In-group we trust,no more? How senders’ legitimacy shapes the success of sanctions-based democracy promotion

Abstract

For decades, pro-democracy states and organizations have been using coercive measures such as sanctions to promote and uphold democratic governance worldwide, yet not always successfully. Recent research has improved our knowledge on such ‘democracy sanctions’, but our understanding of how the different identities of sanctioners shape their success is not complete yet. Focusing on Africa, one of the biggest recipients of such measures but also a primary sender of them especially through the African Union, this article aims to systematically compare how the sanctioning behaviour of regional and extra-regional actors has impacted on the democratic status of the continent since the 1990s. The empirical analysis reconsiders the comparative advantage of regional actors’ legitimacy to explain the likelihood of democracy sanctions success. It shows that ‘in-group’ democracy sanctions in which regional organizations are involved have bigger chances of success than those imposed by higher-capacity yet less-legitimate extra-regional actors only. However, it also shows in-group legitimacy can erode and that cooperation with such local organizations can help contain global democratic erosion, but hardly achieve democratic deepening, due to soft, face-saving compliance criteria and different understandings democracy and its promotion have across the world.

Keywords

Africa coups democracy legitimacy regional organizations sanctions

Introduction

For decades, pro-democracy states and organizations have been using coercive measures such as sanctions to promote and uphold democratic governance worldwide. Yet, such ‘democracy sanctions’ have not always produced satisfactory results. The current sanctions literature still predominantly draws on a capacity-based argument of sanctions effectiveness, investigating if and how severity, namely the imposition of high material costs on the target such as those leveraged by a large front of senders or a global power with strong material linkages with the target, can effectively coerce a target into compliance. According to this argument, sanctions imposed by highly capable senders can produce significant economic pressure, reduce the space for third-party alternatives and thus create incentives for the target to comply. However, this has clearly not always been the case. Counterarguments criticizing this logic have never been lacking, in fact, pointing to the risk of cooperation failure (Miers and Morgan, 2002) or rally-round the flag effects (Galtung, 1967), among others. More recently, at a time in which regional actors are increasing their prominence in the sanction arena, reflections on a largely alternative legitimacy-based perspective have gained more support. Reconsidering the role of ideational factors in this sanction mechanic (Hellquist, 2021, 2022), scholars suggest the identity of the sender and its legitimate authority in the eyes of the target eventually matter. In the related literature of conflict mediation, legitimacy was indeed found to outperform material sources of power in mediating civil wars (Duursma, 2020). Is this comparative advantage of legitimacy-based responses also effective for democracy sanctions? Can legitimacy help explain, better than material factors did, the variation we witnessed in the results of sanctions-based promotion and safeguard of democratic norms worldwide?

To answer these questions, this article builds on and extends recent studies on sender composition in sanctions (Hellquist and Palestini, 2021; von Borzyskowski and Portela, 2018, 2023). Theoretically, it develops a framework on how in-group senders such as regional organizations can help overcome obstacles to sanctions-based democracy promotion by leveraging a comparative advantage in legitimacy-based mediation that generally more highly capable but out-group senders (such as extra-regional global powers) lack. Empirically, it focuses on the African continent – today’s most-sanctioned world region but also one of the most active senders through its regional organizations (Attia and Grauvogel, 2023; Felbermayr et al., 2020; Kirilakha et al., 2021). Against this backdrop, it systematically tests the empirical effects that the very identity of senders, in terms of different degrees of in-group versus out-group legitimacy, has had on the success of all sanctions that have been imposed across Africa to promote and uphold democratic norms since the end of the Cold War.

Africa provides an excellent case to address this puzzle and thus to compare legitimacy- and capacity-based arguments on sanctions effectiveness, especially in the subset of democracy sanctions, because this is the realm in which African regional organizations have lately become particularly active, thus offering a broad and diverse range of democracy sanction senders active on the continent to contrast for a systematic comparative analysis (Charron and Portela, 2015; Cowell, 2011; Eriksson, 2010; Hellquist, 2021; Hohlstein, 2022; Nathan, 2017; Omorogbe, 2011; Powell et al., 2016; Souaré, 2014; von Borzyskowski and Portela, 2018). It is therefore a typical case of external democracy promotion which can inform also other world regions similarly targeted by such initiatives, even more so given its presence of more extreme, unusual conditions in the key explanatory variable of promoters. Indeed, having a richer variety of actors involved in such endeavours, it can teach lessons other world regions may have not experienced yet, or at least not with such a magnitude, and that are relevant also to understanding how to prevent the findings emerging from this African-focused research regarding the undesired effects of in-group legitimacy erosion.

The article proceeds as follows. The first part situates this contribution in the literatures of external democracy promotion and sanctions, specifically reviewing the current state of the art in relation to the impact of sender composition on their success, including the role of regional organizations. In this regard, it also sets the context for the use of sanctions in Africa, clarifying various senders’ sanctions doctrines in the region. Then, it introduces the article’s theoretical framework on the impact that senders’ legitimacy can have on the success of democracy sanctions in Africa. On this basis, it categorizes these multiple senders active in Africa, using a typology of sender composition based on in-group versus out-group legitimacy and the number of senders. The second part applies this framework to Africa, with a systematic empirical test on all the sanctions regimes imposed on African countries over the past 30 years (1990–2021) with the objective of democracy promotion, using various model specifications that take into account possible selection effects between the imposition and outcome stages of sanction regimes. The results of this quantitative analysis inform scholars and policymakers of possible democracy promotion counterproductive effects as they show unilateral or plurilateral sanction regimes which do not include local partners can make sanctions success less likely than sanctions regimes in which regional organizations are also involved. However, the results also show that local partners can have softer, face-saving criteria for compliance also due to different conceptualizations of democracy, and that local partners’ comparative advantage in legitimacy can erode. The conclusion discusses the global and regional implications of these findings.

Coercive democracy promotion: The role of sender composition

Among the various tools the literature on the external dimension of democracy promotion – intended both as support for democratization and protection from autocratization (Leininger, 2022) – has focused on we can certainly find coercive measures such as sanctions. In their seminal work on external coercive democracy promotion, Abel Escribà-Folch and Joseph Wright (2015) describe sanctions as one of the most common foreign policy tools used to this end, together with aid suspensions, naming and shaming campaigns, and military intervention, among others. Although this instrument had long been questioned for its ineffectiveness and even counterproductiveness (Peksen and Drury, 2010), recent research proved it can induce democratization under some conditions, such as when the domestic institutional structures and narratives of the targeted regimes are taken into account (Grauvogel and von Soest, 2014; Escribà-Folch and Wright, 2015; von Soest and Wahman, 2015a). In the broader sanctions literature, scholars’ attention also focused on external factors, including the identity of the actors imposing such measures. Primarily, this concerned the role of unilateral as opposed to multilateral or universal senders in influencing sanctions effectiveness (Bapat and Morgan, 2009; Hufbauer et al., 2007). It is indeed commonly assumed that multilateralism can increase the capacity of the senders’ front and reduce alternatives to a more isolated target (Martin, 1992) thus boosting effectiveness. This is even more so when the front of the senders includes the target’s major trading partners and allies (Kavakli et al., 2020; McLean and Whang, 2010). Yet empirical research also warned about problems of cooperation when too many ‘cooks’ are involved (Miers and Morgan, 2002). This concerns not only the enforcement, due to free riding, but also the decision of what measures to impose (Drezner, 2000).

In investigating the identity of the senders, this debate has lately also touched upon the role of so-called regional senders, namely regional organizations imposing sanctions against their members (von Borzyskowski and Portela, 2018). In this growing debate, scholars have discussed how cooperation between regional and extra-regional senders impacts on sanctions effectiveness (von Borzyskowski and Portela, 2023) as well as the different approach such senders have towards the targets (Hellquist, 2022). The latter finally acknowledged the diverse levels of legitimacy senders may have in the eyes of the target, but stopped short of empirically evaluating, in a systematic fashion, its impact in terms of effectiveness.

Because the literature is still missing such a systematic comparative empirical evaluation of the sanctioning behaviour of such regional and extra-regional powers in terms of effectiveness, including in the realm of democracy promotion, the following section develops a theory which looks specifically into the effect generated by the involvement of regional organizations on democracy sanctions success. In doing so, it integrates theoretical expectations on regional organizations’ behaviour with those on sanctions effectiveness, to explain how in-group regional actors’ higher legitimacy can achieve concessions, overcoming obstacles that out-group coercive democracy promotion commonly faces.

