1. Introduction
Single-party autocratic “machines” have been some of the most common autocracies in the post-World War Two period (Geddes et al., 2014), as they seem to possess an institutional makeup which leads them to be resilient to challenges to their authority (Marin, 2015). In these regimes, there is only one legitimate party in politics which has strong control over society, or occasionally multiple minor political parties that are under the thumb of the dominant political party in the autocracy (Davenport, 2007; Wiegand, 2019). These party-based machine regimes can be understood as autocracies where the party has some degree of constraint and influence over the leader regarding policy, possesses functioning local-level organizations, adopts some of the facades of democracy, and controls most access to government jobs and political power (Peceny et al., 2002). Some of the limited restraints in these types of autocracies can include the legislature blocking the implementation of some executive decrees, approving some types of executive-nominated appointments, and initiating some legislation, as well as the failure of the executive in changing constitutional restrictions to their authority, the presence of an independent or semi-independent judiciary, and the ruling party performing some administrative functions independently of the executive (Shulman, 2008). Although potentially merely gossamer in substance, these limited restraints are embraced by these types of autocracies to emulate the optics and mimic the legitimacy found in deep democracies.
However, on the international stage, it is not infrequent for inconsistencies with democratic norms by single-party autocracies to be highlighted, and frequently the behavior of single-party autocracies can place them at risk from coercive foreign policy tools by outside actors, often in the form of punitive economic sanctions. Economic sanctions by outside actors for undesirable policy behavior in areas such as human rights violations by single-party regimes can tarnish the image of these regimes, drawing the legitimacy of their sole exercise of political power into question both at the international and domestic levels. The hypothesis for this article is that single-party autocracies will engage in damage-control efforts like offering to mediate third-party internationalized military conflicts when they are subjected to adverse international actions, such as being economically sanctioned. The article’s analysis succeeds in obtaining statistically significant evidence that single-party autocracies are at a substantially greater likelihood to offer to mediate third-party internationalized conflicts when being subjected to economic sanctions, providing a base for future research on these types of autocracies as well as specific foreign policy instruments and their effects on these regimes.
2. Literature Review
Interstate conflict and war have proven to be pernicious in the post-World War Two era. From 1945 to the end of the new millennium alone, there were 309 international conflicts, these being defined as militarized conflicts involving states which are continuous and organized, or the demonstration of the intention to utilize military force (Bercovitch, 2000). However, there does seem to be evidence for a post-World War Two trend towards peace (Clauset, 2020), due in part to conflict mediation. Of the 309 conflicts, an impressive 190 of them were mediated (with some of them undergoing several mediation attempts), yielding a total of 1990 international mediation cases, not including 204 mediation offers which were rejected (Bercovitch, 2000). Mediation, as defined by the United Nations, is “a process whereby a third party assists two or more parties, with their consent, to prevent, manage, or resolve a conflict by helping them to develop mutually acceptable agreements” (Akpinar, 2015). Another commonly used definition of mediation conceptualizes mediation as “…a reactive process of conflict management whereby parties seek the assistance of, or accept an offer of help from, an individual, group, or organization to change their behavior, settle their conflict, or resolve their problem without resorting to physical force or invoking the authority of law (Bercovitch & Houston, 1996). The mediation of international conflict is a process where a third party works with at least two disputants to attain an acceptable peace agreement between them. As opposed to other conflict resolution methods involving third parties that are judgmental and binding (like arbitration), mediation is both contractual and voluntary where the third party cannot impose a solution on the disputants but can aid them in their quest to obtain a peaceful end to their conflict (Gartner, 2014). In other words, mediation is legally non-binding. This results in the adherence by disputants to any settlements being self-sustaining to last, as well as voluntary (Gartner, 2014). The lack of third-party enforceability found in mediation, as well as the naturally open-ended nature with respect to whether a settlement will be achieved, profoundly differentiates it from law-based dispute adjudication (Beber, 2012). In fact, third-party mediation is the most frequently employed method to resolve violent interstate conflicts (Gartner, 2014), with states being the most common mediator (Melin, 2013). Considering the relative ineffectiveness of conventional ways of addressing conflicts, mediation is critical as a method for their resolution (Akpinar, 2015).
