How International Actors Help Enforce Domestic Deals

International actors at times seek to help bring peace, democracy, and human rights. Studies of how international actors help enforce political bargains between incumbent governments and their domestic opponents are proliferating. They show that opposition groups have trouble trusting incumbents to adhere to the political bargains they strike because incumbents can use their familiarity with state institutions and can use their asymmetric hold on power during bargain implementation to violate terms by retaining more of the status quo than agreed. International actors can overcome these “reversion problems,” however, by using monitoring mechanisms (often focused on electoral campaigns) and incentives conditioned on compliance. Reversion problems, and enforcement by international actors as a solution, are common across issue areas—arising when domestic actors try to end civil conflict, open elections, and reduce repression—but the literatures in these issue areas have largely remained segregated. This review proposes advancing this research agenda by unifying them and (re)examining the conditions under which this solution works best.


INTRODUCTION
International actors are often involved in the domestic politics of peace, democracy, and human rights. Intrastate peacekeeping missions seeking to secure states enmeshed in civil conflict, for example, rose dramatically after the Cold War; the United Nations deployed more missions in the first five years of the 1990s than in the previous 40 (Fortna & Howard 2008). International involvement has become common beyond conflict-ridden countries. For instance, the prevalence of election observation, monitoring the freedom and fairness of power transitions, has expanded from less than 10% of all elections in the 1960s to almost 80% in the 2000s (Hyde 2011b, Hyde & Marinov 2012. Democracy assistance, a related form of engagement, continued to increase after the Cold War, growing from 2% of foreign aid in 1990 to almost 20% in 2005 (Finkel et al. 2007). While the nature of international involvement in domestic politics has changed in recent years, it remains a central form of interaction between states (Girod 2019).
The intersection of international and domestic politics has been the subject of scholarly interest. A research agenda has developed on institutions that intrude on sovereignty, such as international organizations and treaties, and shape how states interact (e.g., Simmons 1998Simmons , 2010. Another body of work has developed on domestic mechanisms that states use to influence international relations (e.g., Bueno de Mesquita & Smith 2012).
However, a similarly cohesive research agenda has not developed on international actors' increasing assistance in enforcing political bargains between incumbent governments and their domestic opponents. This crucial role of international actors in the domestic politics of peace, democracy, and human rights protection is the focus of this review.
This review first articulates a serious problem in domestic politics-how to get an incumbent to credibly commit to a bargain with opposition groups in the context of weak institutions-and shows that this same problem proves costly across issue areas. Studies of a variety of crucial issue areas-ending civil conflict, opening elections, and reducing repression-describe similar asymmetric commitment problems when incumbents try to strike these formal or informal bargains with their domestic opponents about political issues. The crux of these problems is that incumbents cannot credibly commit to refrain from using their information about, and control over, state institutions during implementation to revert to something close to the status quo. None of the existing work identifies these reversion problems as a category of problems stretching across different issue areas, despite studies in each area examining versions of them. In identifying reversion problems in political bargains, this review stakes out these political problems as a part of the field as others have done for economic problems (e.g., Vreeland 2003).
Next, this review identifies a solution that is also common across the issue areas of ending civil conflict, opening elections, and reducing repression. Reversion problems occur under weak political institutions, so domestic solutions are often absent. Instead, international actors, ranging from foreign states to intergovernmental organizations, help enforce compliance by detecting and punishing violations in many cases. These international enforcers monitor compliance with domestic bargains, often using the period around electoral campaigns, and offer incentives conditioned on that compliance. These conditional incentives typically include foreign aid and beneficial trade terms.
Finally, this review proposes advancing this research agenda by unifying it. I show how studies in disparate issue areas may gain traction by examining questions such as when international enforcement is likely to effectively overcome reversion problems from these different perspectives. Two conditions that account for variation in the use of enforcement mechanisms are the Goldilocks condition (wherein international actors are invested but not too invested) and state desperation. There are also unexplored directions of great importance in the existing literature across these issue areas, including the question of what happens in the longer term. Exploring this unified theoretical framework across issue areas may help us to better understand reversion problems and the possibility for solving them through international enforcement.
This review identifies a research agenda that draws together domestic and international politics to elucidate how peace, democracy, and the protection of human rights can advance in states with weak political institutions. It captures how problems across issue areas that have been examined separately until now are actually underpinned by a common logic-and, in drawing the studies together, it shows how a more sophisticated understanding of international actors' assistance in domestic problems is developing and could develop further. Moreover, addressing these problems is significant for the populations living under these weak institutions, and, because the problems are prone to spread, for policy makers outside these states. Understanding the theoretical underpinnings of these problems and solutions therefore also has important practical applications. This emerging work offers a broader conception of international intervention that suggests an expanded role of global governance.

REVERSION PROBLEMS
Leaders and opposition groups 1 often have incentives to strike a bargain on policy or power distribution in the state. The bargain provides more benefits to each actor than their alternative options, which often entail confrontation (Fearon 1995). These political bargains can be formal or informal political pacts, new political or electoral rules, or the redesign of laws and governing institutions that fundamentally concern how factions contest office, hold office, and use state forces against those out of office. The formality of these bargains is strategically selected and can affect implementation.
Particular problems, however, tend to plague the implementation of these bargains. Especially in the context of weak institutionalization, a distinctive type of credible commitment problems arises: what I call reversion problems, wherein the incumbent specifically can revert to a policy or balance of power similar to the pre-bargain distribution as the incumbent gains a window of opportunity between the bargain and its institutionalization. Credible commitment problems generally occur when a bargain would provide mutual benefits, but, after the actors strike the bargain, power shifts, and then at least one actor is able to violate the bargain to seize even more power (e.g., Fearon 1995Fearon , 1998Powell 2004Powell , 2006Walter 1997Walter , 2002). An especially problematic power shift tends to occur during implementation of a bargain: Changing policies, rules, or larger institutions to provide the de jure match to the de facto challenger demand is often not simultaneous or perfectly precise, and, during this implementation process, one actor typically becomes weaker, incentivizing the stronger actor to renege on the bargain.
Incumbents specifically are more capable of taking advantage of implementation periods than their opponents, reverting during this opportune moment to something closer to the status quo than the bargain that was struck. 2 Incumbents can work the levers of power to renege on changes to the policy or power distribution to which they had agreed. They typically know the state institutions better and have access to them during implementation. Incumbents are typically supposed to change institutions under a bargain to benefit opposition groups, and they are also typically better positioned to resist these changes, including in ways that may not be recognized by other actors. Consider, for example, a context after a civil conflict settlement: Among the more detectable actions, governments may establish new ministries that usurp some of the powers that had been promised to the rebels; 3 among the less detectable actions, governments may drag their feet on registering voters in rebel strongholds when both factions are set to compete in postconflict elections. 4 Incumbents can make excuses for such behavior if caught, too, claiming for instance that they did not have the institutional capacity to produce these changes. Incumbents cannot simply commit to not take advantage of their position, even if their commitment is "genuine," because shifting incentives produce these reversion problems (Simmons & Danner 2010, p. 232). For these reasons, opposition groups rather than incumbents are especially vulnerable to violations under state structures. Reversion problems are therefore asymmetric.

