shopping, and Dispute Settlement in International Trade, 61 Int'l Org. 735 (2007).

⁶ For example, see the history of relations between Colombia and Venezuela, during which these states negotiated the submission of their territorial disagreement to an arbitral process on two occasions—first to the Spanish crown, and later to the Swiss Federal Council. See Gordon Ireland, Boundaries, Possessions, and Conflicts in South America (1971).

⁷ Case law within the law of the sea regime we study here can travel across disparate judicial bodies. To the extent that it differs, however, commitments to particular bodies will allow case law to influence out-of-court bargaining. Moreover, case law is not the only mechanism through which the declarations referenced here influence out-of-court bargaining. As we note and develop below, through such declarations, states also signal a more credible threat (1) to submit the dispute to a court (i.e., a greater commitment to judicial processes overall), and (2) to submit the dispute to a particular court, which—in necessarily moving from arbitration to adjudication (see details presented later)—(a) causes the disputants to lose some control over their dispute's management and outcome (i.e., incur greater costs), and (b) enhances the disputants’ ability to predict how the court will decide (i.e., the court's case law, as well as its reasoning, rules, process, and composition).

⁸ Richard B. Bilder, International Dispute Settlement and the Role of International Adjudication, 1 Emory J. Int'l Disp. Resol. 131, 141 (1986–1987).

⁹ UNCLOS opened for signature in December 1982 and entered into force in November 1994. A regime is a set “of implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge in a given area of international relations.” See Robert O. Keohane, After Hegemony 57 (1984) (citation omitted).

¹⁰ See J.G. Merrills, International Dispute Settlement (4th ed. 2005).

¹¹ This elucidates an ad hoc arbitration procedure to handle disputes specifically over fisheries, environmental protection, scientific research, or navigation.

¹² Natalie Klein, Dispute Settlement in the UN Convention on the Law of the Sea (2005).

¹³ Because states parties can register a preference for multiple compulsory mechanisms, the declarations in favor of individual forums will not sum to the total number of states parties’ making Article 287 declarations. For a detailed breakdown, see Tables A4–5 in the online appendix. We compile these data from the United Nations, at http://www.un.org/depts/los/settlement_of_disputes/choice_procedure.htm. The list of states comes from the Correlates of War (COW) Project, State System Membership dataset, at http://www.correlatesofwar.org/data-sets/folder_listing. Cabo Verde, Cook Islands, Eswatini, Niue, Palestine, or the European Union do not qualify as states under the COW Project criteria. We therefore exclude these actors from our analyses, even though they, too, have made commitments to UNCLOS.

¹⁴ Our dataset captures both peaceful and militarized attempts to settle (or address) interstate maritime disputes. These data allow for non-binding, peaceful attempts (e.g., negotiation), as well as militarization (or the use of force), to occur simultaneously with binding, peaceful attempts (e.g., adjudication or arbitration).

¹⁵ We use “last resort” in a psychological, as opposed to temporal sense. States admittedly do not always use force temporally last among the various management tools available. Rather, our point is that, because force costs more than the alternatives, states will prefer to use the alternative, less costly tools first—but only if they perceive a chance of obtaining their goals through such tools. A state will often use force to bolster its bargaining position. Courts, however, discourage this use of force too because they do not permit such behavior to prejudice court judgments.

¹⁶ Karen J. Alter, Emilie M. Hafner-Burton & Laurence Helfer, Theorizing the Judicialization of International Relations, 63 Int'l Stud. Q. 449 (2019); Paul R. Hensel, Sara McLaughlin Mitchell, Thomas E. Sowers II & Clayton L. Thyne, Bones of Contention: Comparing Territorial, Maritime, and River Issues, 52 J. Conflict Resol. 117 (2008); Paul. R. Hensel & Sara McLaughlin Mitchell, From Territorial Claims to Identity Claims: The Issue Correlates of War (ICOW) Project, 34 Conflict Mgmt. & Peace Sci. 126 (2017).

¹⁷ ICOW identifies 270 interstate maritime claims during the period 1900–2010. For a description of these claims, see Sara McLaughlin Mitchell, Clashes at Sea: Explaining the Onset, Militarization, and Resolution of Diplomatic Maritime Claims, 29 Sec. Stud. 637 (2020).

¹⁸ This does not mean that the target state holds a valid legal claim to the maritime area in question, only that it exercises status quo control over that area. If it did not, the challenger would have no counterpart to challenge.

¹⁹ States involved in a maritime disagreement use military force at some point to manage that disagreement in 29% of diplomatic maritime claims. See Mitchell, supra note 17. The data we use in this study contain 143 dyadic maritime claims in the Western Hemisphere and Europe from 1900–2001. Of these, 115 (or 80.4%) begin in or after 1945. See Hensel & Mitchell, supra note 16. Ninety militarized disputes directly relate to these claims.

²⁰ Delimitation of the Maritime Boundary in the Gulf of Maine Area (U.S. v. Can.), 1984 ICJ Rep. 246 (Oct. 12); Fisheries Jurisdiction (UK v. Ice.), Merits, 1974 ICJ Rep. 3 (July 25); Corfu Channel (UK v. Alb.), 1949 ICJ Rep. 4 (Apr. 9). For an in-depth introduction to the ICOW maritime claims dataset, see Mitchell, supra note 17.

²¹ ICOW identifies 935 peaceful attempts (i.e., negotiations, mediations, conciliations, arbitrations, and adjudications) to settle 270 distinct maritime claims (see Mitchell, supra note 17). States use arbitration and adjudication, which we classify together as “binding forms” of peaceful settlement, in twenty-nine of these 270 diplomatic maritime conflicts.