As such, mediation has increasingly become a favored tool for states to utilize to end interstate conflicts. Notably, mediation amounted to greater than 20 percent of all third-party actions occurring between the conclusion of World War II through the beginning of the new millennium (Frazier & Dixon, 2006; Beber, 2012). There is also evidence that interstate mediation produces results, with mediation efforts leading to a peace agreement in 45.5% of interstate conflicts in addition to resulting in more than twice as many ceasefires relative to civil conflicts (Gartner, 2014). Out of all the activities of statecraft practiced historically by human beings, mediation is ancient. States in the past and present have always intervened in the business of other states (Annan, 1999), with mediation being one of the oldest and most frequently found methods for successfully ending international conflicts when applied correctly (Bercovitch & Lee, 2003). The first mediation effort recorded happened in 209 B.C. when Greek city-states aided Macedonia and the Aetolian League in creating a truce during the first Macedonian war, and after the end of the Cold War, the practice of mediating conflicts became much more prevalent than it was prior (Beber, 2012; Crocker et al., 2005; Melin, 2013). The increasing use of mediation to solve violent interstate disputes has resulted in both painful failures, like the inability of the UN Commission for India and Pakistan in 1948 to facilitate peace, as well as impressive successes, like the Camp David Accords that led to peace between Israel and Egypt (Gartner, 2014).
Mediation can produce what may be construed as global public goods. Wiegand (2019) observed that “On a macro level, public goods…include the reduced likelihood of future conflict (Gibler, 2009; Owsiak, 2012), increased likelihood of rivalry terminations (Owsiak & Rider, 2013), …and international legal recognition of borders that reduce uncertainty and, therefore, future conflict (Simmons, 2005; Carter & Goemans, 2011, 2014)” (pg. 5). The successful resolution of conflicts and disputes creates global public goods, where mutual gains amongst concerned states are facilitated (Wiegand, 2019), which can then behoove a potential mediating state in a variety of tangible strategic ways, including increased legitimacy and respect from international actors. For an autocracy, the obtainment and conservation of legitimacy can be a struggle compared to the popular mandate which is bestowed on democracies via their institutional practices. These regime types can be understood as a political system where a small group of people or a single individual wields power with little constraints, competition, or political participation by the general populace (Frieden et al., 2016), and are often known for abuses such as political imprisonment, repression, and torture (Frantz & Kendall-Taylor, 2014; Akpinar, 2015), which can lead to reputational costs to autocracies. Research indicates that states and their leaders can and do alter their reputations with both their contemporary words and deeds (Lupton, 2018a; Lupton, 2018b; Lupton 2020). By mediating a conflict, autocracies can work to offset potential damage to their reputation which could result from international scrutiny, sanctioning, and condemnation, and instead foster a reputation as a peacemaker (Melin, 2013). Burnell (2006) notes that “In fact, a growing acknowledgement of the international dimensions…suggests that external judgements of [autocratic] regime legitimacy may now carry more weight than perhaps at any previous time…” (pg. 552), while mediation in conflict serves to enhance such legitimacy for an autocracy (Akpinar, 2015) and potentially offset the aforementioned negative effects to their legitimacy and reputation caused by actions of outside actors such as economic sanctions.