Reversion Problems Across Political Issue Areas
To explicate reversion problems, I have selected three issue areas that define fundamental political interactions, in which bargains are likely to be desirable at times. I sought cases where organized groups contest power illegally and legally, which serve as canonical cases, and then added a case where incumbents explicitly use state forces against their opponents, which potentially serves as a harder case. Taken together, these provide diverse cases, which are useful to "illuminate the full range of variation" (Seawright & Gerring 2008, p. 297). In the three issue areas of ending civil conflict, opening elections, and reducing repression, I focused on quantitative studies, which comprise much of what is recently published in these issue areas, and the cases that emerge from them. I also reviewed the top political science journals over the past ten years to further test if the dynamics held in more contexts of these issue areas. 5 Ending civil conflicts. Credible commitment problems, including reversion problems on the part of the incumbent government, specifically, are often seen as a disarmament issue following settlements to end civil conflict (Walter 1997(Walter , 2002. Combatants in many cases can identify a mutually beneficial bargain in which they stop fighting and instead (for instance) share power, because conflict is costly, destroying resources and otherwise disrupting daily life (e.g., Powell 2006). Existing studies show that settlements typically reassert the state's monopoly on force, and so the rebels must demobilize and disarm as a bargain is implemented, even if they will later be integrated into the state's structures, including security forces (e.g., Fearon & Laitin 2007). In these cases, even if the rebels will eventually have some control of state forces, the implementation period still tends to produce moments during which the incumbent can attack at a power advantage, perhaps killing rebel supporters or seizing rebel territory-acts that further shape the power distribution (also see , Lake 2003, Mattes & Savun 2009). Fear of these reversion problems can cause rebels to be reluctant to sign a settlement (protracting the conflict) or insecure about a settlement that they have signed (potentially reigniting the conflict), making settlements relatively rare and prone to failure (e.g., Toft 2009).
For example, in Zimbabwe, the Patriotic Front troops, led by Robert Mugabe against the Rhodesian government, hesitated to sign the 1979 Lancaster House Agreement and bring an end to the civil conflict. The rebels were concerned that Rhodesian forces, supported by South African forces, would attack the assembly points where they were set to gather their troops during the implementation period. Mugabe noted, in an interview and a report cited by Walter (1997, p. 351), that "it would be ridiculous for the settlers who are murdering the Zimbabweans to be intrusted with [our] security during the crucial transitional period." Similar reversion problems persist beyond initial stages of disarmament and demobilization, though. Settlements to end civil conflicts also typically reform institutions that distribute political power among factions in the state, beyond the security forces (e.g., Hoddie & Hartzell 2007). During the implementation of these political changes, the balance of power also tends to shift, producing moments of relative weakness for rebels who do not yet have institutionalized control and providing incentives for incumbents to renege (Matanock 2017a,b). Under these circumstances, too, the incumbent government can revert to something closer to the status quo in ways that can be difficult to detect and counter. 6 El Salvador's civil conflict serves as an example of concerns other than attacks that occur during implementation of a settlement. In 1992, the leftist Farabundo Martí National Liberation Front (FMLN), guerrilla rebels, and the Salvadorian government finalized a settlement. Even after initial demobilization, however, credible commitment problems persisted. The rebels had entered the settlement with the understanding that the FMLN would become a political party, and they accepted a system that maintained bias toward government parties (Matanock 2017b;Wood 2000, p. 87). But the incumbent then subverted aspects of the bargain, further reducing FMLN power through the number of seats won. For instance, the government had agreed to expedite registration cards for eligible voters, including many rebel supporters, but it then failed to meet the timetable (Howard 2008, p. 123;Human Rights Watch 1994, p. 4;LeoGrande 1998, p. 108). Moreover, as elections approached, the government sought to move polling stations in FMLN stronghold municipalities to a department capital, alleging a security threat; the move added travel time and reduced turnout mainly among rebel supporters (Montgomery 1998, p. 131). These are reversion problems affecting political issues, and they extend beyond postconflict settings. 1992, p. 31; also see Savun & Tirone 2011). In fact, reversion problems can challenge the integrity of each election, not just transitional elections. Even if all parties agree to abide by principles of open competition when establishing or revising rules for elections, the incumbent has incentives to use electoral manipulation and even violence during the contests to gain an advantage. The sitting government, which controls the structures of the state, has incentives to commit electoral fraud and repress opponents in weakly institutionalized settings to seize more power than allocated. The informational problems surrounding electoral manipulation, as well as how much each side will tolerate and what will affect the outcomes, exacerbate these credible commitment problems (Beaulieu 2014, Kelley 2009). Even operating under full information, however, the incumbent faces incentives to commit fraud (see especially Beaulieu 2014). Supporting this idea, recent research has found that misconduct is more likely when power is firmly solidified for the incumbent (Donno 2013, Hyde 2011b, Simpser 2013. 7 Reversion problems can also affect bargains struck for protection of minority groups or individuals that extend beyond voting rights (Keohane et al. 2009). Incumbents can promise freedom of the press and the like, but renege on such promises, as well.
The Dominican Republic provides a striking example of the reversion problems that can occur in the context of elections. The incumbent, President Joaquín Balaguer, carefully designed and staffed the body responsible for overseeing and handling any complaints for the 1990 election. Despite a double-digit lead in some pre-election polls, the challenger lost in a process allegedly mired in substantial fraud (Donno 2013, pp. 128-29). Observers treaded lightly in criticizing that election, but the incumbent was on notice. In 1994, the incumbent won again, but this time with more widely validated allegations of substantial manipulation (Donno 2013, pp. 129-30). The incumbent president had undermined the pact of the electoral process twice, including his more overt reversion efforts in 1994, and, at that point, a formalized negotiation produced the Pact for Democracy, wherein a new election was set to be held the following year without the incumbent running. The incumbent managed to cling to power a little longer by postponing that election (Donno 2013, p. 131).
Reducing repression. Reversion problems also prove troublesome when leaders strike bargains with opposition groups that reduce repression and other government abuse with benefits to both sides. Governments often find repression effective in the short term (e.g., Bueno de Mesquita et al. 2003, Valentino et al. 2004), but they face resistance not only in the form of routine noncompliance but also as violent protest and rebellion in the longer term (Carey 2006, Lichbach 1987, Moore 2000, Pierskalla 2010, Ritter & Conrad 2016, Shellman 2006; also see Davenport 2007 for a review of the different directions in which the evidence points on repression and dissent), perhaps especially if average citizens experience grievances during repression (Cederman et al. 2010, Gurr 1970, McAdam 1999. Under particular institutional and structural conditions, leaders therefore often have incentives to avoid these costs through bargains with opposition groups that limit the use of repression (Cingranelli et al. 2013, Conrad 2011, Conrad & Ritter 2019, DeMeritt & Young 2013. These situations, typically described as dynamics of repression and dissent, often entailing protest or even rebellion, are inherently processes of bargaining over behavior, even if informal (e.g., Conrad & Ritter 2019). Democratic institutions are associated with lessened repression (Davenport 2007). Even outside democratic institutions, however, organized opposition parties-particularly where the judiciary is weak-can extract concessions from the government to limit its repression, although these are often viewed as small concessions and not necessarily costly to the government in the short term (a point made by Conrad 2014 andVreeland 2008). In fact, the type of opposition shapes the specifics of the concessions (Conrad 2011). Once governments agree to limit repression, however, they face incentives to revert, especially as opponents are not as highly mobilized and not paying as much attention as before. Organized groups in society, including political parties and minority organizations, potentially pay more attention longer, 8 but they still face multiple demands on their time during implementation, especially because implementing a reduction in repression is often indefinite, involving something like continually restraining security forces. 9 As in the other cases, then, leaders have incentives to make promises in the short term for mutual gains but also have incentives to renege in the longer term (see especially Conrad 2014, Conrad & Ritter 2019, Vreeland 2008)-and they have the information and control asymmetries to do so (e.g., Simmons & Danner 2010, pp. 232-33). This mechanism differs from classic conceptions of some spiral or boomerang models because those often envision opposition groups that cannot initially force the government to strike a bargain and are obliged to rally other pressure to extract even the first concessions (Hawkins 2002, Keck & Sikkink 1998, Risse et al. 1999. The empirical record also shows examples of reversion problems in the realm of repression and other government abuse. While the 1994 genocide in Rwanda was one of the most severe instances of a systematic human rights violation, what happened in its wake shows how hard it can be to hold a government to a bargain over time, even when it has agreed to protect a population. Before the genocide, and in conjunction with ethnic civil conflict, the government had formally signed onto a set of domestic and international agreements that included the protection of human rights (Hafner-Burton 2005, p. 609; US Dep. State 1994). After the genocide, the Tutsi-dominated Rwandan Patriotic Front that had fought its way back into power, facing pressure from all sides, installed a power-sharing government including Hutus in the state's top posts and promised to protect "all Rwandans" (Bonner 1994). Despite the commitment, the army reneged soon after and raided refugee camps housing internally displaced people-and the government distanced itself from the actions, not uncommon with reversion problems, though it had some control over the forces (Hafner-Burton 2005, p. 609).