²² See Mnookin & Kornhauser, supra note 2.

²³ Alter, Hafner-Burton & Helfer, supra note 16.

²⁴ Karen J. Alter, The New Terrain of International Law: Courts, Politics, Rights (2014).

²⁵ As Alter notes, a large percentage of the cases that international courts hear involve other judicial activities, such as administrative review, enforcement, and constitutional review. See id. at 5.

²⁶ Karen J. Alter, The Multiplication of International Courts and Tribunals After the End of the Cold War, in The Oxford Handbook of International Adjudication (Cesare P.R. Romano, Karen J. Alter & Yuval Shany eds., 2014).

²⁷ Alter, Hafner-Burton, and Helfer, supra note 16, describe four phases of judicialized politics: shadow politics, adjudication politics, compliance politics, and feedback politics. Our theory focuses on the first two phases. Shadow politics “refers to mobilization, bargaining, negotiations, and responses generated by a plausible threat of adjudication” (id. at 454), while adjudication politics “encompasses the factors, strategies, and consequences associated with the decision to adjudicate, including which suits are filed, the selection of venue, the gathering of evidence and presentation of arguments, and the decisions of judges, arbitrators, and other adjudicatory bodies” (id. at 455).

²⁸ Id.

²⁹ Tom Ginsburg & Richard H. McAdams, Adjudicating in Anarchy: An Expressive Theory of International Dispute Resolution, 45 Wm. & Mary L. Rev. 1229 (2004); Sara McLaughlin Mitchell & Emilia J. Powell, Domestic Law Goes Global: Legal Traditions and International Courts 217 (2011).

³⁰ Busch, supra note 5; Marc L. Busch & Eric Reinhardt, Bargaining in the Shadow of the Law: Early Settlement in GATT/WTO Disputes, 24 Fordham Int'l L.J. 158 (2000); Marc L. Busch & Eric Reinhardt, Three's a Crowd: Third Parties and WTO Dispute Settlement, 58 World Pol. 446 (2006); Christina L. Davis, Why Adjudicate?: Enforcing Trade Rules in the WTO (2012); Christina L. Davis & Sarah Blodgett Bermeo, Who Files? Developing Country Participation in GATT/WTO Adjudication, 71 J. Pol. 1033 (2009); Julia Gray & Philip Potter, Diplomacy and the Settlement of International Trade Disputes, 64 J. Conflict Resol. 1358 (2020); Eric Reinhardt, Adjudication Without Enforcement in GATT Disputes, 45 J. Conflict Resol. 174 (2001); Gregory C. Shaffer, Michell Ratton Sanchez & Barbara Rosenberg, The Trials of Winning at the WTO: What Lies Behind Brazil's Success, 41 Cornell Int'l L.J. 383 (2008); Richard H. Steinberg, In the Shadow of Law or Power? Consensus-Based Bargaining and Outcomes in the GATT/WTO, 56 Int'l Org. 339 (2002). See also research on international investment arbitration, such as Emilie M. Hafner-Burton, Sergio Puig & David G. Victor, Against Secrecy: The Social Cost of International Dispute Settlement, 42 Yale J. Int'l L. 279 (2017); Emilie M. Hafner-Burton, Zachary C. Steinert-Threlkeld & David G. Victor, Predictability Versus Flexibility: Secrecy in International Investment Arbitration, 68 World. Pol. 413 (2016).

³¹ Mitchell & Powell, supra note 29; Benjamin J. Appel, In the Shadow of the International Criminal Court: Does the ICC Deter Human Rights Violations?, 62 J. Conflict Resol. 3 (2018); Courtney Hillebrecht, The Deterrent Effects of the International Criminal Court: Evidence from Libya, 42 Int'l Interactions 616 (2016); Kathryn Sikkink, The Justice Cascade: How Human Rights Prosecutions Are Changing World Politics (2011); Jacqueline R. McAllister, Deterring Wartime Atrocities: Hard Lesson from the Yugoslav Tribunal, 44 Int'l Sec. 84 (2020).

³² Karen J. Alter, Do International Courts Enhance Compliance with International Law?, 25 Rev. Asian & Pac. Stud. 51 (2003); Laurence R. Helfer, Why States Create Independent Tribunals: A Theory of Constrained Independence, 23 Conf. New Pol. Econ. 253 (2006).

³³ See Ginsburg & McAdams, supra note 29.

³⁴ See Bilder, supra note 8; Krzysztof J. Pelc, The Politics of Precedent in International Law: A Social Network Application, 108 Am. Pol. Sci. Rev. 547 (2014).

³⁵ As an illustration, only 37% of state members to the United Nations recognize the ICJ's optional clause for compulsory jurisdiction. Fewer than 25% of states parties to UNCLOS recognize ITLOS's jurisdiction under an optional Article 287 declaration (see Tables A4–5, online appendix).

³⁶ See Davis & Bermeo, supra note 30; Eric A. Posner & John C. Yoo, Judicial Independence in International Tribunals, 93 Cal. L. Rev. 1 (2005); Áslaug Ásgeirsdóttir & Martin C. Steinwand, Dispute Settlement Mechanisms and Maritime Boundary Settlements, 10 Rev. Int'l Org. 119 (2015).

³⁷ Joost Pauwelyn & Luiz Eduardo Salles, Forum Shopping Before International Tribunals: (Real) Concerns, (Im)possible Solutions, 42 Cornell Int'l L.J. 77 (2009).