Research also indicates that single-party autocracies are more like democracies as opposed to other types of autocratic regimes (Weeks, 2014), including their sensitivity to domestic and international perceptions of regime legitimacy. Differentiating themselves from other types of autocracies like personalist dictatorships, military juntas, or monarchies, party-based autocracies have a hierarchical structure where political elites aren’t personally bound to the leader of the autocracy (enabling them to be able to act to remove the leader from power without fear of losing their job or life), and where regime insiders and leaders rise up through the ranks based on seniority and merit as opposed to family or otherwise personal relationships with the autocratic leader (Conrad et al., 2014; Weeks, 2008, 2014). Hypothetically, this leads the leaders of single-party autocracies to get rid of policies that don’t satisfy the interest of the party (Shulman, 2008). In this type of autocratic regime, domestic institutions aren’t necessarily “rubber-stamp” organizations, the party regularly holds intraparty elections competitively for certain positions, as well as regime insiders and factions potentially coalescing around specific policy issues and competition for important jobs (Weeks, 2008; Davenport, 2007). In addition, because they attempt to parrot democracies, this can cause party-based autocracies to be vulnerable to pressure to live up to their supposedly democratic image. Kendall-Taylor and Frantz (2014) observed that unlike other autocratic regimes, autocracies which utilize pseudo-democratic institutions “…often still seek to portray themselves as ‘democrats,’ making them likely to be more vulnerable to international and domestic criticism that highlights the inconsistency between their behaviors and democratic norms” (pg. 81). The pressure to escape the highlighting of inconsistencies with democratic norms may result in party-based autocracies offering most often to mediate international conflicts when compared to other types of autocracies. Indeed, in Autocracies as Mediators in Conflicts (Honig, 2022), evidence indicated that single-party autocracies are statistically more likely to offer to mediate third-party internationalized military conflicts when subjected to economic sanctions than other types of autocracies (military juntas, personalist dictatorships, and monarchies), which supports the idea of sensitivity to questions of their image and legitimacy by actors outside of the regime. To investigate the degree of difference between the likelihood of mediation offers by single-party autocracies both before and after sanctioning, this article utilized the following quantitative research analysis.
3. Methodology
This article hypothesizes that single-party autocratic regimes will be substantially more likely (defined as >50%) to offer to mediate a third-party internationalized military conflict when being subjected to economic sanctions. This figure was chosen as in many areas of research, 50% is seen as a decisive tipping point in likelihood. For example, when something has a 50% greater likelihood of happening, it’s often interpreted as a fundamental change in expectations, making the shift significant for clear analysis and discussion. A greater than 50% likelihood marks a critical point where there’s a clear shift from the event being equally likely to happen or not, to one side becoming reliably more probable. As Hohle and Teigen (2017) observed, “…probability distributions have a known lower and upper bound representing impossibility and certainty (0% and 100%), and a 50/50 midpoint below which chances are often described as low (less than even) and above which they are high (more than even)” (pg. 26). To test this hypothesis, it used Probit regression as the primary testing method. The use of a Probit regression model was due to the data being time-series cross-sectional with binary dependent and independent variables. Further, the analysis also used OLS linear probability models as a robustness check on direction and significance as well as likelihood of offers to mediate, due to its unbiasedness and its distinction of having the least variance among all linear unbiased estimators, its linear properties, its asymptotic unbiasedness, and its status as a consistent estimator. Most importantly, the analysis utilized predicted probabilities to determine the difference of likelihood of offers of conflict mediation by single-party regimes pre- and post-sanctioning while conducting an F-Test to determine statistical significance, as predicted probabilities show real likelihood and substantive effects above and beyond what Probit and OLS on their own would provide. So as to obtain the correct estimates and their confidence intervals, the analysis simulated the Probit model using the “Clarify” program, which King et al. (2000) note “…uses Monte Carlo simulation to convert the raw output of statistical procedures into results that are of direct interest to researchers, without changing statistical assumptions or requiring new statistical models” (pg. 3). To avoid problems with heteroscedasticity of observation, the analysis also tested for robust standard errors.