Political Rather Than Economic Issues
The bargains plagued by reversion problems can be about political issues-e.g., how factions contest and hold office, how those who are in control use state forces against those who are not-or about economic issues. To the extent that the existing literature considers the systematic problem of constraining the government, its focus is usually on economic issues. Problems between domestic actors such as property disputes, or, especially, problems between domestic and international actors such as expropriation, are somewhat analogous (this is an extensive literature, going back to Williamson 1971Williamson , 1985North 1981North , 1994; and more recently Acemoglu & Robinson 2001, Acemoglu et al. 2005, Easterly & Levine 2003, and Helpman 2004. Theory developed in the context of contract enforcement and taxation more broadly arrives at the conclusion that 8 Indeed, "organizational" rights, such as the ability to establish an opposition party, are more successful in establishing protection than "individual" rights, such as freedom; this may be because organizations have incentives to protect themselves by mobilizing their members, although it is not clear that these collectives then have any direct effect on bargains concerning other repression (e.g., Chilton & Versteeg 2016). 9 Exploring dynamics within the government, such as considering principal-agent problems in security forces, demonstrates additional challenges (e.g., recently, Conrad  www.annualreviews.org • International Actors and Domestic Deals contracting with the government can be especially difficult because the government's ability to renege differs from the abilities of other domestic actors (Acemoglu 2003). 10 While political and economic issues overlap on aspects of contract enforcement, for example, this review focuses on political reversion problems, which are often more difficult to detect than economic reversion problems and are often less likely to automatically invoke international actors such as multinational corporations and their home countries. Existing work on economic issues can inform our understanding of reversion problems that concern political issues, but the two categories also diverge in ways that tend to make different solutions available to leaders and their opponents. Consider governments breaking contracts (Wellhausen 2015) or hold-up problems (Carnegie 2014). These challenges in the economic realm occur in relatively strong political institutions that produce self-enforcing solutions (e.g., North & Weingast 1989, Ordeshook 1992, Przeworski 1991, Weingast 1997. If a government knows that consensus exists on how much it can tax, for example, and that other actors will collectively respond to a violation, the government will avoid transgressions that would incite its removal (Weingast 1997, p. 246). But consensus on limits and commitment to defend those limits are needed to make self-enforcement work (Weingast 1997, pp. 251-52). Consensus on limits is unlikely where reversion problems arise in transitioning states or divided societies-those with weak political institutions are inherently unstable and also tend to lack enforcement (Levitsky & Murillo 2009)-which removes the self-enforcement options in many cases when political issues such as civil conflict arise (e.g., Fearon 2011;Wantchekon 2004, pp. 17, 27).
Economic and political issues also face different types of international interest in these situations. Economic issues do not necessarily receive self-enforcement, as with expropriation threats to multinational companies; however, international actors are often motivated in these cases to threaten enforcement (e.g., Büthe & Milner 2008, Elkins et al. 2006, Ginsburg 2005, Kerner 2009, and, especially, Vreeland 2003; see Post 2014 for an overview). The international actors are motivated by their own powerful members when they are intergovernmental organizations, by lobbies, or even by constituents who financially benefit from that enforcement. In contrast, political issues do not similarly motivate international actors. Some of these political problems could cause later economic problems, and they may also pose transnational threats, including the spread of conflict, crime, and even disease; however, these are diffuse risks (e.g., Crocker 2003; Eizenstat et al. 2005;Fearon & Laitin 2004;Fukuyama 2004, p. 92). No single actor is motivated to act alone, and even for collective action the cost needs to be low (Eizenstat et al. 2005). Moreover, most reversion problems related to political issues, unlike economic issues, are specific to the particularities of the bargains, so complete specification in the bargain is difficult and the task for international actors harder. But work on each category informs the other.