³⁸ See Posner & Yoo, supra note 36, at 28.

³⁹ This is best done via a single issue area—in our case, the law of the sea. For a similar approach, see Mitchell & Powell, supra note 29; Busch & Reinhardt, supra note 30; Gray & Potter, supra note 30; Appel, supra note 31.

⁴⁰ Ian Brownlie, Principles of Public International Law 180 (6th ed. 2003); Jack L. Goldsmith & Eric A. Posner, The Limits of International Law 59 (2005).

⁴¹ Although the three-mile limit was customary law, Scandinavia, Spain and Portugal, and Russia pursued four-, six-, and up to 100-mile limits. Id. at 60.

⁴² Earlier conventions laid the groundwork for UNCLOS, but left many noteworthy ambiguities. For example, the Convention on the Territorial Sea and the Contiguous Zone (1958) sets no limit on the territorial sea—other than that it must logically be less than twelve miles (Article 24). Similarly, the Convention on the High Seas overlaps the high seas with the contiguous zone, something that UNCLOS later corrected (Article 86). Convention on the Territorial Sea and the Contiguous Zone, Apr. 29, 1958, 516 UNTS 205; Convention on the High Seas, Apr. 29, 1958, 450 UNTS 11; Convention on Fishing and Conservation of the Living Resources of the High Seas, Apr. 29, 1958, 559 UNTS 285; Convention on the Continental Shelf, Apr. 29, 1958, 499 UNTS 311; United Nations Convention on the Law of the Sea, Dec. 10, 1982, 1833 UNTS 397; see also Jane's Exclusive Economic Zones (Martin Pratt & Clive Schofield eds., 2d ed. 2000).

⁴³ See Klein, supra note 12, at 33; Donald R. Rothwell & Tim Stephens, The International Law of the Sea 439 (2010).

⁴⁴ See Klein, supra note 12; Rothwell & Stephens, supra note 43.

⁴⁵ “States Parties shall settle any dispute between them concerning the interpretation or application of this Convention by peaceful means in accordance with Article 2, paragraph 3, of the Charter of the United Nations and, to this end, shall seek a solution by the means indicated in Article 33, paragraph 1, of the Charter.” United Nations Convention on the Law of the Sea, supra note 42, Art. 279.

⁴⁶ “1. When a dispute arises between States Parties concerning the interpretation or application of this Convention, the parties to the dispute shall proceed expeditiously to an exchange of views regarding its settlement by negotiation or other peaceful means. 2. The parties shall also proceed expeditiously to an exchange of views where a procedure for the settlement of such a dispute has been terminated without a settlement or where a settlement has been reached and the circumstances require consultation regarding the manner of implementing the settlement.” Id. Art. 283.

⁴⁷ “Nothing in this Part impairs the right of any States Parties to agree at any time to settle a dispute between them concerning the interpretation or application of this Convention by any peaceful means of their own choice.” Id. Art. 280.

⁴⁸ “1. If the States Parties which are parties to a dispute concerning the interpretation or application of this Convention have agreed to seek settlement of the dispute by a peaceful means of their own choice, the procedures provided for in this Part apply only where no settlement has been reached by recourse to such means and the agreement between the parties does not exclude any further procedure. 2. If the parties have also agreed on a time-limit, paragraph 1 applies only upon the expiration of that time-limit.” Id. Art. 281.

⁴⁹ “If the States Parties which are parties to a dispute concerning the interpretation or application of this Convention have agreed, through a general, regional or bilateral agreement or otherwise, that such dispute shall, at the request of any party to the dispute, be submitted to a procedure that entails a binding decision, that procedure shall apply in lieu of the procedures provided for in this Part, unless the parties to the dispute otherwise agree.” Id. Art. 282.

⁵⁰ See the discussion of the Southern Bluefin Tuna and Mox Plant cases in Klein, supra note 12, which illustrates how dispute settlement under Part XV of UNCLOS intersects both with states parties’ commitments under other treaties and their obligations with respect to Article 282.

⁵¹ “1. A State Party which is a party to a dispute concerning the interpretation or application of this Convention may invite the other party or parties to submit the dispute to conciliation in accordance with the procedure under Annex V, section 1, or another conciliation procedure. 2. If the invitation is accepted and if the parties agree upon the conciliation procedure to be applied, any party may submit the dispute to that procedure. 3. If the invitation is not accepted or the parties do not agree upon the procedure, the conciliation proceedings shall be deemed to be terminated. 4. Unless the parties otherwise agree, when a dispute has been submitted to conciliation, the proceedings may be terminated only in accordance with the agreed conciliation procedure.” United Nations Convention on the Law of the Sea, supra note 42, Art. 284.

⁵² “Subject to section 3, any dispute concerning the interpretation or application of this Convention shall, where no settlement has been reached by recourse to section 1, be submitted at the request of any party to the dispute to the court or tribunal having jurisdiction under this section.” Id. Art. 286. See also Rothwell & Stephens, supra note 43, at 445–46.

⁵³ Twenty-one judges comprise ITLOS, which has broader ratione personae jurisdiction than the ICJ because it recognizes the standing of international organizations. See Hugo Caminos, The International Tribunal for the Law of the Sea: An Overview of Its Jurisdictional Procedure, 5 L. & Prac. Int'l Cts. & Tribs. 13 (2006).

