International Organizations: Enablers or Impediments for Authoritarian International Law?

International organizations (IOs) provide space for the exchange of ideas. Particularly since the Cold War ended, many expected that this exchange would inevitably lead to more democratization and liberalization around the globe. Instead, some of the largest non-democratic actors on the global stage have functioned within these organizations for decades without liberalizing, while others joined as full or newly transitioned democracies just to see those qualities slowly erode. As Tom Ginsburg's recent article concludes, today's autocrats might instead use international law—including the legal apparatus of IOs—to further their own authoritarian agendas. This essay engages with Ginsburg's thoughtful piece by suggesting that IOs both enable and resist the emergence of “Authoritarian International Law” (AIL). Creating or joining IOs is a costly but attractive strategy for revisionist states since members equally influence IO evolution. Fortunately for democracy's advocates, IOs are usually status quo entities, and liberalism is deeply embedded in many existing today. Cross-temporal observations of changes in IO membership, members’ regime types, and IO features beyond the founding documents are needed to fully understand how organizations simultaneously perform these paradoxical functions.

have the luxury of designing an organization according to their combined preferences to meet certain rational ends. 6 When states join existing IOs, however, they do not immediately have the same creative control over the organization. Joining is costly, often requiring some domestic reforms to align with the group. Current members determine the accession conditions, usually placing an onus on "fit." 7 States might accept these costs in pursuit of domestic political or economic benefits. For example, liberal theories suggest that "weak democracies" join international organizations in order to tie the hands of future elites, thus ensuring democratic governance continues. 8 States also might accept these external conditions in order to gain international legitimacy, signaling that the state meets the requisite conditions of membership. 9 Post-accession, though, all member states, new and old, belong to the "in group," meaning they are all able to influence the organization's trajectory, at least when acting with a majority of other members.
While some IOs are actors in their own right, they function, at minimum, as environments for regular interaction among member state representatives. As a consequence, IOs allow for the exchange of ideas beyond formal economic, security, or other functional purposes. Because many IOs are not exclusively comprised of democratic states, leaders are exposed not only to liberal ideas within these organizations, but to authoritarian ones as well. Today's new autocrats learn effective means of sheltering their authoritarian stripes from international criticism by participating in an IO setting. IO membership is thus a "permissive condition" for the emergence of AIL. 10 We have witnessed this over the past decade in the EU. Established in 1951 as the European Coal and Steel Community, the EU has expanded drastically in scope and membership from the original six members to twentyeight currently. Hungary did not join the EU until 2004, after emerging as a "success story" among the former Soviet bloc states during the democratic transitions. 11 It obviously did not participate in the early discussions that shaped the Union, especially those that formalized the conditions of membership in the early 1990s. Hungary was a "rule-taker," adjusting its domestic political and economic system to earn membership. Since 2010, however, Hungarian Prime Minister Viktor Orbán's Fidesz party has dismantled the country's democratic system beyond recognition-an authoritarian lion among liberal sheep. 12 How has Hungary avoided political sanction for so long? The answer is multifaceted, but a few aspects of it are relevant for AIL. First, Hungary uses the EU's inefficiency to its advantage. 13 For example, the General Affairs Council of the EU was poised to discuss the European Parliament's 2018 resolution that triggered the suspension clause (Article 7(1) Treaty on European Union) on September 19, 2019, when-merely one week prior-the Hungarian government circulated a 158-page rebuttal. 14 The delegation then relied heavily upon this updated information during the meeting, curtailing serious debate since others had no time for proper vetting or review. 15 Second, Fidesz has used its membership in the European People's Party, the largest political group in the European Parliament, to evade other sanctions. 16 A third reliable strategy Hungary has used is likening its constitutional reforms to laws and practices of other member states. This practice of stitching together a "Frankenstate" 17 has allowed Orbán to deflect criticism from other EU members, most recently Finland. 18 As the EU continued to flounder in its response to the Hungarian government, Orbán's playbook was taken up by leaders of the ruling Polish Law and Justice (PiS) party in 2015. Having observed the EU's in-(or at least insufficient) action, the PiS government began systematically dismantling its own democratic institutions-most notoriously the judiciary. Clearly, IO membership simplifies sharing authoritarian policies-and, as a phenomenon, is not exclusive to less democratically dense regions.
IOs thus enable AIL by exposing recalcitrant member states to new ideas and anti-democratic strategies, and by setting precedents of inaction when even one member models effective anti-democratic strategies for other wouldbe autocrats in the group. If organized, autocrats may therefore attempt to revise those very institutions they pined after. 19 As I will now discuss, though, the permissive condition of IO membership does not necessarily lead to the autocratization of international law and organizations. IOs impose barriers to revisionist states for the same reason critics are frustrated by the EU's inaction in Hungary or Poland: institutional change is slow.