ENFORCEMENT BY INTERNATIONAL ACTORS AS A SOLUTION
How then do international actors help leaders and opposition groups overcome reversion problems and commit to mutually beneficial bargains across political issue areas? International actors at times have incentives to help enforce these political bargains in order to remove specific problems that threaten to disrupt their relationships with these states or to remove the public disorder these problems produce, provided the costs are sufficiently low and can be shared through mechanisms such as coordinated intergovernmental organization action (e.g., Fearon & Laitin 2004).
Also, because political institutions are typically too weak to constrain domestic actors from violating bargains when reversion problems bind, domestic actors at times have incentives to use international actors' enforcement to make future noncompliance detectable and subject to punishment. 11 Opposition groups are often most concerned, but leaders also stand to benefit from committing to these bargains because they secure the payoffs of peace, reduced resistance, and so forth. The conditions under which international and domestic actors are most likely to have these incentives to establish enforcement mechanisms are discussed in the final section of this review. The question in this section is, provided such conditions exist, how do international actors help in a cost-effective way?
I posit that international actors across political issue areas can enforce compliance in domestic bargains when they can detect and punish noncompliance. To serve this role, international actors need mechanisms for monitoring and for providing incentives conditioned on compliance.

Monitoring
International actors first must verify compliance by domestic actors. 12 In some cases, compliance with the terms of an agreement will be obvious-for example, if negotiators establish how many fighters must be demobilized or how many weapons decommissioned, then they can track those metrics-but, in other cases, it will not. Indeed, identifying violations is often difficult, as governments can use state structures to subtly seize more power than they were allocated, reverting to fuller control of their institutions than agreed while politically disadvantaging opposition groups. Identifying the difference between violations and other actions, including unintentional noncompliance, is crucial to overcome any commitment problem (e.g., Fortna 2008), but it is especially important with the subtleties of reversion problems. International actors therefore need to create clear standards for what compliance with bargains looks like, either by pushing domestic actors to state public benchmarks and regular milestones or by offering their own interpretation of these if not made available, and then they need to check domestic actors' behavior against these standards, especially at important moments for implementation.
International actors can use various mechanisms to monitor compliance. They can station foreign observers in a state, continuously or periodically; investigate reports of violations in the state; or triangulate between such reports from outside the state. Across issue areas, international actors employ a variety of these mechanisms, some more than others. Having dedicated observers in the state or at least continually observing the state may be the most effective but not the most cost-efficient mechanism (e.g., Fortna 2008). 13 Elections and other systematized spotlights, where institutions repeatedly draw international actors' attention to crucial moments for domestic bargains, are especially useful to help international actors monitor compliance. Systematized spotlights do two things: First, they establish what compliance means by predetermining public benchmarks and regular milestones that then are known to all actors; second, they mobilize detection of noncompliance at moments that are crucial for the bargains. Certain institutions thereby increase international attention and accountability that ease detecting and punishing violations. Elections serve this role by establishing these benchmarks in each cycle (Fearon 2011), specifically when power is (re)distributed between actors (Matanock 2017b(Matanock , 2018. In addition, elections draw the attention of domestic actors (Fearon 2011) and, especially since the end of the Cold War, the attention of international actors, too (Matanock 2017a,b). International actors, primarily intergovernmental organizations and representatives of democratic governments, provide international election observation and other monitoring in many cases (Hyde 2011b, Kelley 2012. Bargains can use alternative institutional designs to create aspects of systematized spotlights, especially adopting strong obligations on precise terms in publicly signed treaties to help form easily assessable benchmarks and milestones (based on theory on treaties in general; see Koremenos 2016;Lipson 1991;Morrow 2007;Simmons 2009, p. 277). These agreements, however, often fail to produce the same attention after the initial moment of interest. For example, human rights treaties draw attention during an initial signature, but less afterward (Nielsen & Simmons 2015), unless they somehow regularize and draw attention to reporting (e.g., Kelley & Simmons 2013). Alternative monitoring mechanisms-for example, sending peacekeepers to be on the ground continuously or establishing a committee specifically staffed to monitor human rights as part of a preferential trade agreement-are often more costly but potentially also effective.

Providing Incentives Conditioned on Compliance
International actors next must incentivize compliance with political bargains, either by providing rewards for good behavior or punishments for bad. 14 Monitoring compliance is not sufficient because, if one side is temporarily stronger than its opponent during implementation, it has an incentive to renege on even a mutually beneficial bargain (as the government does with reversion problems). Overcoming commitment problems generally entails somehow tying hands (Fearon 1997). In these cases, international actors are providing these mechanisms for domestic actors, as the international actors provide a cost to offset the domestic actors' benefit of reneging on a bargain. The threat alone may be enough to maintain compliance, as long as it is credible. I discuss when international actors have incentives to use, and condition, their resources in the final section of the review.
International actors can use a spectrum of tools to impose costs on domestic actors for noncompliance with bargains. These range from limiting choice through the threat or use of force ("coercion")-including not only compellence but also power to arrest and sentence to punish past action and even deter future action-to a range of conditional incentives (also called "inducement"), and even potentially beyond material mechanisms to naming and shaming (also called "persuasion"). [These terms follow Howard (2019), although they are not consistent across studies; Krasner (2015), for example, uses "imposition" in place of "coercion."] The mechanisms involving force or coercive action may be most productive to shape behavior but also prohibitively costly in many cases-even to end civil conflict (e.g., Fortna 2008). 15 As with continuous monitoring, few cases can rally sufficient international involvement to implement coercion (e.g., Simmons 2010, p. 277). Using force may also be viewed as disproportionate in many cases, as most reversion problems concern mainly the subtle redistribution of political power. Again, across issue areas, there are cases of all these mechanisms, but some are used more than others.
Considering conditional incentives, rewards such as foreign aid or preferential trade terms are often the most available mechanism by which international actors can help solve reversion problems. International actors can condition these incentives on compliance with a domestic bargain to indirectly alter its costs and benefits. 16 In fact, economic, legal, and political benefits can all be withdrawn in response to violations (e.g., Donno 2013). 17 Offering these benefits, and threatening to withdraw them, is typically more feasible for international actors than direct enforcement. International actors tend to provide substantial funding to states facing transitions and other crises (e.g., Collier & Hoeffler 2004, p. 1136Findley 2018), including not only development aid but also democracy assistance, which expanded during the 1990s (Bjornlund 2004, Bush 2015, Carothers 1999, and some of this funding can be effectively conditioned on compliance. Democracy assistance, for example, tends to be conditioned for all parties on compliance with constitutional rules and procedures that are tied to bargains in many cases (Azpuru et al. 2008;Carothers 1999;McFaul 2010, p. 20). This reinforces the usefulness of elections as systematized spotlights, as described above, because elections often receive not only observers but also conditional incentives. While initial arguments suggested that donors might be unwilling to cut off aid-the "Samaritan's Dilemma" (Buchanan 1975)-recent research shows that threats to withdraw assistance are often carried out (e.g., Wright & Winters 2010). These levers, together with trade deals and other incentives associated with normal interactions, are often available to international actors (e.g., Finkel et al. 2007, Hafner-Burton 2005, Matanock 2017b, Matanock & Lichtenheld 2019 and provide economic, political, and legal tools to shape behavior.