⁵⁴ Klein notes that “the appeal of arbitration may be found in the possibility of secrecy, party control over the composition of the tribunal and the questions addressed to the tribunal as well as the ability to avoid a third State's intervention in the proceedings.” Klein, supra note 12, at 56. A precursor to the UNCLOS compulsory dispute settlement process appears in the Optional Protocol of Signature Concerning the Compulsory Settlement of Disputes. That protocol, however, privileged the ICJ (as a default judicial body), remained underdeveloped, and as its title would suggest, was optional. Optional Protocol of Signature Concerning the Compulsory Settlement of Disputes, Apr. 29, 1958, 450 UNTS 169.

⁵⁵ See Klein, supra note 12, at 57.

⁵⁶ For the optional declarations under UNCLOS, see the specific website for the United Nations Convention on the Law of the Sea (1982), at https://treaties.un.org/pages/ViewDetailsIII.aspx?src=TREATY&mtdsg_no=XXI-6&chapter=21&Temp=mtdsg3&clang=_en.

⁵⁷ We gather these data from the declarations that states parties to UNCLOS have made as of December 31, 2020. Id.

⁵⁸ See Rothwell & Stephens, supra note 43, at 449.

⁵⁹ Agreement for the Implementation of the Provisions of the United Nations Convention on the Law of the Sea of 10 December 1982 Relating to the Conservation and Management of Straddling Fish Stocks and Highly Migratory Fish Stocks, Art. 30, Aug. 4, 1995, 2167 UNTS 3, at https://treaties.un.org/pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXI-7&chapter=21&clang=_en; United Nations Educational, Scientific, and Cultural Organization, Records of the General Conference, Sess. 31, Vol. I: Resolutions, 50, 56 (2001), at https://unesdoc.unesco.org/ark:/48223/pf0000124687.page=56; see also Helmut Tuerk, The Contribution of the International Tribunal for the Law of the Sea to International Law, in Maritime Boundary Disputes, Settlement Processes, and the Law of the Sea 253 (2009).

⁶⁰ Caminos, supra note 53, at 19.

⁶¹ United Nations Convention on the Law of the Sea, supra note 42, Art. 297.

⁶² Id. Art. 298; Ted L. McDorman, An Overview of International Fisheries Disputes and the International Tribunal for the Law of the Sea, 40 Can. Y.B. Int'l L. 119 (2002); see also Klein, supra note 12, at 227.

⁶³ Lori Fisler Damrosch, Military Activities in the UNCLOS Compulsory Dispute Settlement System: Implications of the South China Sea Arbitration for U.S. Ratification of UNCLOS, 110 AJIL Unbound 273 (2016–2017).

⁶⁴ A single state party can initiate conciliation with another state party under Article 298, but only with regards to the first category of exceptions (para. 1(a); boundary delimitations maritime boundaries, or historic bays or titles). See United Nations Convention on the Law of the Sea, supra note 42, Art. 298, para. 1(a); Klein, supra note 12, at 257. This has occurred once to date, in the 2018 East Timor/Australia Timor Sea Conciliation (see Table A1, online appendix).

⁶⁵ Anne Sheehan, Dispute Settlement Under UNCLOS: The Exclusion of Maritime Delimitation Disputes, 24 U. Queensland L.J. 165, 183 (2005).

⁶⁶ Figure 1 suggests that fewer states parties make Article 298 declarations than we claim here. The time series data presented in the figure rely on the Correlates of War (COW) Project's State System Membership dataset, supra note 13. The COW Project's criteria for such membership causes it to omit micro-states (e.g., smaller Pacific or Caribbean islands). This explains the apparent discrepancy between the figure and text (see also Tables A4–5, online appendix).

⁶⁷ See Mitchell, supra note 17, at 645; Treaty Between Uruguay and Argentina Concerning the Rio de la Plata and the Corresponding Maritime Boundary (1973), available at https://www.un.org/Depts/los/LEGISLATIONANDTREATIES/PDFFILES/TREATIES/URY-ARG1973MB.PDF.

⁶⁸ Our approach builds on the idea that we can learn more about disputes by studying the “naming,” “blaming,” and “claiming” processes, as well as how conflicts arise and progress through each stage. As we note earlier, our data focuses on interstate disputes; it does not consider non-controversial, law enforcement actions that states take against private citizens within or beyond their territorial seas. On the dispute origination process, see William L.F. Felstiner, Richard L. Abel & Austin Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming …, 15 L. & Soc'y Rev. 631 (1980–81).

⁶⁹ International relations researchers generally consider the major states in the post-World War II period to be the United States (1945–present), the United Kingdom (1945–present), France (1945–present), the Soviet Union/Russia (1945–present), China (1949–present), Germany (1990–present), and Japan (1949–present). See COW State System Membership dataset, supra note 13.

⁷⁰ Following the standard approach in quantitative international conflict studies, our unit of analysis is the dyad-year. A dyad-year consists of two states paired with a year—for example, United States-Canada 1970, United States-Canada 1971, and so on. The first criteria generates 93,047 dyad-years, while the second (contains major power) adds another 32,944 dyad-years. As Mitchell notes, supra note 17, ICOW researchers begin with state-state pairs that could have overlapping maritime zones, as identified through Jane's Exclusive Economic Zones, supra note 43. They then use various documents (e.g., newspapers, history books, journal articles, and government documents) to discover the cases in which diplomatic conflicts (or in the ICOW Project's terminology, “maritime claims”) occurred. These cases include competing claims over maritime boundary delimitation, the legal bases for coastal states’ claims (e.g., the status of rocks vs. islands, or historical rights), access to marine resources, and navigational issues.