As already mentioned, the focus will be on Africa because it is a region heavily targeted by sanctions-based democracy promotion. Emblematically, it is the continent in which the United Nations imposed the only two sanction programmes ever approved during the Cold War, against the two apartheid regimes of Rhodesia and South Africa. Moreover, more than any other world region, since the 1990s, it has presented a strong record of both regional and extra-regional senders, thus with rich variation in in-group and out-group legitimacy and capacity, a pre-requisite for a comparative analysis on the effect of sender’s legitimacy on democracy sanctions compliance.

After the Cold War, more than the UN, which imposed sanctions primarily for non-democratic objectives, the extra-regional senders most active in imposing democracy sanctions in Africa were the European Union (EU), a sui-generis regional organization which imposes sanctions also against non-member states through its Common Foreign and Security Policy and the conditionality clauses introduced in its trade agreements in case of democratic violations (Saltnes, 2017), and the United States, the single-most active user of sanctions with an oft-neglected soft spot for Africa, as one-third of all its post-1990 sanctions cases have been on African targets (Felbermayr et al., 2020).

Regionally, instead, the actor endowed with the most developed legal frameworks against democratic sanctions violations has been the African Union (AU), the continent-wide organization created in 2002 to replace its predecessor, the Organization of African Unity (OAU), founded in 1963. Because of its founding principle of non-interference in the domestic affairs of member states (except for cases of colonialism and apartheid), the OAU had shown much indifference to actual and potential violators of civil liberties and political rights across the continent in the previous decades. The very institutional features of the OAU (meeting only once a year, save for extraordinary sessions convened by a two-thirds majority, with no explicit authority for expulsion or suspension of its members – Udombana, 2002) were unfit to carry out rapid external intervention including the imposition of sanctions (Piccone, 2004)¹. Only in the late 1990s did its spirit of ‘non-interference’ begin to change towards a new ‘non-indifference’ principle. And democracy, more specifically the respect of democratic principles in accessing power, was precisely the theme for which they began to ponder the use of sanctions².

Upon and following the creation of the AU in 2000–2002, African states adopted some key documents which form the legal basis for its sanctioning behaviour³ and which authorize the AU assembly or, since 2004, also the more versatile, 15-member-strong Peace and Security Council (PSC) to suspend or further sanction with other measures African state authorities, and those who supported them, who violated the constitutional order of their countries⁴. The violations listed in these documents concern the replacement of elected governments (thus unconstitutional changes of government, UCGs) either by a coup d’état, or by mercenaries, armed dissident groups and rebel movements, as well as the refusal of incumbents to relinquish power to the winning party after free and fair elections.⁵ The perpetrators of these anti-democratic actions are also banned from standing for elections according to more recent provisions contained in the 2007 Addis Charter, unless exceptions are negotiated (Nathan, 2017).

After the AU, ECOWAS is the subregional organization which has engaged the most with sanctions⁶. ECOWAS did not include democratic sanctions provisions in its founding treaty (1975, revised in 1993) but adopted a Protocol on Democracy and Good Governance in December 2001, ensuring it could deal specifically with UCGs with an automatic mechanism similar to that of AU, at least as long as suspension is concerned (other types of sanctions or military interventions still need to be agreed with unanimity – Cowell, 2011).

Building norms softly: Explaining democracy sanctions success through in-group legitimacy

Drawing insights from research that has pushed forward a comparative advantage of regional actors’ legitimacy in conflict mediation, this article proposes that such in-group legitimacy certainly has a positive impact on sanctions effectiveness as well, through a camaraderie mechanism that builds on, on one hand, a perceived obligation of the target to obey, and on the other hand a gentler mode of execution that aims to mediate the correction of the wrongdoing as quickly as possible, even at the cost of softer compliance, so as to save the face and not ridicule the wrongdoer. But this article also warns of potential risks. One risk is legitimacy erosion, namely the possibility of undermining the perceived legitimacy of such regional actors when the very camaraderie principles are disavowed. Another risk is that in-group legitimacy can face limits when it comes to ensuring robust democratic compliance as the softer, face-saving nature of camaraderie can hinder the depth of democratic changes required to lift the restrictive measures.

Compliance through soft in-group camaraderie

From a non-normative, Weberian perspective, legitimacy is a relational concept between two entities in which one sees the claims made by the other as being justified, and therefore follows them on the basis of a perceived obligation to obey (Gerschewski, 2018). The role of legitimacy has received growing consideration in several internationalist sub-literatures (Franck, 1988; Pevehouse, 2002), ranging from conflict resolution to sanctions proper, as something which ‘influences [regional and international organizations’] ability to secure compliance with [. . .] rules and norms’ (Tallberg and Zürn, 2019), after years in which capacity-based explanations had dominated. As one of the latest such investigations, Duursma (2020) has recently considered how African third parties are effective in regional conflict mediation because of the higher degree of legitimacy they have compared to non-African third parties. Similar regional legitimacy-based considerations can be applied to the field of sanctions and external democracy promotion. In doing this, a useful starting point is the distinction introduced in this literature between in-group and out-group sanctions by Hellquist (2021, 2022), on the basis of the kind of relation between the sender and its targets.

Collective identity and belonging, such as the one derived from shared in-group membership in a regional organization, can definitely be a source through which one achieves and maintains legitimacy (Tallberg and Zürn, 2019). Such in-group legitimacy is based on the acceptance of the rightfulness of a system or institution by its members, and specifically its claim to rule (including over promoting and upholding democratic governance in that space), that derives from the sense of belonging to it. In the African context, for example, regional actors such as the African Union enjoy such in-group legitimacy not only because of the presence of contractual bonds with their member states – the latter do yield part of their sovereignty to the former, which is legally entitled to intervene on agreed matters – but also because of a deeper knowledge of the local context and cultural proximity (Sabrow, 2017; Tavares, 2010). The result is an environment of ‘camaraderie’⁷ in which members feel comrades and see the regional actor’s demands as something that has to be obeyed.

Moving to a sanctions-specific scenario, it is widely acknowledged that, in general, to obtain compliance you need not only credible threats but also credible assurances (Cebul et al., 2021). Consequently, a target is expected to be more inclined to see a sender’s demands as something for which it is worth making concessions when they are imposed by someone which credibly ensures restraint in its actions, respecting proportionality and avoiding ever-lasting measures. This reputation for restraint is typically the case of in-group senders which are moved by a logic of preservation of their group and its norms, not by a logic of punishment or competition. In addition, in such cases relief, in the form of diplomatic and economic reintegration, generally follows more transparent concession benchmarks and is assured as soon as such concessions are pledged. This transparency should also make these in-group measures less contested because acknowledged ex ante when joining the group and more consistently applied as all members are generally considered as one and the same. Therefore, ceteris paribus, in spite of the limited material leverage⁸ and the presence of reputational costs only (for having broken the unity of the group), the certainty of response combined with the credibility of relief at the basis of this logic should give in-group senders such as regional actors higher leverage to compel defiant targets to modify their behaviour than extra-regional actors. Importantly, lack of contestation and consequently higher leverage are valid only as long as targets are not ridiculed. Indeed, camaraderie is based on partnership and transparent benchmarks, not on paternalism. Comrades do not instruct each other, but rather gently persuade them to respect the principles they agreed on. This is all the more true in regions that went through colonialism’s practices of domination and subjugation⁹. Instructing a comrade breaks the bond that ties members and that makes each try to look good to each other.

Softness in the criteria for compliance

The camaraderie mechanism through which in-group legitimacy induces higher compliance is similar to Hellquist’s (2021) likening of AU sanctions to an academic peer review which combines ‘community-derived authority with equal treatment and constructive criticism’ (p. 452). However, it differentiates from it by making more explicit an element that is hardly compatible with peer reviewers’ deontology but certainly is with comrades, namely the idea that some wrongdoing is tolerated for the sake of togetherness and the quickest return to the group.