Solutions Across Issue Areas
To explore enforcement by international actors as a solution to reversion problems, this section revisits the same issue areas that were used above to illuminate the problems themselves: ending civil conflicts, opening elections, and reducing repression.

Ending civil conflicts.
Ending civil conflicts is the area in which international actors are most likely to use more costly mechanisms such as coercion, and yet the cases still show significant use of conditional incentives and of monitoring through systematized spotlights.
While peacekeepers can at times use coercion, the United Nations and other intergovernmental organizations tend to threaten force to protect civilians and deter attacks at most. Foreign states operating as peacekeepers are more prone to do so. Some peacekeepers also patrol buffer zones, arrest violators, and even at times try them through tribunals that are majority foreign. In addition, these peacekeepers provide continuous monitoring while on the ground (e.g., Clayton et al. 2017, Fortna 2008, Howard 2019, Hultman et al. 2016, Walter 1997. In Zimbabwe, the Patriotic Front and the government signed a settlement that provided each side with protections, especially focusing on the rebels. For example, not only the rebels but also Rhodesian civilians who supported the settler government were required to disarm, both sides disbanded forces, and the assembly points for demobilizing rebels were strategically located far from Rhodesian bases and close to Mozambique and Zambian borders to provide additional protection. However, the rebels were still quite concerned about the process, and so they actually signed the Lancaster House Agreement only after Britain agreed to send 1,200 British and Commonwealth forces to the state (Walter 1997, p. 353). Asked if this provision of forces was important to the rebel 16 Conditionality can refer to the strategic use of foreign aid to induce governments to undertake economic structural adjustments (Haggard & Kaufman 1992), but I use the term to mean any promised or imposed reward or related punishment in response to domestic actors' behavior (Girod & Tobin 2016). 17 While the literature on economic issues emphasizes traditional sanctions, it also raises the possibility of removing benefits, as with International Monetary Fund conditionality (Stone 2004, Vreeland 2003, complementing this research on political issues. decision, Mugabe, the leader, stated that "it was the really vital element and it prevented the Rhodesian forces, you see, supported by the South African forces-who were in the country, don't forget-from attacking our assembly points" (interview quoted by Walter 1997, p. 351). The British and Commonwealth soldiers watched and were ready to act with force if needed. The settlement succeeded.
Recent work, however, shows that many postconflict missions are not mandated for coercive force (e.g., Howard 2019, Matanock & Lichtenheld 2019). While coercion can be used to enforce domestic bargains, international involvement rarely threatens force (broadly, see Simmons 2010). While UN peacekeepers can use force-especially when control over a state such as Cambodia or Kosovo has been entirely delegated to the organization-the mandates of almost half of all missions do not provide for coercion (Matanock & Lichtenheld 2019). However, conditional incentives, alongside monitors, are employed with stabilizing effects (Matanock & Lichtenheld 2019). Policy makers, including a prominent UN official, note that, in postconflict settings, foreign aid tends to be "the main repository of 'sticks' and 'carrots' to keep a peace process on track" (Arnault 2006, p. 12). While peacekeeping missions provide some monitoring, the use of elections with former combatant parties as part of peace agreements has become widespread (Matanock 2017b), allowing additional observation of government and rebel party compliance throughout implementation. International actors can help monitor compliance with the agreed power distribution at these crucial moments, through unarmed observers focused on elections, and use donor aid and other assistance to incentivize it (also see Flores & Nooruddin 2016).
International actors used conditional incentives to help overcome reversion problems in El Salvador. The elections in postconflict El Salvador were a central component of international intervention: Not only had the rebels agreed to become a political party as part of the powersharing agreement, but also, in 1993, the international community turned out to enforce this transition. The UN mission doubled in size when it formed an electoral division, which deployed 900 additional observers (Call 2002;Howard 2008, p. 100), and 4,000 total observers were dispatched by various international and regional bodies (Montgomery 1998, p. 131). Election observation was also tied to the provision of foreign aid (Sullivan 1994, pp. 86-87) and was hugely important ( Johnson 1993, Karl 1992, Rosa & Foley 2000. Provision of aid was clearly, although unofficially, conditioned on both sides abiding by the peace accord (e.g., Rosa & Foley 2000, p. 150). As expected, then, after the 1992 peace agreement, when the government was suspected of delaying voter registration cards prior to rebel party participation, potentially leaving rebel supporters unable to vote, the response was swift: UN officials went to municipalities to investigate complaints, and, once they verified government noncompliance, the United States incentivized compliance by freezing $70 million in foreign aid. The government then complied, registering 90% of eligible voters in time for the election (Human Rights Watch 1994). When the government sought to move polling stations from FMLN strongholds to a department capital, again disenfranchising rebel supporters, international actors responded: The UN electoral observation mission disputed security claims, offered to provide logistical support, and pressured the government to comply with the peace agreement (Montgomery 1998, p. 131). Elections accompanied by foreign aid can thus be used to help enforce a bargain about ending civil conflict.
Peacekeeping is perhaps an area most likely to see continuous deployment and coercive force, yet, even here, international actors often use unarmed observers focused on systematized spotlights such as elections and donor assistance to incentivize compliance with these bargains.