⁷¹ The Hecate Strait lies between the Canadian Queen Charlotte Islands and the British Columbia mainland. The Dixon Entrance lies between Queen Charlotte Island and the Alaskan Prince of Wales Island. For claim dates using the criteria in the discussion that follows, see Hensel, Mitchell, Sowers & Thyne, supra note 16.

⁷² David Nowell, The Canada-United States Dispute in Dixon Entrance: History, Current Issues and Prospects for Conflict Resolution, International Boundaries and Boundary Conflict Resolution: Proceedings of the 1989 IBRU Conference (Carl Grundy-Warr ed., 1990).

⁷³ See McDorman, supra note 62, at 373.

⁷⁴ We do not develop a theory of maritime claim onset here, but rather examine how legalization and judicialization influence bargaining over existing maritime claims. Daniels and Mitchell, using a dataset with broader spatial coverage, directly examine the question of maritime claim onset. The authors show that diplomatic conflicts over maritime areas are more likely to occur between more powerful, more democratic, and more economically advanced states, as well as those with a history of militarized conflict. Dyads containing UNCLOS states parties are also less likely to experience the onset of a new maritime claim, a finding we confirm below. See Kelly Daniels & Sara McLaughlin Mitchell, Bones of Democratic Contention: Maritime Disputes, 20 Int'l Area Stud. Rev. 293, 296 (2017).

⁷⁵ Delimitation of the Maritime Boundary in the Gulf of Maine Area, supra note 20.

⁷⁶ Facts on File, 8/1/1991. This is not a non-controversial law enforcement action because the incident occurs in disputed waters.

⁷⁷ See Hensel, Mitchell, Sowers & Thyne, supra note 16; Hensel & Mitchell, supra note 16.

⁷⁸ See Goldsmith & Posner, supra note 40, at 23.

⁷⁹ See Posner & Yoo, supra note 36.

⁸⁰ See Bilder, supra note 8.

⁸¹ For a foundational treatment of legalization, see Kenneth W. Abbott, Robert O. Keohane, Andrew Moravcsik, Anne-Marie Slaughter & Duncan Snidal, The Concept of Legalization, 54 Int'l Org. 401 (2000).

⁸² Although the obligation, strictly speaking, only becomes legal after ratification (i.e., once a state party), signing a convention indicates a non-negligible acceptance of the standards a convention contains. For this reason, we explore the effects of both signing and becoming a state party to (i.e., ratifying) UNCLOS. If signing UNCLOS does not alter state behavior, then the empirical analysis will indicate that.

⁸³ A focal point solves a coordination problem because it allows bargaining actors’ expectations to converge; it often takes the form of a conspicuous option among many alternatives (e.g., a convention) and “can defend itself with the argument ‘If not here, where?’” See Thomas C. Schelling, The Strategy of Conflict 70 (1960). Through its ability to act as a focal point, the UNCLOS regime's existence may also affect states not parties, especially given the large number of states parties to UNCLOS (see Table A4, online appendix). Our empirical models consider this possibility by including a systemic measure of UNCLOS (i.e., whether UNCLOS exists or has entered into force).

⁸⁴ In our empirical analyses, however, we find that signing (but not yet ratifying) UNCLOS influences the onset and management of dyadic claims more so than ratifying UNCLOS. In other words, the existence of the treaty—at a systemic level—changed most states’ dyadic diplomatic behavior even before the treaty legally entered into force.

⁸⁵ For additional evidence supporting this argument, see Stephen C. Nemeth, Sara McLaughlin Mitchell, Elizabeth A. Nyman & Paul R. Hensel, Ruling the Sea: Managing Maritime Conflicts Through UNCLOS and Exclusive Economic Zones, 40 Int'l Interactions 711 (2014).

⁸⁶ On institutions, see Keohane, supra note 9. On organizations, see Kenneth W. Abbott & Duncan Snidal, Why States Act Through Formal International Organizations, 42 J. Conflict Resol. 3 (1998); Sara McLaughlin Mitchell & Paul R. Hensel, International Institutions and Compliance with Agreements, 51 Am. J. Pol. Sci. 721 (2007).

⁸⁷ Evidence supports this position. For example, as two states each belong to a greater number of “peace brokering intergovernmental organizations,” the likelihood rises that they experience peaceful conflict management in their territorial dispute, particularly management that involves a third party. See Megan Shannon, Preventing War and Providing the Peace International Organizations and the Management of Territorial Disputes, 26 Conflict Mgmt. & Peace Sci. 144 (2009).

⁸⁸ See Alter, supra note 24, at 64.

⁸⁹ The Grisbådarna Case (Nor. v. Swe.), Award of the Tribunal (Perm Ct. Arb. Oct. 23, 1909), available at http://www.worldcourts.com/pca/eng/decisions/1909.10.23_Norway_v_Sweden.pdf.

⁹⁰ Fisheries Case (UK v. Nor.), 1951 ICJ Rep. 116 (Dec. 18), available at http://www.icj-cij.org/docket/files/5/1811.pdf.

⁹¹ As Schultz notes, unresolved territorial disputes create other types of uncertainty, such as a lack of clear guidance about property rights for businesses and investors. This uncertainty curtails business dealings, including foreign direct investment. Many unresolved maritime claims have a similar effect, especially when they involve the extraction of natural resources (e.g., disputes in the South China Sea and Gulf of Tonkin). In such cases, ambiguities concerning sovereignty complicate negotiations between the disputing states and private companies (e.g., Suriname-Guyana), thereby stalling economic development. See Kenneth A. Schultz, Borders, Conflict, and Trade, 18 Ann. Rev. Pol. Sci. 125 (2015); Beth A. Simmons, Trade and Territorial Conflict in Latin America: International Borders as Institutions, in Territoriality and Conflict in an Era of Globalization 251 (Miles Kahler & Barbara Walter eds., 2006); Hoon Lee & Sara McLaughlin Mitchell, Foreign Direct Investment and Territorial Disputes, 56 J. Conflict Resol. 675 (2012).