Indeed, the camaraderie logic may engender softness not only in the mode (the above-mentioned lack of paternalism, manifested in the absence of ridicule and the presence of mediation and dialogue), but also in the criteria for compliance with the senders’ demands. This possible side effect stems from the fact that, between brothers and sisters you are inclined to condone some misconduct – especially the less visible one. To be sure, this face-saving feature may characterize also out-group sanctions between close partners. With regard to in-group sanctions, this feature may not be universal either, but rather depend on the characteristics of the group, in this case the above-mentioned regional background of subjugation, and the more minimalist understanding of democracy in many developing regions. Most African countries themselves, for instance, have a markedly procedural definition and practice of democracy, focused on (often seemingly) multiparty elections but ignoring broader civil liberties dimensions (Mattes, 2019). This may eventually influence in a significant way the very criteria for democracy sanctions compliance. A thin, minimalist, some would even say fallacious (Diamond, 1996), sense of democracy limited to how power is obtained and not to how it is exerted subsequently such as this can certainly be associated to higher success rates than more liberal conceptualizations of democratic success can be. As a result, in-group legitimacy may well be theorized as the engine of a mechanism which can increase the likelihood of compliance, but following the prevalent local definitions thereof, which in the African context means at the cost of more superficial standards for such compliance, most likely largely criticized by extra-regional observers but which are those more widely accepted by the members of that group.

Eroding legitimacy

It is also worth noting that this mechanism needs to be sustained. While this theoretical frameworks’ main argument remains that in-group legitimacy can certainly influence regional organizations’ ability to secure higher compliance with commonly agreed democratic rules and norms, it also acknowledges the possibility of contestation over such in-group legitimacy. In-group legitimacy, just like capabilities but also other forms of non-innate legitimacy, can indeed erode over time. Africa is, in this regard, enlightening. Although an African regional organization’s in-group legitimacy may also partly build on the innate collective Pan-African identity (Witt, 2023), it also acquires it from concrete experiences and performance as a political institution. Consequently, apart from the above-mentioned contestation generated by paternalism that shatters the true principles of camaraderie, a possible major root cause of in-group legitimacy erosion lies in the lack of consistency over time and across member states in how the regional organization enforces its norms and policies in practice.

Data from local Afrobarometer’s surveys and PRIF’s interviews paint indeed a worsening picture about popular and elite support for some recent AU and RECs’ interventions than generally thought (Afrobarometer, 2023; Birchinger et al., 2023; Chacha, 2023). Recent highly visible instances of contestation of African regional organizations’ actions (by African rulers and citizens alike, such as in Mali regarding ECOWAS’ post-2020 sanctions, Niger regarding ECOWAS’ 2023 military threat and sanctions) confirm this. Such contestation targets especially RECs such as ECOWAS accused of being pawns of foreign powers, thus emulating the very extra-regional actors they tried to differentiate from, but also the AU for its lack of consistency in interventions and in the criteria for sanctions removal.

The latter can certainly also depend on the peculiar emphasis that African senders’ legal frameworks have always placed on mediation and compromise with sanctioned entities (Nathan, 2017), that can generate some inconsistencies in the duration of these sanctions policies. Yet, first and foremost, it has to do with recently growing inconsistencies in the very decision of whether to sanction a member state in the first place, substantiating charges of double standards and hypocrisy that are denting regional actors’ credibility and therefore erode their in-group legitimacy. Empirical evidence shows that inconsistencies of this kind, in the African context, with the AU and RECs failing to suspend and sanction democratic norm violators, including coup-born regimes and overstayers, intensified after 2014, when late in the year both the AU and ECOWAS failed to suspend Burkina Faso after the overthrow of President Blaise Compaoré (who himself had just tried to extend his power) and the junta’s unwillingness to transfer power to a civilian authority. This is why this article postulates that in-group legitimacy can erode over time. With regard to the African context, it argues a different (i.e. decreasing) effect after 2014 precisely due to such in-group legitimacy erosion on the continent. It also argues that this different effect distinguishes sanctions involving regional actors more than those without them, thus having not only a different, but also a differentiating effect (meaning larger sizes of decrease than the other, equally decreasing, sender categories), to be analysed with appropriate methodological specifications in the statistical tests.

Sender typology for empirical analysis

To operationalize this framework in the following empirical part, focused on Africa, I draw on this in-group versus out-group distinction of the senders’ legitimation strategies in relation to African target states, as well as on the size of them (unilateral or plurilateral), to create a typology of sender composition as follows (Table 1): individual out-group (i.e. sanctions by a single extra-regional entity only, such as the US or the EU, with no multiple-actors coalition); plurilateral out-group (i.e. multiple out-group senders such as combined sanctions regimes by the US and EU, irrespective of their coordination, but as long as they share the same overall objective against the target); unilateral in-group (i.e. imposed by a single African state, but short of any regional organization’s endorsement); plurilateral in-group (i.e. sanctions imposed by a regional organization to which the target is member of); plurilateral mixed (i.e. combined sanctions regimes imposed by multiple senders including both in-group and out-group). Because of the automatic clauses triggering at least the suspension of the violators from regional activities present in the AU’s legal frameworks on UCGs, in most of these mixed plurilateral regimes in which extra-regional senders join local ones, the in-group African regional organizations are the ‘first movers’, thus making these regimes largely African-led and shielded from accusations of out-group illegitimacy¹⁰. Although never imposed on African targets for primarily democracy aims, UN sanctions would be considered like plurilateral out-group, given that they are decided by a body, the UNSC, which the target is not necessarily member to.

Table 1.

Typology of sender composition based on senders’ legitimacy (in- vs out-group) and size (uni- vs pluri-lateral).

		Sender legitimacy
		In-group	Out-group	Mixed
Sender size/capacity	Unilateral	i.e. by one single African state (albeit not primarily for democracy aims, Eritrea’s sanctions on Ethiopia in 2001–2018, after the end of the UN regime)	i.e. by one single extra-regional power (US’ sanctions on Cameroon 2019–ongoing)	n/a by definition
Sender size/capacity	Plurilateral	i.e. by regional organizations only, against their members (AU’s Comoros 2007–08)	i.e. combined sanctions regime by extra-regional senders such as US and EU (Zimbabwe 2002–ongoing)	i.e. combined sanctions regimes involving both African and extra-regional powers (Guinea 2009)

In post-1990 Africa, in-group sanctions have almost invariably involved one or more extra-regional actors, eventually creating plurilateral mixed regimes¹¹. As a result, in the empirical analysis, it is not possible to statistically compare such plurilateral Africa-strong regimes with African-only regimes due to lack of data for the latter. Yet, it is possible to compare the former with both unilateral and plurilateral extra-regional sanctions regimes which do not involve regional support, therefore still testing whether – ceteris paribus – the inclusion of local actors increases the probability of democratic concessions. In such tests, if everything else, especially the size and capacity of the sender front, is held equal, a change in the probability of success then may well be the result of the inclusion of in-group actors.

Hypotheses

This article’s legitimacy-based perspective challenges traditional materialist accounts according to which senders with stronger economic and military linkages with targets have greater leverage on them and can therefore impose more severe measures. According to this capacity-based perspective, extra-regional actors’ initiatives are expected to have higher effectiveness than regional ones. This is due to the fact that the largest economic and security partners of most developing states are notoriously not regional but extra-regional ones (see Mold, 2022 for the case of Africa), so that there is little that regional neighbours can do on their own to exert material pressure and enforce coercive policies on targets through economic or financial restrictions. But this perspective fails to explain the existing variation in the effectiveness of highly capable extra-regional initiatives, as they are not always successful in obtaining the targets’ compliance, even when enforcing similar economic pressure. The legitimacy-based perspective suggests non-material factors play a role in this, through the inclusion of regional actors in the sender front. For this reason, the empirical section will investigate the legitimacy-based mechanism in opposition to the traditional capacity-based one, with the aim of understanding which better explains the success of sanctions-based democracy promotion.

H1 (in-group legitimacy). The probability of democratic sanctions success in regions targeted by external democracy promotion increases when regional organizations are involved in a sanction regime, not when the severity of the imposed restrictive measures increases.