Opening elections.
Opening elections is a situation where the most obvious type of intervention to enforce agreements is periodic monitoring alongside conditional incentives such as aid or preferential trade terms, although other mechanisms are also occasionally employed.
As with postconflict contexts, coercion is occasionally used. Foreign-imposed regime change (FIRC), for example, at times seeks with limited success to promote democracy-but FIRCs overall are rare (see Krasner & Weinstein 2014). Despite the democratizing rhetoric, many of these missions also simply may not prioritize more open elections over other foreign policy goals (e.g., Bueno de Mesquita & Downs 2006).
In many cases, international election observation is deployed to provide monitoring (Hyde 2011b, Kelley 2012, and it is often tied to conditional incentives where observation is a mechanism for "unlocking foreign aid or loans or averting foreign support for opponents" (McCoy 1993, p. 133), although crises still make reversion problems clear to international actors without monitors and at times draw even stronger sanction. Highly conditional democracy aid initially was sent mostly to regions close to Western powers (Azpuru et al. 2008, Youngs 2008, but, by the turn of the century, it flowed to 100 states (Carothers 1999). On the whole, results are mixed as to whether these tools produce stronger democracies (Bjornlund 2004;Carothers 1997;Dunning 2004;Hyde 2011a,b;Kelley 2012;Knack 2004). However, systematic evidence shows that electoral integrity is often monitored and incentivized through these international mechanisms (e.g., Donno 2013). Studies also indicate that donors withdraw foreign aid for egregious violations in particular (Boulding & Hyde 2005). Negative reports from observers can especially strengthen opposition parties' claims of broken bargains, although such reports can spark violence if the violation is not remedied rapidly (e.g., von Borzyskowski 2019). These challenges can even remove reverting leaders (e.g., Cooper & Legler 2006).
International observers in the Dominican Republic advocated for negotiations between the incumbent and the challenger after the 1994 elections were widely perceived as fraudulent. These negotiations culminated in a pact to hold new elections that would exclude the incumbent the following year, and the international community exerted influence to support the bargain. Both the Organization of American States, which was viewed as relatively neutral, and the United States, which held significant economic levers, pressured the parties to enact and implement a pact whereby the incumbent's win would be overturned with new elections; the incumbent did manage to delay, but ultimately caved to these international forces and stepped aside for new contests in 1996 (Donno 2013, pp. 130-33;Hartlyn 1998, pp. 228, 254-55, 262-64).
Opening elections is often an occasion for periodic monitoring alongside conditional incentives to enforce agreements. However, given the systematized spotlights provided by elections, at times international actors need only to triangulate between reports on such situations because the violations can already be identified through the campaign or voting processes. Observation also, at times, may not be linked to conditional incentives-although systematic data on the co-occurrence are difficult to obtain-and then perhaps the pressure falls more into the realm of naming and shaming (a mechanism discussed more in the context of protecting human rights, below).

Reducing repression.
Reducing repression is a third area in which costly mechanisms are occasionally used. Periodic monitoring and investigating reports alongside conditional incentives are also often used in the context of reducing repression. However, some of these international institutions have a lighter touch than in the other issue areas. 18 The mechanisms for monitoring and conditioning incentives on reduced repression are diverse, and, perhaps because they at times include generalized treaties or courts into which governments opt, creating international legal obligations, they are often studied separately. These 18 There are myriad studies of the broader subject of compliance with human rights standards and the effect of international human rights law; this review focuses on the mechanisms international actors use in cases with reversion problems.
www.annualreviews.org • International Actors and Domestic Deals human rights institutions in theory lock leaders into bargains or even unilateral reductions in repression, among other provisions (Landman 2005, Moravcsik 2000. These institutions have grown as even authoritarian states sign on (e.g., Vreeland 2008). They notably include the UN Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (CAT) and the International Criminal Court (ICC). As in the other issue areas, at times these mechanisms employ coercive force or something akin to it. The ICC is one of the most binding institutions-a "self-binding commitment" (Simmons & Danner 2010, p. 234). International courts and tribunals, although they face some limitations, have the ability to prosecute cases against signatories independently, even cases brought by other sources (e.g., Schabas 2016, Schoner 2019), and act directly to punish those in their custody (Ritter & Wolford 2012). Instances in which courts punish are rare, as are missions that use force against perpetrators to quell violence, but they do perhaps have broader effects, such as deterring future violations (Appel 2016;Bellamy 2009;Jo & Simmons 2016;Keohane et al. 2009;Kim & Sikkink 2010;Moravcsik 2000;Murdie & Peksen 2014;Simmons 2009Simmons , 2010Simmons & Danner 2010).
While diverse, and often studied separately, other mechanisms by which international actors monitor and condition incentives on compliance are also clearly operating in the context of repression. Creating effective monitoring, especially systematized spotlights, is difficult in many of these cases, but it can pay off: Firm language about human rights recommendations, for example, produces more implementation (Linos & Pegram 2016). Many international institutions, such as the CAT (e.g., Vreeland 2008) and the UN Commission on Human Rights (e.g., Lebovic & Voeten 2009), have some committee or other mechanism for monitoring, but, in addition to problems like limited staff, they are often not reinforced by the focus of systematized spotlights. This makes violations harder to detect. States committed to "hard" human rights standards tied to preferential trade agreements and other material rewards that are conditional often establish their own monitoring (Hafner-Burton 2005): For example, the Lomé and Cotonou Agreements dictating terms of ties between the European Union and African, Caribbean, and Pacific states "require habitual assessments of national developments concerning human rights" (Hafner-Burton 2005, pp. 606-7, quoting the agreement language), which meant that 16 violations had produced consultations by 2005 (Bradley 2005, Mackie & Zinke 2005. Even these mechanisms may miss less obvious crises. While domestic opposition groups push for their states to opt into international institutions protecting human rights, for instance, it is not always clear exactly how much power various mechanisms have over leaders. For example, the CAT allows leaders to opt out of two optional protocols that would otherwise provide more leverage to punish repression (see Hill 2015 on reservations broadly), making the cost of signing on relatively small and likely at best to bind in the longer term for incumbents (Conrad 2014, Vreeland 2008. 19 Many other conventions similarly have perhaps minimal positive effects at best (Hathaway 2002, Simmons 2009, Von Stein 2005, although some studies are broadly more optimistic on these legal mechanisms (e.g., Sikkink 2019), and the rising standards of accountability over time may produce some of the more pessimistic results (Fariss 2014). Beyond the international institutions, the practice of naming and shaming is often treated as a mechanism international actors can use directly to impose costs on states for repressing, perhaps through their reputation (Nielsen & Simmons 2015). The direct evidence on naming and shaming is generally positive (DeMeritt 2012, Franklin 2008, Hendrix & Wong 2013, Krain 2012, despite notable exceptions (Hafner-Burton 2008). But naming and shaming can also serve as a monitoring device that then induces some other type of punishment (Barry et al. 2013, DeMeritt 2012, Franklin 2008, Hafner-Burton 2008, Hendrix & Wong 2013, Krain 2012, Lebovic & Voeten 2009, Murdie & Peksen 2014, Nielsen 2014, Peterson et al. 2016. Alongside this or another form of monitoring, the incentives conditioned on compliance to punish reneging include the reduction of foreign aid, direct investment, and even exports in the studies just cited , Murdie & Bhasin 2011. Economic sanctions are often studied separately and show more negative effects (Peksen 2009, Peksen & Drury 2009, Wood 2008. Now, mechanisms that work may not always work through international monitoring and conditional incentives-for example, the domestic coordination may play a crucial role (Conrad & Ritter 2019), or these mechanisms may interact with domestic coalitions and institutions, potentially adding power (e.g., Lupu 2015, Powell & Staton 2009, Simmons 2009)-but there is substantial evidence of at least some similar dynamics (especially recently, see Carnegie & Marinov 2017).
International actors did apply pressure in Rwanda. Their efforts were ineffective at the start of the fast-moving genocide, indicating some of the limits of this type of intervention. But the European Commission through the Lomé Agreement did subsequently use its leverage when the government took power and violated refugees' rights. This international actor, which had designated millions of dollars to help rebuild the state, suspended its funding until the army perpetrators were punished (Hafner-Burton 2005, p. 610). These violations were identified because, particularly after crises, the Lomé Agreement called for monitoring and periodic review. More broadly, however, the Rwandan government has manipulated other international processes strategically, so this positive outcome does not characterize the whole case (Loyle & Davenport 2016).
While the effects are mixed, perhaps in part because systematized spotlights and even conditional incentives are often less available for human rights violations than power transitions are, even reducing repression is an area in which international actors use some similar mechanisms. The cases of interest in this review entail an opposition group making an informal bargain with the government, at a minimum, but repression also typically has more diffuse effects than issue areas where organizations are the target, potentially producing some differences (more on this below). 20 Many of these cases show periodic monitoring but also investigation of complaints, often triangulating between staff reports from international organizations, alongside conditional incentives to enforce agreements.