⁹² See Goldsmith & Posner, supra note 40.

⁹³ International courts like the ICJ do not follow the doctrine of stare decisis. Judges may consider previous rulings when hearing cases, but they are not legally bound by their (or other) court's decisions. Nevertheless, most international courts consider previous judgments and rarely contradict the logic of their (or other) court's rulings. A norm of common law adjudication consequently exists in the international system. See Chester Brown, A Common Law of International Adjudication (2007).

⁹⁴ See Mnookin & Kornhauser, supra note 2; Ginsburg & McAdams, supra note 29.

⁹⁵ Paul K. Huth, Sarah E. Croco & Benjamin J. Appel, Bringing Law to the Table: Legal Claims, Focal Points, and the Settlement of Territorial Disputes Since 1945, 57 Am. J. Pol. Sci. 90 (2013).

⁹⁶ Gilligan, et al. argue that strong courts could reduce states’ incentives to negotiate and to reveal information in the pre-trial stage, thereby deleteriously affecting bargaining. Given the highly legalized character of UNCLOS, we believe judicialization further strengthens the regime by clarifying ambiguities that the treaty did not initially address. A typical maritime dispute case adjudicated through the ICJ, for example, reconciles (a) competing focal point(s) on matters that UNCLOS left unclear (up to the time of the case). The setting of or selecting among these focal points (i.e., case law) then benefits both states parties and non-parties alike. Michael Gilligan, Leslie Johns & B. Peter Rosendorff, Strengthening International Courts and the Early Settlement of Disputes, 54 J. Conflict Resol. 5 (2010).

⁹⁷ See Klein, supra note 12, at 59.

⁹⁸ See Ginsburg & McAdams, supra note 29, at 1319–22; see also North Sea Continental Shelf Cases (W. Ger. v. Den.; W. Ger. v. Neth.), 1969 ICJ Rep. 3 (Feb. 20), available at https://www.icj-cij.org/public/files/case-related/52/052-19690220-JUD-01-00-EN.pdf.

⁹⁹ See Convention on the Continental Shelf, supra note 42, Art. 6. To complicate matters, Article 12 of the convention allows states parties to make reservations on, inter alia, Article 6, thereby exempting them from its provisions. Nevertheless, Germany was not a state party, so the legal question engaged whether the equidistance principle contained in Article 6 was part of customary international law. The court determined it was not. For further discussion, see L. D. M. Nelson, The North Sea Continental Shelf Cases and Law-Making Conventions, 35 Mod. L. Rev. 52 (1972); North Sea Continental Shelf Cases, supra note 98.

¹⁰⁰ Mnookin & Kornhauser, supra note 2, refer to these as transaction costs; they can be financial, emotional, or psychological. For a deeper comparison among conflict management tools, see Greig, Owsiak & Diehl, supra note 3. On judicial processes as political cover, see Todd L. Allee & Paul K. Huth, Legitimizing Dispute Settlement: International Legal Rulings as Domestic Political Cover, 100 Am. Pol. Sci. Rev. 219 (2006).

¹⁰¹ According to Mitchell & Owsiak, the average ICJ case—from initial application to judgment—unfolds over approximately four years. See Sara McLaughlin Mitchell & Andrew P. Owsiak, The International Court of Justice, in The Routledge Handbook of Judicial Behavior 445, 452 (Robert M. Howard & Kirk A. Randazzo eds., 2017). Davis & Bermeo, supra note 30, similarly demonstrate that high start-up costs prevent less powerful states from using the GATT/WTO judicial process. On international legal traditions and international courts, see Mitchell & Powell, supra note 29.

¹⁰² Compliance rates with international courts are high, and multilateral pressures often enforce such compliance. States should therefore expect to comply with a court judgment they receive, or incur significant reputation costs from non-compliance. See Mitchell & Hensel, supra note 86.

¹⁰³ Stephen E. Gent & Megan Shannon, Decision Control and the Pursuit of Binding Conflict Management: Choosing the Ties that Bind, 55 J. Conflict Resol. 710 (2011).

¹⁰⁴ Leslie Johns, Courts as Coordinators: Endogenous Enforcement and Jurisdiction in International Adjudication, 56 J. Conflict Resol. 257 (2012).

¹⁰⁵ Songying Fang, The Strategic Use of International Institutions in Dispute Settlement, 5 Q. J. Pol. Sci. 107 (2010).

¹⁰⁶ On the term “last resort,” see comments at note 15 supra.

¹⁰⁷ Research on the ICJ uncovers a similar dynamic to the one argued here. If two states have each made an optional declaration that recognizes the ICJ's jurisdiction, they are more likely to strike and comply with agreements to resolve their ongoing diplomatic conflicts. See Mitchell & Powell, supra note 29.

¹⁰⁸ This implies that states who make Article 287 declarations should (try to) use the forums specified in their declarations, if they go to court. For this reason, we do not find the growing use of Annex VII arbitration worrisome for our argument—unless alternate judicial bodies were acceptable and preferable to the default procedure for both disputants. Future research might explore this further as the ICOW maritime claims dataset expands.