For the independent variable, the analysis used instances of economic sanctions (Sanctions) leveled at an autocracy, with the dependent variable being offers of conflict mediation (Mediation Offer) by an autocracy. Burnell (2006) noted that for autocracies “…international legal recognition and support, whether material and/or symbolic—that is to say external legitimation—are very valuable to the manufacture of legitimacy…[and] a regime’s main claim to legitimacy can shift over time, adjusting to…changes in the international environment” (pg. 549). If sanctions were levied and/or threatened during any calendar year in the dataset on a mediating state or potential mediating state, this independent variable was coded as 1 (and 0 if not). If mediation was offered for an international conflict during any calendar year in the dataset, the dependent variable is coded as 1 (and 0 if not). The data for sanctions were derived from the Threat and Imposition of Sanctions (TIES) Dataset 4.0 (Morgan et al., 2013), which covers sanctioning cases from 1945 - 2005, which was then merged with the Bercovitch International Conflict Dataset. The unit of analysis for this hypothesis was state year. The analysis merged the Bercovitch (2000) dataset with the Geddes, Wright, and Frantz (2014) dataset covering authoritarian regime types from 1946-2010. This dataset divides autocracies into four different types of regimes: monarchies, personalist dictators, party-based autocracies, and military juntas. The basis for dividing these groups up is based primarily on the rules that delineate leadership groups, as well as how these groups represent specific interests in the making of decisions (Geddes et al., 2014). Part of the reasoning for choosing the Geddes et al. (2014) dataset was that the specific autocratic regime-types that they identify can be tied directly to specific levels of regime flexibility, transparency, and accountability, grounded in how these regimes measure on scores of these concepts and on existing academic work on the subject (Mattes & Rodrigues, 2014). Via the same line of reasoning, the analysis derived information for states with less than a million inhabitants which were omitted from the Geddes et al. (2014) dataset from Magaloni, Chu, and Min’s (2013) dataset on autocracies of the world from 1950 through 2012, as well as via the CIA World Factbook’s (2021) database. To interpret the results, the analysis tested for joint significance. Replicating the analysis found in Autocracies as Mediators in Conflicts (Honig, 2022), the analysis coded single-party autocracies as a 1 (Party), with all other types of autocratic regime-types (military juntas, personalist dictatorships, and monarchies) combined coded as a 0. In addition, the analysis interacted the independent variable (Sanctions) with the variable for single-party autocracies (Party). This was done to attempt not only to replicate the findings found in Autocracies as Mediators in Conflicts (Honig, 2022), but more importantly to produce the necessary data to analyze the predicted probabilities required to determine the difference of likelihood for offers of mediation pre- and post-sanctioning.
The analysis also included several control variables which may affect a state’s decision to offer mediation, which were fixed at their modes. These included Same Polity, which compares the polity of the mediator with the polities of the parties of the conflict. This variable was operationalized by categorizing different types of state regimes based on differences in government structures and sources of political decision-making power, the information on which was derived from the Bercovitch dataset. These categories were democracies, single-party states, authoritarian regimes, monarchies, and military regimes. This variable is coded 0 if the mediator’s polity is different than the belligerent states’ polities, 1 if it is the same as one state, and 2 if it is the same as both states. Potential mediators might be more likely to offer their services if they share a government type with one or more of the conflicting parties due to common political bonds and beliefs. Another control variable included was Same Region, which indicates if the mediator shares the same geographic region as one or more of the conflicting parties (these regions being North America, Central and South America, Africa, Southwest Asia, East Asia/the Pacific, Middle East, and Europe). This variable was coded as 0 if the mediator doesn’t share the same region as either belligerent party, 1 if it shares its region with one party, and 2 if it shares its region with both parties. States that share their geographic region with one or more of the conflicting parties may be more motivated to attempt to mediate. A further control variable included was UN Security Council Member, which indicated if a mediator is a permanent member of the UN Security Council (UNSC). The analysis included this variable since powerful states are thought to act as mediators more frequently and with greater success (Crescenzi, Kadera, & Mitchell, 2011). This variable was coded as 1 if the mediator is a permanent member of the UNSC, and 0 if not. These control variables were all derived from the Bercovitch dataset (2000). The last control variable was Previous Relationship, which is based on previous alliances and/or economic relationships between the conflicting parties and the potential mediator. This control variable was derived from Gibler’s (2009) International Military Alliances dataset and Kohl et al.’s (2013) Trade Agreement Dataset. This variable was coded as 0 if there was no previous relationship between the mediator and either belligerent party, 1 if the mediator has a previous relationship with one of the conflict parties, and 2 if the mediator has a previous relationship with both conflict parties.