EMERGING CONCLUSIONS AND REMAINING QUESTIONS
This review identifies a research agenda that incorporates both domestic and international politics across a range of topic areas. International actors frequently intervene in practice as policy makers seek to help end civil conflict, open elections, and reduce repression, but their actions across issues have not yet been systematically studied by scholars.
The major payoff to examining this phenomenon across issue areas is the understanding that international actors can enforce domestic bargains and overcome reversion problems, often by using conditional incentives and systematized spotlights. For domestic actors, political bargains involving international enforcement have a more efficient outcome; for international actors, they reduce transnational threats, often at a low cost.
This research agenda also has the potential to answer larger questions, especially about when enforcement mechanisms work and what their limitations are. Studies so far suggest that their effectiveness typically depends on international and domestic actors' interests and resources. The literature across issue areas already suggests two key conditions, discussed in detail below, under which international enforcement will solve reversion problems on political issues: (a) International actors must experience a Goldilocks condition in terms of their interests, and (b) domestic actors must experience state desperation in terms of their resources. Other results, however, are mixed and may benefit from additional theorizing and testing across issue areas. This work will include considering the design of the response but also differences in issue area. Such studies have the potential to yield policy recommendations, including for the current era.

Incentives and Resources
In many cases, overcoming reversion problems requires that international actors are interested in the domestic bargain but not so invested in a particular side that they ultimately cannot hold both accountable for compliance. This just-right interest on the part of international actors is the Goldilocks condition (Matanock 2017b). Studies across issue areas have identified two related sets of factors that shape the conditions, potentially producing a Goldilocks condition with the appropriate international involvement to help overcome reversion problems. First, the dynamics of the major powers in the system matter. During the Cold War, the Soviet Union and the United States were primarily interested in ideological loyalty from governments, so each superpower accordingly backed a particular side in other states (e.g., Bearce & Tirone 2010, Bermeo 2016, Dunning 2004. If states signed onto the Universal Declaration of Human Rights or other international treaties as part of domestic bargains, they were largely alone in implementing their commitments (e.g., Mayall 1996, p. 5). After the Cold War, the United States and its allies provided significant conditional incentives and were more willing to pressure states and engage in their affairs to promote peace and democracy, producing Goldilocks conditions in many cases (e.g., Feste 2003, Lebovic & Voeten 2009). Since the turn of the century, a shift away from a Goldilocks condition has emerged in many states: The attacks of September 11, 2001, for instance, focused these international actors on backing supportive partners in the War on Terror (Girod 2019), and other aid providers such as China often do not condition assistance on compliance with domestic bargains (Girod 2018). A Goldilocks condition therefore is less likely in certain eras-and also for certain partners in the current era.
The ties between international actors and domestic actors also matter in two distinct-and sometimes conflicting-ways. Closer relationships with governments reduce the likelihood of a Goldilocks condition but also augment international actors' leverage. First, close and crucial ties can mean that international actors favor their domestic partners over outcomes such as peace. In other words, international actors can become so partisan-often toward the government, given its hold on the levers of trade, communication, and more-that they can no longer help to overcome reversion problems. Instead, outside the Goldilocks condition, these international actors typically have incentives to ignore their preferred partners' violations of domestic bargains, affecting their willingness to use their leverage. Outsiders' strong economic and even strategic interestsincluding trading partnerships providing needed natural resources, protection of major foreign business properties, or, as noted above, partnerships combating common security threats-can thus make threats to punish government violations of domestic political bargains less credible (e.g., Carothers 1999, p. 45;Girod 2012Girod , 2018Hafner-Burton et al. 2019;Karreth et al. 2019;Nielsen 2014;von Borzyskowski & Vabulas 2019). 21 The factors that shape these ties, and even that motivate intervention, are beyond the scope of this review but are an important and related topic (see Bueno de Mesquita & Smith 2012).
From the perspective of the particular state, these same ties affect the government's responsiveness, as does state desperation (Girod 2012). When international actors possess strong ties with governments, in particular, they are more likely to hold sufficient leverage to sanction violations, even if they are also more likely to be partisan. International actors can use aid, trade, and intergovernmental ties as levers: Membership in trade networks or intergovernmental organizations may be the most compelling promise to produce democratization, for example (e.g., Börzel & Hackenesch 2013, Pevehouse 2005. But foreign aid is typically a widely available inducement in most unstable contexts (e.g., Collier & Hoeffler 2004). In fact, without leverage and close ties between states, international actors do not have the ability to meaningfully shape incentives for domestic actors (e.g., Levitsky & Way 2005. Wider regional or neighborhood dynamics may also affect the expectations that international actors will engage at all (e.g., Gleditsch & Ward 2006. Finally, when states are more dependent on international trade and foreign aid, for instance, they have stronger motivation to respond to conditional incentives offered by international actors. If governments have access to ample extractable natural resources, they may not be concerned about losing whatever aid they receive and so may not respond to calls to open their elections, for example (Girod 2012). Similarly, they are less likely to constrain their use of repression . Some work even tests the Goldilocks condition (specifically the government's strategic importance to the international actor) and the desperation factors together, showing that foreign aid garners more compliance in nonstrategic states needing access to these funds (Girod 2012, Girod & Tobin 2016. Some studies find other conditions under which states are receptive to international enforcement of bargains that otherwise face reversion problems, but they are not as consistent across cases, and they even suggest some systematic differences. Factors such as initial levels of government control, the strength or weakness of other domestic institutions, and even the regime type shape the willingness of governments to sign onto domestic bargains and of international actors to back them. For example, weak states experiencing violence as well as strong states with substantial human rights protections sign onto the ICC (Simmons & Danner 2010). These same factors then also affect compliance. Existing work shows that international forces are most likely to reduce repression in democratizing and newly democratic states (e.g., Hafner-Burton & Tsutsui 2007, Landman 2005, Neumayer 2005; see Simmons 2010 for a review). In contrast, autocracies may be more sensitive to naming and shaming when considering whether to reduce repression (Hendrix & Wong 2013). Research on electoral openness tends to treat democratization as variation to be explained (e.g., Donno 2013), but at times it is used as an explanatory factor for decisions to invite international actors (e.g., Hyde 2011b), and more of this might help explain mixed findings. The evidence so far suggests that more open regimes benefit more from observation, at a minimum (e.g., Kelley 2012). In postconflict studies, variables for regime type have a negative effect on aid conditionality, but this effect is not consistent when such factors as strategic importance are included in the models (Girod 2012). Theorizing more consistently about regime type, for example, and testing its effects within particular methods but across issue areas may help explain mixed findings currently in the literature. The Goldilocks condition and the desperation of the state already emerge across issue areas.