However, drawing from the above-mentioned evidence from Africa, the aforesaid risk of eroding legitimacy of African regional organizations implies that effectiveness based on in-group legitimacy can decline. As long as Africa is concerned, we should expect a different and differentiating effect of such sender composition’s legitimacy-based mechanism on sanctions success over the past decade (i.e. post-2014) compared to the period before (pre-2014). Certainly, the rise of global authoritarian powers since 2014 has created an environment conducive to the resilience of non-democratic principles that may play a role in reducing the likelihood of post-2014 external democracy promotion across all the senders’ spectrum in any region¹². But because, as long as Africa is concerned, post-2014 growing inconsistencies emerged specifically from regional actors, the largest post-2014 decrease in the likelihood of democratic compliance is expected for the sender category inclusive of African regional organizations.

H2 (legitimacy erosion). In the African context, the diminishing legitimacy of regional organizations over the past decade has lessened such effects, especially for regionally led democratic sanctions.

What ‘democratic concessions’ means empirically, though, requires a more elaborated strategy. Against the above-mentioned caution for softer compliance criteria among in-group comrades, but also regarding some out-group partners, it is useful to operationalize softer and tighter criteria for sanctions success, with two more empirically specific yet different dependent variables. By identifying political rights improvement as a proxy for softer compliance and civil liberties improvement as tighter compliance, I further hypothesize that those sanctions regimes in which regional organizations are involved are more successful than other regimes in creating or restoring a political environment which is formally plural and based on free and fair electoral processes (H3a), but are unlikely to have a substantive effect beyond this, thus failing to improve also more robust, liberal democratic principles related to a wider array of civil liberties (H3b), because of the camaraderie’s face-saving logic and a different – namely softer, minimalist, in the African context – regional conceptualization of democracy.

H3 (soft face-saving compliance)

H3a. Regionally led democratic sanctions increase the probability that electoral processes, political pluralism, and participation improve after their imposition.

H3b. Regionally led democratic sanctions do not effectively increase the probability of improving a wider array of civil liberties.

Research design

Given the absence of a global data set ready to use for such empirical analyses, I assembled an original data set which focuses on Africa and operationalises the variables listed below on the basis of a series of existing data collections and repositories such as the Global Sanctions Data Base (GSDB, Kirilakha et al., 2021), NELDA, SIPRI, UCDP/PRIO, UNCTAD, World Bank indicators, V-Dem and Freedom House, as referenced below. The hypotheses are therefore tested on all the sanction regimes imposed against African targets in 1990–2021, with logistic regressions in which the unit of analysis is the African country-year. The primary sample consists of all African country-years in the period 1990–2021 in which imposed democracy sanctions were present. A secondary sample, used to test additional model specifications that account for possible selection bias (Heckman Probit models) and thus take into account factors that may influence the very imposition of democratic sanctions in Africa¹³, consists of all African country-years in the period 1990–2021.

Operationalization of the dependent variables

For each country-year, it was dichotomously recorded whether a given African country was under democracy sanctions and, if it was, whether those imposed democratic sanctions achieved their aims in that given year. They are coded as such if in the Global Sanctions Data Base (GSDB, Kirilakha et al., 2021), from which this article’s sanctions-related variables are sourced, they were recorded as total success, partial success or negotiated settlement. This is a broad operationalization that reflects mediation-induced restorations of constitutional orders typical of African regional organizations’ sanctions doctrine. Because the theoretical framework also expected that some sender categories could have overall softer criteria for compliance due to in-group face-saving logics and different conceptualizations of democracy and, consequently, of democracy sanctions success, two alternative and more specific dependent variables are created. In this way, they capture more fine-grained aspects of the change in the democratic status of sanctioned targets, namely contrasting whether political rights (taken as a proxy for soft criteria for compliance) and civil liberties (tighter criteria) have improved the subsequent year (Y + 1) after the one in which sanctions were in place (Y)¹⁴. Given the GSDB data’s inappropriate information for this empirical application, variables PR change and CL change are both drawn from Freedom House (2022) and operationalized dichotomously to code the presence of a subsequent yearly improvement respectively in the Political Rights and Civil Liberties rankings. These two rankings fit best on theoretical and empirical grounds the aim of this research to separately capture, with discrete values, sizable changes in both political rights and civil liberties dimensions¹⁵.

Operationalization of independent variables

Determinants of democracy sanctions success

To test whether the inclusion of African regional organizations in the senders’ front has an effective impact on success, first I created a variable African involvement which is dichotomously coded as 1 when, in a given country-year, imposed democracy sanctions are upheld also by an African regional organization (I coded senders as EU, US, AU, ECOWAS, SADC, UN, other external actors, other African actors¹⁶). In all the other cases, it is coded as 0. Then, to better examine the effect this African involvement generates in comparison with other sanctions regimes in which African organizations were not involved, I created a categorical variable (Sender composition) which distinguishes cases of unilateral extra-regional democracy sanctions (coded as 0), from plurilateral extra-regional democracy sanctions (coded as 1, namely sanctions by more than one extra-regional actor, yet without the involvement of African actors), from sanctions which involved at least one African organization (coded as 2). Because African-only sanctions regimes (i.e. AU’s Comoros 2007–08) are too rare to be coded as an independent category, they cannot be included as a stand-alone category in the statistical analysis. Therefore, all the cases characterized by the involvement of an African regional organization in the empirical analysis are de facto ‘mixed plurilateral’.

To test the theorized laxness of African regional organizations’ sanctions enforcement and rising contestation of their in-group legitimacy on the continent after 2014, a post-2014 dichotomous variable is added to all the models. In some, it is used only as a control variable, while in others it is used in interaction with the sender composition variables to test also the existence of the above-mentioned differentiating effect across the diverse sender categories.

The other independent variables included in the models are those traditionally used in the literature of sanctions and coercive democracy promotion as plausible explanations of effectiveness. Given the absence of a GSDB variable estimating sanctions costs, a dichotomous variable Sanctions severity estimates the severity of the measures imposed by coding whether they consisted of coexisting trade and financial restrictions, which the literature generally expects to be the most severe and comprehensive (Bierstecker et al., 2016). The Sanctions heterogeneity of the measures imposed, instead, counts the different restrictive measures listed in the GSDB and accounts for the counterproductive effects that the inappropriateness of an excessive number of dimensions addressed in a sanctions regime (diplomatic, economic, financial or military) may generate, in the form of rally-round-the-flag effects benefitting the targeted regime which plays the victimization card to boost domestic political support. The natural log of the GDP per capita PPP (World Development Indicators (WDI), 2022) is included to control for the wealth of the target. The percentage of the target’s military expenditure on the GDP (SIPRI, 2022) captures the military’s loyalty to the regime during sanctions. The percentage of the target’s total trade with all the (Western or African) sender(s), sourced from UNCTAD (2022, data covering only the years 1995–2021), on its own GDP (sourced from the WDI) measures the potential economic cost sanctions can have on the target, given that the literature expects that the stronger the linkage between the target and the senders, the more vulnerable the target to their pressure. In addition, political ties with extra-regional major powers are measured through the alignment of their votes at the UN General Assembly (UNGA) (Voeten, 2013): therefore, the political closeness variable dichotomously codes whether there is strong alignment in UNGA votes (> .8 in a 0–1 scale) between the target country and the two major non-democratic powers, China and Russia (an average of the two is calculated), accounting for the political protection targets could receive from them from punitive measures against democratic norm violations. On the contrary, the democratic status of the target’s sub-region, calculated as the average V-Dem polyarchy value of the African sub-region’s member states (Coppedge et al., 2022), is included to check whether, through diffusion mechanisms, more democratic neighbours facilitate the target’s acquisition of democratic values. In the models, all these factors are lagged by 1 year to ensure temporal precedence to sanctions’ effects, except for variables related to features of the sanction policy itself, namely sender composition, sanctions severity and heterogeneity. Finally, as both can increase the domestic vulnerability of the target, the models also control for the existence of ongoing political instability (operationalized with a dichotomous conflict variable drawn from UCDP/PRIO armed conflict data set – cf. Davies et al., 2022; Gleditsch et al., 2002) and for the presence of a concurrent non-democratic sanctions regime against the country targeted by democracy sanctions.