4. Results
The analysis succeeded in replicating the previous findings found in Autocracies as Mediators in Conflicts (Honig, 2022), as well as providing evidence supporting its current hypothesis. Sanctions interacted with Party was positively correlated and strongly statistically significant for both OLS and Probit. After conducting a joint significance test, significance was indeed found. These results reconfirm the findings contained in Autocracies as Mediators in Conflicts (Honig, 2022), that Party-based autocracies are the most likely to offer to mediate international conflicts when being confronted with sanctions. These autocracies are the most affected by sanctions statistically, which has a negative effect on all other regimes combined. Most importantly, by using predicted probabilities in conjunction with the Probit analysis, single-party autocracies were found to be at a respectable 82% greater likelihood to offer to mediate third-party internationalized military conflicts when they are confronted with sanctions. This finding was statistically significant at the 95% confidence level. Interestingly, by utilizing predicted probabilities with the OLS analysis, autocracies were found to be at a staggering 1403% greater likelihood to offer to mediate third-party internationalized military conflicts when being economically sanctioned, which was statistically significant at the 99% confidence level. The large discrepancy between the results of the Probit model and the OLS model might be because Probit models are specifically designed to handle the nature of binary outcomes more efficiently, whereas OLS might lead to inefficient or biased estimates when the dependent variable is binary (such as in this analysis). Nevertheless, both results lend support for the article’s hypothesis, that single-party autocratic regimes will be substantially more likely (defined as >50%) to offer to mediate third-party internationalized conflicts when being subjected to sanctions. The control variables behaved largely as expected. A previous relationship between a potential mediator and one belligerent, a previous relationship between a potential mediator and both belligerents, UNSC membership, and a shared geographic region between a potential mediator and both belligerents were all positively correlated and strongly statistically significant for both OLS and Probit.
5. Discussion
The results of this study reveal a significant insight into the behavior of single-party autocracies in the international arena: these regimes seem acutely sensitive to the risks of a negative international reputation and engage in damage-control efforts, such as playing the part of peacemaker, to mitigate these risks when necessary (such as when being singled out for sanctioning). Unlike their multi-party counterparts, which may maintain a degree of political flexibility and resilience through internal political competition, single-party autocracies face unique vulnerabilities when it comes to their global standing. These regimes rely heavily on international legitimacy and recognition, particularly in an era where global governance structures, media attention, and foreign investment are all deeply intertwined with the perception of a regime’s governance. For these states, external perception is not merely a matter of diplomatic or economic concern—it is hand in glove with the political survival of the ruling party. However, it’s important to consider limitations and counterarguments to this finding. While the study finds that these regimes seem sensitive to the tarnishing of their reputation and legitimacy, this sensitivity may not be uniform across all types of international interactions. For instance, regimes that have access to key economic or military alliances may find ways to mitigate a tarnished international reputation. In these cases, external pressure might be less effective in shaping the behavior of the regime than the internal power structures or the ability of the leadership to control the narrative domestically. Moreover, the focus of this study on single-party autocracies excludes other forms of autocracies—such as military juntas or monarchies—which may exhibit a different set of vulnerabilities or strategies in response to international reputation concerns. Further research could explore these variations, comparing how different autocratic structures manage external perceptions and how these reputational concerns shape foreign policy decisions, including playing or seeming to seek to play a mediation role in outside conflicts.
6. Conclusion
The degree of the lack of research on single-party autocratic regimes and their complex role in mediation raises important questions about research into both comparative regime-types as well as militarized conflicts and mediation, although this could be a function of a lack of updated data, which could increase the volume of research on these topics. Rectifying this issue, although slow and steady work, would be worthwhile in rounding out the study of the responses of autocracies to punitive measures internationally, as well as the comparative differences in autocratic regime-types. To examine the normative implications of this article’s findings, it is important to consider whether sanctions might simply encourage mediation offers by single-party autocracies that are more symbolic than substantive and only meant to burnish a state’s reputation in times of need, or whether they indeed do lead to accepted offers and/or successful conflict resolution outcomes. Although this analysis did seek to shed light on these questions, efforts to quantitatively determine whether sanctions were correlated with greater acceptance of genuine mediation offers by single-party autocracies and/or successful outcomes to mediation did not achieve statistical significance, so at present it is unknown whether the evidence suggests that sanctions are merely symbolic to improve reputation, or if they are correlated with genuine or successful conflict resolution efforts. Clearly there is much to learn here and more research to be done, but such evidence as there is on the sensitivity which single-party based autocracies have to tangible forms of outside pressure is a boon not only for academia (which all too often seems to neglect investigating the comparative differences in autocratic regimes and their behavior), but also offers a potentially valuable tool for affecting the behavior of many repressive states around the world while they are engaged in perceived negative policy behavior.