Designing the Response and Different Issue Areas
The design of the response arises as a particularly important factor when considering when these mechanisms work and what their limitations are. What information is provided through the monitoring, for example, may make the difference between likely success and failure (as Hendrix & Wong 2013 argue about the effects of regime type, described above). Moreover, while www.annualreviews.org • International Actors and Domestic Deals comparing monitoring and conditional incentives to other mechanisms such as coercion (e.g., Matanock & Lichtenheld 2019), for instance, recent work also usefully explores whether carrots or sticks work better to stabilize postconflict peace and encourage democratization (e.g., Cil & Huth 2015, Ethier 2003). As noted above, economic sanctions specifically also may have negative effects (also see Krasner & Weinstein 2014). Work on repression has focused more on why states sometime choose carrots versus sticks as a response, potentially an important precursor to understanding their effects (e.g., Hazelzet 2001, Kotzian et al. 2011). These design issues raise other related questions about which actor initiates the international enforcement, whether international actors always intend to serve this role, how purposeful the international actors are in their design, which international actors perform these functions in most cases and whether they do so through intergovernmental organizations, how private sector or traditional leaders (highlighted in other work) fit, and how long international actors have to continue to enforce political bargains over time. We know very little about the answers to these questions and how they affect the mechanisms-for example, the request for postconflict international enforcement often comes from the domestic opposition, perhaps shaping how well it works (e.g., Matanock 2017b)but they are crucial to policy because they determine, for instance, whether and when domestic institutions take over as guarantors over time.
In addition, differences in issue areas may shape how effective these mechanisms are in addressing reversion problems. The nature of the reversion problems and, particularly, how likely the problems are to affect international actors can influence the provision of incentives conditioned on compliance. As noted, reversion problems in some issue areas may be more obvious than others. Violations of constitutions and other formalized electoral rules are more evident to monitors, for example. Some issues inherently center on systematized spotlights such as elections, but others, like postconflict power sharing, have evolved to now incorporate the same mechanisms (Matanock 2017b). Protecting human rights, however, may not lend itself as easily to these mechanisms that make reversion easier to detect. Existing research has not yet shown that a similar institutional design can clarify noncompliance. Conditional incentives are also affected by issue type. Because civil wars have caused more casualties than interstate wars in the modern era and pose contagion concerns (e.g., Eizenstat et al. 2005), international actors may be especially willing to invest in ending these armed conflicts. Repression, however, is the issue area "perhaps the least likely to engage the attention of other states," which do not see improved strategic interest or "joint gains" from engaging in these cases (Simmons 2010, p. 288). Some issue areas may thus be more likely than others to prompt international enforcement, but the difference is due to strategic concerns that can also shape how well enforcement mechanisms work. Understanding these differences allows policy makers to know when these mechanisms are even available as an option.

Changing Concerns
New questions for study also arise from considering policy makers' concerns. One of the major considerations for policy makers seeking to secure peace, democracy, and human rights is to avoid making the situation worse. Backlash is a concern. Some work on this topic suggests that when domestic actors strike a bargain or have other broad incentives to invite international enforcement, backlash may be less likely (e.g., Glasius 2009). But various factors can foster backlash in the issue areas discussed in this review (e.g., Carothers 2006, Risse & Babayan 2015, Terman 2016. For example, recent work on China and Russia shows that leaders can be very strategic in fomenting backlash if they object to critiques or constraints over time (e.g., Ambrosio 2009, Gruffydd-Jones 2018). Another consideration for domestic and international actors alike is whether international enforcement mechanisms continue to exist at all with changes in US policy in the present era (e.g., Girod 2019).
Overall, this review shows that international actors can often use conditional incentives and systematized spotlights to overcome reversion problems and enforce domestic bargains on political issues. The Goldilocks condition, wherein international actors are invested but not too invested, and state desperation explain variation in the use of these mechanisms. Other conditions may also be important, but future work is needed to ascertain them. Many current studies focus only on one issue area-postconflict contexts, for example-but exploring across issue areas with this unified framework in mind may help us better understand reversion problems and the possibility for solving them through international enforcement.

DISCLOSURE STATEMENT
The author is not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.