Triggers of democracy sanctions

To take into consideration possible selection effects (Van Bergeijk and Siddiquee, 2016), additional models also include an imposition equation made up of factors which can influence the very imposition of democracy sanctions in Africa. These factors are drawn from the expectations generated by the review of the legal frameworks above and the literature. This equation therefore includes dichotomous variables capturing the occurrence, in a given country-year, of successful coups (Powell and Thyne’s, 2011 coup data), controversial elections (elections in which the opposition was harassed by the government as in variable 15 of the NELDA data set – Hyde and Marinov, 2012), and a variable for less visible triggers such as the year-on-year decrease in an African country’s civil liberties score (Freedom House, 2022). The lagged variable of the imposition of non-democracy sanctions is also included to account for the role that existing non-democracy sanctions may have in facilitating the imposition of democracy sanctions. The imposition of sanctions could further be facilitated or constrained by the vulnerability of the target and the costs that breaking such economic ties with a potential target may generate on the most exposed sender, therefore some of the controls used in the outcome equation are also used here such as the presence of conflict in the target state but also inflation, the natural log of the GDP per capita (PPP) of the target, all sourced from the WDI (2022) data set, together with a variable estimating the weight of the trade ties between the target and the largest potential Western democratic sender on the latter’s GDP (reported in basis points).

Empirical results and discussion

Table 2 presents the results of the quantitative analysis of the determinants of democracy sanctions success in Africa 1990–2021, across different model specifications. Depending on the model, the sample of cases refers either only to the African country-years in which democracy sanctions have been imposed (261 observations, reduced to 174 in models with full controls due to some missing data from the sources, especially those related to pre-1995 trade data, used to compile this data set) or to the entire sample of 1782 country-years across 54 African countries (reduced to 1269 observations after adding controls).

Table 2.

Determinants of democratic sanctions success in Africa.

DV:	1a	1b	2a		2b		2b ⁺ (legitimacy)		2b^* (severity)
Sanctions Success	Logit	Logit	Heckman		Heckman		Heckman Interaction		Heckman Interaction
			Outcome stage		Outcome stage		Outcome stage		Outcome stage
African involvement	1.383*** (0.350)		0.772*** (0.185)
Sender composition (ref. cat.: sanctions with AROs)
• Extra-regional unilateral		-2.595*** (0.809)			-1.397*** (0.375)		-1.611*** (0.395)		-1.734*** (0.454)
• Extra-regional plurilateral		-1.314*** (0.387)			-0.755*** (0.205)		-0.958*** (0.237)		-0.547** (0.261)
Sender composition ✕ post 2014
• Extra-regional unilateral							5.671*** (0.982)
• Extra-regional plurilateral							5.341*** (0.422)
Sanctions severity	0.463(0.664)	0.344(0.677)	0.212(0.342)		0.159(0.346)		-0.00674(0.425)		0.453(0.493)
Sender composition ✕ severity
• Extra-regional unilateral									1.802** (0.883)
• Extra-regional plurilateral									-0.992(0.650)
Sanctions heterogeneity	-0.323(0.263)	-0.445* (0.268)	-0.173(0.138)		-0.238* (0.133)		-0.191(0.138)		-0.260* (0.133)
Political closeness China-Russia _t-1	-1.398** (0.623)	-1.341** (0.591)	-0.811** (0.351)		-0.783** (0.335)		-0.831** (0.324)		-0.705** (0.324)
GDP per capita _{PPP (thous.) (log) t-1}	1.035*** (0.359)	1.002*** (0.305)	0.583*** (0.179)		0.586*** (0.148)		0.557*** (0.142)		0.708*** (0.146)
Military expenditure _{(% GDP) t-1}	-0.538*** (0.162)	-0.576*** (0.149)	-0.321*** (0.0944)		-0.337*** (0.0840)		-0.292*** (0.0881)		-0.314*** (0.0917)
Bilateral trade flows _{(% on target GDP) t-1}	0.0162(0.0362)	0.0197(0.0304)	0.0121(0.0189)		0.0120(0.0160)		0.0120(0.0156)		0.0112(0.0155)
Neighbours’ democracy levels _t-1	4.947(3.168)	3.564(2.664)	2.639(1.656)		1.822(1.455)		2.019(1.325)		0.792(1.581)
Concurrent non-democratic sanctions	-0.277(0.538)	-0.283(0.453)	-0.121(0.283)		-0.147(0.253)		-0.160(0.239)		-0.236(0.235)
Ongoing violent instability	1.337** (0.597)	1.533** (0.673)	0.776** (0.341)		0.844** (0.368)		0.724** (0.368)		0.773* (0.409)
Post 2014	-2.054** (0.856)	-2.045*** (0.781)	-1.136*** (0.427)		-1.117*** (0.408)		-6.033*** (0.452)		-0.925** (0.447)
				Imposition stage		Imposition stage		Imposition stage		Imposition stage
Coups				1.674*** (0.248)		1.677*** (0.248)		1.675*** (0.254)		1.679*** (0.245)
Controversial elections				0.611*** (0.151)		0.617*** (0.150)		0.616*** (0.150)		0.618*** (0.151)
Worsening civil liberties				0.205(0.173)		0.211(0.173)		0.213(0.175)		-0.207(0.173)
Non-democratic sanctions _t-1				0.692*** (0.238)		0.699*** (0.239)		0.700*** (0.239)		0.698*** (0.239)
Inflation _t-1				0.00748*** (0.00191)		0.00743*** (0.00188)		0.00743*** (0.00188)		0.00748*** (0.00192)
Bilateral trade _{(% GDP top sender, bps) t-1}				-0.0279(0.0186)		-0.0277(0.0186)		-0.0277(0.0186)		-0.0278(0.0187)
Ongoing violent instability				0.353* (0.206)		0.358* (0.206)		0.358* (0.206)		0.359* (0.206)
Constant	-2.127* (1.176)	0.280(0.989)	-1.280(0.803)	-1.593*** (0.197)	0.155(0.739)	-1.603*** (0.199)	0.175(0.714)	-1.604*** (0.199)	0.415(0.838)	-1.603*** (0.199)
Observations	174	173	1269	1269	1268	1268	1268	1268	1268	1268
Pseudo R ²	0.1828	0.2007

***

p < 0.01, **p < 0.05, *p < 0.1. Robust errors clustered at country level in parentheses. AROs = African Regional Organizations.

First, two simple logit models with clustered standard errors at the African country level are used (Models 1a and 1b). The difference between Models 1a and 1b lies in the operationalization of the involvement of African regional organizations, as in Model 1a the dichotomous variable African involvement is used, while Model 1b uses the categorical Sender composition. The robustness of each model is also confirmed with the inclusion of the cubic polynomials of the years passed since the onset of the latest sanction episode, in order to account for possible temporal dependence, and with the use of a multilevel logistic regression with country fixed effects. To take into account possible selection effects, these analyses are also tested with two two-step Heckman Probit models (Models 2a-2b). A Heckman probit model consists of a selection and outcome equation. In this empirical application, the former is set to focus on the factors that may influence the very imposition of sanctions and so is made up of the set of variables described in the previous section as plausible triggers or facilitators of democracy sanctions imposition in Africa. The dependent variable of this equation is precisely the effective imposition of democracy sanctions in a country-year. The outcome equation, instead, focuses on sanctions success so is tested on cases in which the dependent variable of the selection equation was present (1 instead of 0), hence those country-years in which sanctions were imposed, using the same determinants as in Models 1a-1b. Differently from those models, however, Models 2a-2b’s more elaborated structure also accounts for the possible effect the variables contained in the selection equation may have also on the outcome stage.

The results show that, across all these model specifications, the involvement of African regional organizations increases the likelihood of sanctions success compared to cases in which democracy sanctions are imposed without the co-participation of African organizations. Model 1a shows the involvement of African organizations increases the probability of sanctions success compared to all the cases with no such involvement. The more fine-grained Model 1b – which has sanctions regimes inclusive of African regional organizations as the hidden reference category – further specifies this effect is significant when these regimes are compared both to unilateral and plurilateral extra-regional sanctions cases in which African regional organizations are not involved¹⁷. When selection effects are taken into consideration in Heckman Probit models (2a and 2b), it is noted that the occurrence of highly visible democratic violations as well as the target’s economic vulnerability and the existence of previous non-democratic sanctions regimes, affect democratic sanctions imposition but not the final effect of sender composition on sanctions success, which remains statistically significant. As a result, empirical evidence strongly confirms H1. The significance of Models 1a to 2b’s post-2014 variable, however, also confirms that, over the past few years, the mechanism of in-group-legitimacy has fallen into disrepair. This is all the more clear in Model 2b⁺, testing not only the presence of a post-2014’s lessening effect, but also a differentiating impact across the diverse sender categories. The interaction between the sender categories and the post-2014 variable is indeed significant and the marginal effects displayed in Figure 1 show the sender category that suffered from the largest decrease in the probability of success after 2014 is the one involving local regional organizations. While the plurilateral extra-regional category also has a negative effect, its magnitude is much smaller than the other. On the contrary, unilateral extra-regional sanctions show no statistically different effects contrasting pre- and post-2014 evidence.