¹⁰⁹ This definition of politically relevant dyads has a long history in quantitative research on international conflict. See Douglas Lemke & William Reed, The Relevance of Politically Relevant Dyads, 45 J. Conflict Resol. 126 (2001). On major states, see note 69 supra. See also COW State System Membership dataset, supra note 13.

¹¹⁰ See Hensel, Mitchell, Sowers & Thyne, supra note 16.

¹¹¹ Although not yet available, preliminary estimates from the remaining regions suggest that 270 total dyadic maritime claims existed globally from 1900–2001. See Hensel & Mitchell, supra note 16; Mitchell, supra note 17. UNCLOS provisions articulate coastal states’ rights more clearly than the guidelines for delimiting overlapping zones. In previous work, we argue that states manage the delimitation of overlapping maritime boundary cases in ways similar to traditional territorial (i.e., boundary) issues, and that states therefore find EEZ maritime claims, on average, more salient than non-EEZ maritime claims. The empirical evidence in that earlier study is consistent with such an argument; states attempt settlement—via both peaceful and non-peaceful (i.e., military) mechanisms—more often in their EEZ, as opposed to non-EEZ, maritime claims, placing the former on par with the management of their more salient territorial disputes. The creation of optional exceptions for maritime boundaries further demonstrates the salience of delimitation issues (Article 298). Andrew P. Owsiak & Sara McLaughlin Mitchell, Conflict Management in Land, River, and Maritime Claims, 7 Pol. Sci. Res. & Methods 43 (2019).

¹¹² Faten Ghosn, Glenn Palmer & Stuart A. Bremer, The MID3 Data Set, 1993–2001: Procedures, Coding Rules, and Description, 21 Conflict Mgmt. & Peace Sci. 133 (2004).

¹¹³ Twenty-eight percent of ICOW maritime claims experience at least one MID. More generally, 2.4% of politically relevant dyads experience at least one MID. The discrepancy occurs because (1) not all politically relevant dyads historically have maritime claims, and (2) not all MIDs in politically relevant dyads map onto a specific diplomatic maritime claim. See Hensel and Mitchell, supra note 16.

¹¹⁴ Strictly speaking, signatories are not “states parties” to UNCLOS—meaning that the signature carries a potential signaling function, but no true legalization or judicialization. Whether signing UNCLOS (i.e., the signal) affects bargaining behavior remains an empirical question, one that we investigate.

¹¹⁵ Alternative versions of the empirical models accounted for Article 298 declarations as well. The Article 298 declarations did not affect the peaceful or militarized management of maritime claims. We therefore present models in this study that exclude Article 298 declarations—a decision that simplifies the theoretical and empirical discussion greatly.

¹¹⁶ See Lee & Mitchell, supra note 91; Mitchell & Powell, supra note 29, at 113; Hensel, Mitchell, Sowers & Thyne, supra note 16.

¹¹⁷ See also Nemeth, Mitchell, Nyman & Hensel, supra note 85.

¹¹⁸ More specifically, we control for “peace years.” Conflict scholars propose that the longer a dyad remains at peace (i.e., has no MID), the less likely it is at any given time to start one. For a detailed discussion of this variable's coding and source, see Table B1 (online appendix).

¹¹⁹ For similar results, see Nemeth, Mitchell, Nyman & Hensel, supra note 85.

¹²⁰ One anonymous reviewer noted that (1) discussion of EEZs entered into the third United Nations Conference on the Law of the Sea (1977–1978; before UNCLOS), and (2) UNCLOS reflects customary law, implying that UNCLOS may not exert the empirical effects we attribute to it. In response, we first note that states did not finalize their consensus on EEZs until UNCLOS. In addition, UNCLOS necessarily reflects customary law. Indeed, UNCLOS codifies that customary international law, which suggests that it represents a behavioral indictor that the community reached a broad consensus. Given this symbiotic relationship, it is extremely difficult (if not impossible) to separate UNCLOS from customary law. For more on this long-standing debate about the empirical effects of treaties, see George W. Downs, David M. Rocke & Peter N. Barsoom, Is the Good News About Compliance Good News About Cooperation?, 50 Int'l Org. 379 (1996). As argued in note 129 supra, though, we find that the effects of UNCLOS on peaceful negotiations of maritime issues strengthen over time, which demonstrates that institutionalization of customary law generates stronger effects on state behavior.

¹²¹ Although we highlight arbitration cases heard through the PCA in Table A1 (online appendix), maritime issues need not be arbitrated through the PCA under UNCLOS. We therefore intend Table A1 to be illustrative, rather than exhaustive. The ICOW dataset—and therefore our analysis—captures all cases of arbitration over maritime claims, including non-PCA arbitration cases (e.g., British arbitration in the Beagle Channel dispute between Argentina and Chile).

¹²² See Mitchell and Hensel, supra note 86. A full investigation into the effects of UNCLOS on compliance with maritime awards lies beyond the scope of this study. To our knowledge, these compliance rates have yet to be gathered for ITLOS and the PCA. We note compliance rates for the ICJ here, however, not only because we have these data, but also to highlight that states should generally expect to comply with the terms of an ICJ judgment if the ICJ issues a ruling in their case. We assume that expectation would carry over to other binding forums and motivate out-of-court bargaining there. Whether that assumption holds requires further empirical investigation.

¹²³ Of the ten states parties that have active Article 287 declarations in favor of Annex VII arbitration (as of December 2020), only two rank order the default procedure behind another forum (i.e., as not their first preference). Six list both the default and (an)other forum(s) as a first-order preference. The final two, Egypt and Slovenia, list a first-order preference for Annex VII arbitration, but no preference for any other forum. See United Nations Convention on the Law of the Sea, supra note 56.