IO Stickiness as an Impediment to AIL
IOs usually evolve gradually-progressively or regressively-earning the label of "sticky." 20 IOs also tend to last for a long time 21 -and while more IOs exist in the world today than ever before, IO creation has dropped to early Cold War levels ( Figure 1). Fortunately, 70 percent of states today are already members of at least one regional IO that commits its members to democracy, so new authoritarians have work to do if they want to steer one in a different, illiberal direction. 22 This empirical reality suggests that we must study organizations beyond their time of creation to understand if and how they foster AIL. However, generating such cross-regional, time-series data requires deep substantive knowledge of each institution, not to mention a lot of time. A second-best approach-and a way to see if this enterprise is worthwhile-is to compare IOs created during different time periods.
To do so, we must first establish the appropriate cut point. If the central question is what today's "new authoritarians" are trying to accomplish with international law, there are many reasons to think their design incentives were different during and after the Cold War. 23 Ginsburg starts off his essay by convincingly arguing that the present-day authoritarian approach is distinct from that of the past, at least because autocratic states are already "integrated into the global capitalist economy"; ideology is no longer their main motivator; and domestic institutions in autocracies, namely elections and courts, parody those of a democratic state. 24 Along with these new characteristics of autocrats, the modal internal threat to authoritarian (and weak democratic) leaders has shifted since the 1990s from military or executive coups to more nuanced tactics. 25 Also, and perhaps most obviously, states have only been able to act and interact relatively independently of either American of Soviet influence since the Cold War ended. For all of these reasons, we should only expect to observe new autocrats with AIL tendencies in the post-Cold War period. Next, we need to establish proper conceptualizations for the IO's average member state regime type and its features, respectively the independent and dependent variables, both of which can be conceptualized as a democratic-autocratic dichotomy. 26 Categorizing individual states as democratic or autocratic is straightforward if one adopts a widely accepted quantitative measure, such as from the Polity IV dataset. 27 Averaging the member states' regime scores provides the "democratic density" of the IO. 28 Categorizing IOs as either authoritarian or democratic is less straightforward-and here Ginsburg's innovative conceptualization of AIL is very helpful. He suggests AIL has several distinct characteristics in substance and form that separate it from pro-democracy (or general) international law. Authoritarian treaties emphasize sovereignty and stability versus democracy, freedom, and human rights, using international law to promote autocracy rather than democracy, all while hedging primarily against internal security threats. 29 Thus, an authoritarian IO's documents will reflect "legal rhetoric, practices, and rules specifically designed to extend the survival and reach of authoritarian rule across space and time," while a democratic IO's laws "protect and extend the sphere of democratic governance." 30 Using these definitions, he codes seventysix organizations based on whether the founding documents contain democratic features (mention democracy, the rule of law (ROL), human rights, or international law), authoritarian features (security), and third-party dispute resolution mechanisms (courts or dispute resolution).
When we separate the IOs into Cold War and post-Cold War periods, new patterns in treaty features emerge. The first observation we can make is that most of the post-Cold War IOs are either comprised of democratic or "hybrid" states (thirty-six of forty-five), based on the democratic density of member states. 31 Given the changes in dictatorial tactics in this era, many new authoritarian states fall into this hybrid category. Perhaps predictably, post-Cold War IOs also reference democratic features at higher rates than the Cold War IOs (Columns a, b). References to democracy, human rights, and the ROL nearly double even among democratic IOs (Columns c, d). Hybrid organizations are even more likely to reference democratic features in their founding treaties than democratic IOs of the same era; in fact, post-Cold War hybrid IOs almost universally reference human rights while barely half of the democratic IOs do (Columns h, d). This trend signals an entrenchment of these liberal values in formal institutions-which, as has been discussed, are then difficult to walk back.
The fact that hybrid organizations adopt democratic language at even higher rates than democracies could signal a few different mechanisms. It is possible these IOs' founders are disingenuously mimicking democratic IOs, using the same tactics Ginsburg observes at the domestic level: taking on the veneer of rule-of-law-abiding states that collectively value democracy and respect human rights. Alternatively, it might suggest aspirational international lawmaking, in line with the aforementioned assumption that more integration catalyzes liberalization and democratization. Only by evaluating changes to the formal IO treaties, to IO membership, and in member states' regime types across time can we ascertain whether international law and organizations are or are not being used to further authoritarian ends. 26 In general, I conceptualize these as a spectrum, but will adopt the simplification used by Ginsburg for better engagement with his argument. 27 As is common, Ginsburg (supra note 2, 233) treats a score of +6 or higher as a "democracy" (on a 21-point scale of -10 to +10, where negative scores are more autocratic and positive scores more democratic). 28  rhetorical emphasis of sovereignty and respect for self-determination as a guise for anti-democratic activity, but I leave that aside here. 31 Pevehouse & Russett, supra note 28.