Figure 1.

Average marginal effects (95% CIs) on the likelihood of democracy sanctions success of one-unit change from pre- to post-2014, across all the sender categories as in Model 2b⁺.

Whether severity has played a role in this needs to be checked in two stages. First, a t-test explores whether sanctions severity has remained constant across time (before and after 2014) and sender categories (with or without regional organizations). The results show indeed there is no statistical difference ((Pr(|T| > |t|)) is greater than 0.05 in both cases). That is, the continuous presence of highly capable extra-regional actors in supporting and enforcing these measures ensures also African-led sanctions have remained overall severe. Then, to further check its effect on the probability of sanctions success, besides its use as a control, Model 2b^* also includes an interaction between severity and the sender categories, to verify possible differentiated effects with this variable as well. However, as Model 2b^* and Figure 2 show, severity does not contribute to explaining democracy sanctions success and its variation across the different sender categories.

Figure 2.

Average marginal effects (95% CIs) on the likelihood of democracy sanctions success of one-unit change from less to high severity, across all the sender categories as in Model 2b^*.

Therefore, the overall empirical analysis in Table 2 suggests that it is indeed the sender category, and more specifically the inclusion of in-group senders, that increases the probability of democracy sanctions success. Yet, because sanctions success as coded in GSDB remains ambiguous as to its effective empirical achievements in terms of democracy promotion and does not allow to clarify whether softer demands and compliance criteria are in place, in Table 3 the two alternative and more specific dependent variables are used in place of the GSDB sanctions success measure to test hypotheses H3a-b. The same model specifications of Table 2 are used, except for the dependent variables. The results confirm the in-group legitimacy advantage by showing that, as long as the procedural, minimalist compliance was concerned (Models 3a-3b), sanctions involving regional actors do have a higher probability of sanctions success than other sender categories (in particular extra-regional plurilateral¹⁸). Yet, it also confirms the softness hypotheses by suggesting that, although African-mediated sanctions regimes are more likely than other sanctions regimes to ensure political rights improvement, they are no more likely to increase the likelihood of civil liberties improvement (Models 4a-4b). To be sure, the model also suggests the latter is not systematically different from any other sender category. Therefore, overall, the table proves once again that, ceteris paribus, the inclusion of local actors (in-group legitimacy) increases the probability of fundamental democratic concessions, at least until 2014, but does not in any way alter the achievement of deeper, liberal values, which remains similar across all the sender categories. The time clarification is required because Table 3 also substantiates the same post-2014 effect emerged in Table 2, suggesting that an erosion in local actors’ legitimacy has cancelled their comparative advantage from 2014 onwards, even when softer compliance criteria are concerned.

Table 3.

Distinguishing softer (political rights improvement) and tighter (civil liberties improvement) criteria for sanctions success.

DV:	3a	3b	4a	4b
PR improvement _(Y ₊ ₁₎ and CL improvement _(Y ₊ ₁₎	Heckman PR	Heckman PR	Heckman CL	Heckman CL
Sender composition (ref. cat.: sanctions with AROs)
• Extra-regional unilateral	0.142(0.569)	-0.133(0.537)	0.442(0.342)	0.425(0.282)
• Extra-regional plurilateral	-0.750** (0.332)	-0.983*** (0.328)	-0.139(0.288)	-0.157(0.325)
Sender composition ✕ post 2014
• Extra-regional unilateral		6.297*** (0.743)		0.0647(0.861)
• Extra-regional plurilateral		6.234*** (0.651)		0.0872(0.881)
Post 2014	-1.120*** (0.420)	-6.720*** (0.637)	-0.894*** (0.317)	-0.951(0.749)
Constant	0.258(0.886)	0.317(0.829)	-1.790(1.109)	-1.778*(1.011)
Observations	1268	1268	1268	1268

***

p < 0.01, **p < 0.05, *p < 0.1. The table only displays the variables of interest with, in parentheses, clustered errors at the target country. Full models with the controls are available in the data repository. Models 3a and 4a are Heckman probit models replicating Table 2’s Model 2b, whereas Models 3b and 4b replicate Table 2’s Model 2b⁺, with the interaction with post-2014. PR = political rights; CL = civil liberties; ARO = African regional organizations.

A more complete image in this regard emerges from Figure 3, displaying the predictive margins of the three different sender composition categories for the Heckman Probit models 2b+; 3b; 4b, distinguishing the pre- and post-2014 effects. It confirms that the probability of democratic sanctions success decreases after 2014 across any sender category and model specification, but more markedly for sanctions involving regional actors. For the latter, after 2014, their comparative advantage has disappeared. Before 2014, instead, this sender category enjoyed a statistically higher likelihood of success than any other, except for civil liberties improvements.

Figure 3.

Predictive margins (90–95% CIs) for Sender composition categories in Heckman Probit Models 2b⁺,-3b, 4b (varying only in the operationalization of the dependent variable), before and after 2014.

The post-2014 effects for this category are not surprising in light of the numerous inconsistencies African regional organizations displayed over the past few years, such as not sanctioning UCGS in Burkina Faso (2014), Zimbabwe (2017), and Chad (2021) and not sanctioning unconstitutional preservations of power such as third-term violations and election rigging that still happen across the continent¹⁹. Differently from the past, some of the very democratic violations of recent years, including military coups, have been portrayed by their perpetrators precisely as supposed corrective measures to such regional actors’ growing inconsistencies. Empirical evidence has shown that these violators eventually end up resorting to the same practices they criticized and promised to correct, starting from the rigging of the elections in favour of their preferred candidate to eventually entrench themselves in power (Elischer and Hoyle, 2023), but their anti-regional narrative is nonetheless telling of how much the legitimacy of those regional actors has eroded. Without consistent support for such democratic norms and application of their related policies (i.e. about the length of the supposed post-UCG transition or the very decision to suspend UCG violators, including incumbents who try to overstay in power, manipulating constitutions and elections against the spirit if not the very letter of AU legal frameworks), these organizations expose themselves to the same accusations that local critics move against out-group senders, namely being not only disproportionate but also inappropriate (Tournons la page, 2023), to the detriment of the in-group legitimacy advantage that had, until not long ago, ensured higher compliance from African targets.

A case which helps illustrate the comparative advantage of a sender front inclusive of regional actors, at least before their post-2014 declining legitimacy in the eyes of African targets, is Togo. The failure of the mid-1990s EU’s unilateral attempt to promote democratic reforms and human rights in the country, following President Gnassingbé Eyadéma’s repression of opposition political parties, contrasted starkly with the results achieved in 2005 by a front of local and extra-regional actors. They jointly intervened to promote the restoration of an elected government when, that year, after the sudden death of Eyadéma, his son Faure sidelined the designated successor who was temporarily abroad and unconstitutionally took power. ECOWAS was the first mover and soon after the AU also imposed one of its first sanctions regimes ever. For the first time, African neighbours displayed ‘a show of unity’ and such ‘unrelenting’ pressure and mediation that they finally directed towards the Togolese regime eventually managed to isolate Faure, ‘forcing [his] resignation’ (Duodu, 2005). Lacking support for this move within and outside the region, he was convinced to step back and let the Vice-President of the Assembly lead the country to new elections later in spring, so as to reconquer access to his community (‘I sincerely hope that the West African community (. . .) remains at our side to guarantee the organization of free, transparent and honest elections’ – Reuters, 2005). Faure eventually won the election (as regional actors did not prevent him from running back then – Nathan, 2017), but, on a more positive note, such pressure helped move the country to national reconciliation and to a partially freer status in the coming years, according to the major democracy indices.