¹²⁴ Mitchell & Powell, supra note 29, find that dyads (i.e., pairs of countries) that accept the compulsory jurisdiction of the ICJ are significantly more likely to experience militarized disputes, consistent with our findings. States in the Western Hemisphere drive these results. European dyads that accept the ICJ's compulsory jurisdiction experience a reduced risk of interstate conflict. This makes intuitive sense, since European states created the precursor to the ICJ, the Permanent Court of International Justice.

¹²⁵ On the spatial and temporal limitations to our study, see Part IV.

¹²⁶ Because other instruments reference the courts we study (e.g., the ICJ), one could argue that the effects we observe derive from one or more instruments other than UNCLOS. Although plausible, such commitments would struggle to explain the constellation of findings we uncover and attribute to UNCLOS here.

¹²⁷ The control variables in our models behave as the studies from which they derive anticipate. Space constraints, as well as a desire to simplify the discussion, preclude us from discussing these results in detail. It is nevertheless worth noting that we obtain no unexpected results. The full statistical models, as well as coefficient plots for the control variables associated with each model, appear in Appendix B (online).

¹²⁸ States may have good reason to manage territorial claims differently than maritime claims—for example, a lack of institutions to address territorial issues, or the ICJ itself, which plays a prominent role in territorial claim management. Future research might investigate these possibilities further. See Owsiak & Mitchell, supra note 111; Hensel & Mitchell, supra note 16.

¹²⁹ To evaluate whether the effects of UNCLOS change over time, Appendix D (online) re-estimates our models for two sub-periods: (1) UNCLOS exists, but is not in force (1982–1993); and (2) UNCLOS has entered into force (1994–2001). We find some evidence that the effects of UNCLOS strengthen over time (e.g., the effect of Article 287 declarations on bilateral negotiations is stronger in the latter, as opposed to former, period; see Table B12, online appendix). Because these periods are short, however, we hesitate to draw strong conclusions from them and prefer the models we present in the main text. In addition, we ran versions of our models that included an interaction between our variables of interest and time—another means through which to account for dynamic temporal effects. We uncover no temporal effects through these models. Given, however, the low number of states that make joint Article 287 declarations and have maritime claims, we might expect such findings when parsing the data more finely, as would be necessary with such an interaction.

¹³⁰ Áslaug Ásgeirsdóttir & Martin C. Steinwand, Distributive Outcomes in Contested Maritime Areas: The Role of Inside Options in Settling Competing Claims, 62 J. Conflict Resol. 1284 (2018).

¹³¹ The design of UNCLOS drives the endpoints of the range. Annex VII serves as the default compulsory procedure. Annex VIII pertains not to delimitation issues, but rather to fisheries, the preservation of natural resources, marine research, navigation, and pollution.

¹³² Because our empirical models control for military capabilities, this finding does not attribute simply to major-state democracies (e.g., the United States, United Kingdom, or France) holding a more global territorial portfolio (i.e., having more maritime boundaries throughout the world). See also Daniels & Mitchell, supra note 76, who obtain a similar result with a more direct control of major state status. For a deeper discussion of democracies and their maritime conflicts, see Mitchell, Sara McLaughlin & Prins, Brandon C., Beyond Territorial Contiguity: Issues at Stake in Democratic Militarized Interstate Disputes, 43 Int'l Stud. Q. 169 (1999)Google Scholar.

¹³³ “Islamic law state” refers to a state whose domestic legal system follows Islamic law—one of the three major legal traditions that have both substantial geographic reach and long-lasting influence. The other two traditions are civil law and common law. See Badr, Gamal Moursi, Islamic Law: Its Relation to Other Legal Systems, 26 Am. J. Comp. L. 187 (1978)CrossRefGoogle Scholar.

¹³⁴ See Mitchell & Powell, supra note 29.

¹³⁵ The negative effect of Annex VII on maritime claims flips from negative to positive in some models, while the effect of joint ICJ commitments is weaker in most peaceful settlement attempt models.

¹³⁶ See Mitchell & Powell, supra note 29; Appel, supra note 31; Hillebrecht, supra note 31; Jo, Hyeran & Simmons, Beth, Can the International Criminal Court Deter Atrocity?, 70 Int'l Org. 443 (2016)CrossRefGoogle Scholar. Rome Statute of the International Criminal Court, July 17, 1998, at https://treaties.un.org/Pages/ShowMTDSGDetails.aspx?src=UNTSONLINE&tabid=2&mtdsg_no=XVIII-10&chapter=18&lang=en#:~:text=In%20accordance%20with%20its%20article,Italy%20until%2017%20October%201998.

¹³⁷ See, e.g., Busch & Reinhardt, supra note 30; Gray & Potter, supra note 30.

¹³⁸ See Mitchell & Powell, supra note 29, at 212–18.

¹³⁹ See Posner & Yoo, supra note 36. The idea is that if a priori forms of compulsory jurisdiction (Article 36(2) declarations) improve the chances for out of court bargaining, then the percentage of cases that the PCIJ/ICJ would hear that come through such forms of compulsory jurisdiction should be smaller because fewer of those cases end up on the court's docket.

¹⁴⁰ See Ginsburg & McAdams, supra note 29.

¹⁴¹ See Mitchell & Powell, supra note 29, at 72.

¹⁴² Mnookin & Kornhauser, supra note 2.

¹⁴³ See Alter's transnational politics model, supra note 24.

¹⁴⁴ Id.