Similar external pressure from regional and extra-regional actors also worked against the Guinean coup plotters in 2009. Guinea had also been targeted by unilateral EU sanctions years earlier, successfully so – a reminder that not all the extra-regional unilateral measures eventually fail. However, the Guinean case is illustrative also of the past few years’ declining legitimacy of African regional organizations (Togo, instead, did not register any democracy sanction in 2014–2021, when the data set ends). When the AU and ECOWAS, together with other extra-regional actors, imposed sanctions in reaction to the September 2021 coup that removed President Alpha Conde, the regional actors were highly contested for the double standards displayed just six months earlier, when the AU failed to sanction Chad’s constitutional coup (ECOWAS had no jurisdiction on that case). Unsurprisingly, they have since failed in negotiating with the violators both the scheduling of new elections and the creation of a civilian transitional government.

Conclusion

This article aimed to contribute to ongoing research on the effectiveness of promoting and upholding democratic governance worldwide through coercive measures such as sanctions, focusing specifically on the role of sender composition. Given material-based explanations of sanctions-based external democracy promotion do not fully explain the existing variation in the effectiveness of highly capable extra-regional initiatives, this research aimed to examine the merits of an alternative, legitimacy-based explanation that animates not only the sanctions literature but also neighbouring fields such as those related to conflict resolution. It did so by focusing on democracy sanctions in Africa, the most-sanctioned world region, which is also characterized by a varied front of sanctioners.

The theoretical framework proposed an in-group legitimacy-based explanation for compliance with democracy sanctions, through a camaraderie mechanism that leverages softness in modes and often in compliance criteria too to induce a change of behaviour in the target. The subsequent empirical analysis of democracy sanctions in Africa over the years 1990–2021 showed the involvement of African regional organizations – which are heavily based on such in-group legitimacy rather than capacity – is fundamental to increase the likelihood of democracy sanctions success in the region compared to both unilateral and plurilateral sanctions imposed on African targets without the involvement of African institutions. However, the empirical results also confirmed that this is true only when using softer criteria of compliance. Moreover, the results show that these effects have decreased since 2014 across all the sender categories but mostly for those involving African regional organizations, due to still ongoing processes of in-group legitimacy erosion on the continent following their violations of the most basic principles of the in-group camaraderie logic.

These findings have global and regional implications. On one hand, for extra-regional actors active in democracy promotion worldwide these findings mean that, instead of imposing long, comprehensive and ineffective democracy sanction regimes that over time may also turn counterproductive (Meissner and Mello, 2022), they would benefit more from coordinating with regional actors. On the other hand, it is important to this end, though, that local actors maintain, or rehabilitate if lost, their legitimacy and do not emulate extra-regional actors’ strategies, to preserve themselves from (often unsubstantiated) claims of being pawns of foreign powers, which would otherwise obliterate their in-group legitimacy comparative advantage. As long as Africa is concerned, the paternalistic narratives used after the 2023 coup in Niger and the related threat of military intervention certainly disavowed the traditional camaraderie approach and eventually failed. Albeit remarkable advances were achieved in just 20 years, instead of ‘indulg[ing] [. . .] in blissful self-satisfaction’ (Faki, 2024), African regional organizations need redress the growing inconsistencies displayed in their democracy promotion initiatives and avoid the double standards that since 2014 have eroded their legitimacy to recover their comparative advantage. The status of democracy on the continent largely depends on that.

Footnotes

Acknowledgements

Earlier versions of this article were discussed or presented at the 35th Annual Conference of the Italian Political Science Association (Sapienza University of Rome, 2022), at the General Conference of the European Consortium for Political Research (University of Innsbruck, 2022), at the Institute for Peace and Security Studies (Addis Ababa University, 2023), at the 36th Annual Conference of the Italian Political Science Association (University of Genoa, 2023). I thank all the participants, and especially Yuleng Zeng, Clara Portela, Katharina Meissner, Kerim Can Kavakli, Francesco Giumelli, Laura Berlingozzi and Jesutimilehin Akamo, for their helpful comments and suggestions. I also thank the anonymous reviewers and the journal editors for their constructive comments and feedback, which were very helpful in improving the article.

Data availability statement

I will upload my data set and replication file in a public data repository in case of publication

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Tiziana Corda

Notes

Author biography

Tiziana Corda is a PostDoctoral Research Fellow at the Department of Social and Political Sciences of the University of Milan. Her research focuses on authoritarianism and the comparative analysis of political regimes; sanctions and other forms of foreign pressure; regional organizations and extra-regional powers in Africa. Her most recent research articles appeared on Democratization and Defence and Peace Economics.

References

Afrobarometer (2023) Afrobarometer data. Available at: http://www.afrobarometer.org

Amani (2023) Beyond subsidiarity: understanding the roles of the AU and RECs/RMs in peace and security in Africa. Special Research Report Amani Africa. Available at: https://amaniafrica-et.org/beyond-subsidiarity-understanding-the-roles-of-the-au-and-recs-rms-in-peace-and-security-in-africa/

Attia

Grauvogel

(2023) International sanctions termination, 1990–2018: Introducing the IST dataset. Journal of Peace Research 60(4): 709–719.

Bapat

Morgan

(2009) Multilateral versus unilateral sanctions reconsidered: A test using new data. International Studies Quarterly 53(4): 1075–1094.

Bierstecker

Eckert

Tourinho

(2016) Targeted Sanctions. the Impacts and Effectiveness of United Nations Action. Cambridge: Cambridge University Press.

Birchinger

Jaw

Bah

, et al (2023) ‘Siding with the people’ or ‘Occupying force’? Local perceptions of African Union and ECOWAS interventions in the Gambia. PRIF, Report 3/2023. Frankfurt: PRIF. Available at: https://www.prif.org/publikationen/publikationssuche/publikation/siding-with-the-people-or-occupying-force-local-perceptions-of-african-union-and-ecowas-interventions-in-the-gambia

Boese

(2019) How (not) to measure democracy. International Area Studies Review 22(2): 95–127.

Cebul

Dafoe

Monteiro

(2021) Coercion and the credibility of assurances. Journal of Politics 83(3): 975–991.

Chacha

(2023) Public attitudes toward external democracy promotion in Africa. Democratization 30: 1552–1581.

10.

Charron

Portela

(2015) The UN, regional sanctions and Africa. International Affairs 91(6): 1369–1385.

11.

Coppedge

Gerring

Knutsen

, et al (2022) V-Dem Codebook v12. Varieties of Democracy. Available at: https://www.v-dem.net/static/website/img/refs/codebookv12.pdf

12.

Corda

(2024) Democratic sanctions in Africa. A reassessment of their selective and inconsistent use across multiple senders. In: Bultrini

Giumelli

Portela

, et al (eds) International Sanctions in Practice. London: Routledge, pp. 161–178.

13.

Cowell

(2011) The impact of the ECOWAS protocol on good governance and democracy. African Journal of International and Comparative Law 19(2): 331–342.

14.

Davies

Pettersson

Öberg

(2022) Organized violence 1989-2021 and drone warfare. Journal of Peace Research 59(4): 593–610.

15.

Diamond

(1996) Is the third wave over? Journal of Democracy 7(3): 22.

16.

Djinnit

(2021) The case for updating the African Union policy on unconstitutional changes of government. ACCORD Policy and Practice, Brief No 54. Available at: https://www.accord.org.za/publication/case-for-updating-au-policy-unconstitutional-changes-government/

17.

Drezner

(2000) Bargaining, enforcement, and multilateral sanctions: When is cooperation counterproductive? International Organization 54(1): 73–102.

18.

Duodu

(2005) Togo’s new dictator is forced to step down. The Guardian. Available at: https://www.theguardian.com/world/2005/feb/27/theobserver1

19.

Duursma

(2020) African solutions to African challenges: The role of legitimacy in mediating civil wars in Africa. International Organization 74(2): 295–330.

20.

Elischer

Hoyle

(2023) Electoral contests in the aftermath of military coups: how domestic constraints motivate praetorian conduct. Contemporary Politics 30: 44–64.

21.

Eriksson

(2010) Supporting Democracy in Africa – African Union’s Use of Targeted Sanctions to Deal with Unconstitutional Changes of Government. Stockholm: FOI, Swedish Defence Research Agency.

22.

Escribà-Folch

Wright

(2015) Foreign Pressure and the Politics of Autocratic Survival. Oxford: Oxford University Press.

23.

Faki

(2024) Strategic remarks of the chairperson. African Union Speeches. Available at: https://au.int/en/speeches/remarks-ceremony-high-level-20th-anniversary-colloquium-au-psc

24.

Felbermayr

Kirilakha

Syropoulos

, et al (2020) The global sanctions data base. European Economic Review 129(C): 103561.

25.

Franck

(1988) Legitimacy in the international system. The American Journal of International Law 82(4): 705–759.

26.

Freedom House (2022) Freedom in the World Data. Available at: https://freedomhouse.org/reports/freedom-world/freedom-world-research-methodology

27.

Galtung

(1967) On the effects of international economic sanctions, with examples from the case of Rhodesia. World Politics 19(3): 378–416.

28.

Gerschewski

(2018) Legitimacy in autocracies: Oxymoron or essential feature. Perspectives on Politics 16(3): 652–665.

29.

Gleditsch

Petter

Wallensteen

, et al (2002) Armed conflict 1946–2001: A new dataset. Journal of Peace Research 39(5): 615–637.

30.

Grauvogel

von Soest

(2014) Claims to legitimacy count. Why sanctions fail to instigate democratisation in authoritarian regimes. European Journal of Political Research 53(4): 635–653.

31.

Hellquist

(2021) Regional sanctions as peer review: The African Union against Egypt (2013) and Sudan (2019). International Political Science Review 42(4): 451–468.

32.

Hellquist

(2022) The African union in a world of sanctions. In: Charron

Portela

(eds) Multilateral Sanctions Revisited. Montreal, QC, Canada: McGill-Queen’s University Press, pp. 99–117.

33.

Hellquist

Palestini

(2021) Regional sanctions and the struggle for democracy: Introduction to the special issue. International Political Science Review 42(4): 437–450.

34.

Hohlstein

(2022) Regional Organizations and their Responses to Coups. Bristol: Bristol University Press.

35.

Hufbauer

Schott

Elliott

, et al (2007[1990]) Economic Sanctions Reconsidered (3rd edn). Washington, DC: Peterson Institute for International Economics.

36.

Hyde

(2023) International actors and domestic politics in an era of global challenges to democracy," in Bartels

et al (eds), The Forum: Global Challenges to Democracy? Perspectives on Democratic Backsliding. International Studies Review 25(2): 11–15.

37.

Hyde

Marinov

(2012) Which elections can be lost? Political Analysis 20(2): 191–201.

38.

Kavakli

Chatagnier

Hatipoglu

(2020) The power to hurt and the effectiveness of international sanctions. Journal of Politics 82(3): 879–894.

39.

Kirilakha

Felbermayr

Syropoulos

, et al (2021) The Global Sanctions Data Base (GSDB): an update that includes the years of the Trump presidency. In: Van Bergeijk

PAG

(ed.) Research Handbook on Economic Sanctions. Cheltenham: Edward Elgar Publishing, pp. 62–106.

40.

Leininger

(2022) International Democracy Promotion in Times of Autocratization: From Supporting to Protecting Democracy (IDOS Discussion Paper 21/2022). Bonn: German Institute of Development and Sustainability (IDOS).

41.

McLean

Whang

(2010) Friends or foes? Major trading partners and the success of economic sanctions. International Studies Quarterly 54(2): 427–447.

42.

Martin

(1992) Coercive Cooperation: Explaining Multilateral Economic Sanctions. Princeton, NJ: Princeton University Press.

43.

Mattes

(2019) Democracy in Africa: Demand, supply, and the ‘dissatisfied democrat. Afrobarometer, Policy Paper No 54. Available at: https://www.afrobarometer.org/wp-content/uploads/migrated/files/publications/Policy%20papers/ab_r7_policypaperno54_africans_views_of_democracy1.pdf

44.

Meissner

Mello

(2022) The unintended consequences of UN sanctions: A qualitative comparative analysis. Contemporary Security Policy 43(2): 243–273.

45.

Miers

Morgan

(2002) Multilateral sanctions and foreign policy success: Can too many cooks spoil the broth? International Interactions 28(2): 117–136.

46.

Mold

(2022) The economic significance of intra-African trade Getting the narrative right. Brookings Global, Working Paper 44. Available at: https://www.brookings.edu/articles/the-economic-significance-of-intra-african-trade-getting-the-narrative-right/

47.

Nathan

(2017) A survey of mediation in African coups. African Peacebuilding Network, APN Working Papers No 15. Available at: https://www.ssrc.org/publications/a-survey-of-mediation-in-african-coups/

48.

Omorogbe

(2011) A club of incumbents? The African Union and coups d’état. Vanderbilt Journal of Transnational Law 44: 123–154.

49.

Peksen

Drury

(2010) Coercive or corrosive: The negative impact of economic sanctions on democracy. International Interactions 36(3): 240–264.

50.

Pevehouse

(2002) With a little help from my friends? Regional organizations and the consolidation of democracy. American Journal of Political Science 46(3): 611–626.

51.

Piccone

(2004) International mechanisms for protecting democracy. Democracy Coalition Project, Washington, DC, p. 8.

52.

Powell

Thyne

(2011) Global instances of coups from 1950 to 2010: A new dataset. Journal of Peace Research 48(2): 249–259.

53.

Powell

Lasley

Schiel

(2016) Combating Coups d’état in Africa, 1950–2014. Studies in Comparative International Development 51(4): 482–502.

54.

Reuters (2005) Togo president to step down, 27 February. Available at: https://www.theage.com.au/world/togo-president-to-step-down-20050227-gdzogj.html

55.

Sabrow

(2017) Local perceptions of the legitimacy of peace operations by the UN, regional organizations and individual states – A case study of the Mali conflict. International Peacekeeping 24(1): 159–186.

56.

Saltnes

(2017) Norm collision in the European Union’s external policies: The case of European Union sanctions towards Rwanda. Cooperation and Conflict 52(4): 553–570.

57.

SIPRI (2022) SIPRI military expenditure database. Available at: https://www.sipri.org/databases/milex

58.

Souaré

(2014) The African Union as a norm entrepreneur on military coups d’état in Africa (1952–2012): an empirical assessment. The Journal of Modern African Studies 52(1): 69–94.

59.

Staeger

(2023) Resource mobilization in security partnerships: Explaining cooperation and coercion in the EU’s partnership with the African Union. Cooperation and Conflict 58: 502–521.

60.

Tallberg

Zürn

(2019) The legitimacy and legitimation of international organizations: introduction and framework. Review International Organization 14: 581–606.

61.

Tavares

(2010) Regional Security: The Capacity of International Organizations. London: Routledge.

62.

Tournons La Page (2023) Tournons La Page calls for dialogue and restraint in Niger. Available at: https://tournonslapage.org/fr/actualites/tournons-la-page-appelle-au-dialogue-et-a-la-retenue-au-niger

63.

Udombana

(2002) The institutional structure of the African Union: A legal analysis. California Western International Law 33: 69.

64.

UNCTAD (2022) International merchandise trade. Data Centre. Available at: https://unctadstat.unctad.org/wds/ReportFolders/reportFolders.aspx?sCS_ChosenLang=en

65.

Van Bergeijk

PAG

Siddiquee

MSH

(2016) Biased sanctions? Methodological change in economic sanctions reconsidered and its implications. International Interactions 43(5): 879–893.

66.

Voeten

(2013) Data and analyses of voting in the UN general assembly. In: Reinalda

(eds) Routledge Handbook of International Organization. London: Routledge, pp. 149.

67.

von Borzyskowski

Portela

(2018) Sanctions cooperation and regional organizations. In: Aris

Snetkov

Wenger

(eds) Inter-organizational Relations in International Security. London: Routledge, pp. 240–261.

68.

von Borzyskowski

Portela

(2023) Global governance from below: Regional sanctions as drivers of UN sanctions. Journal of Conflict Resolution 67: 1930–1958.

69.

von Soest

Wahman

(2015a) Are democratic sanctions really counterproductive? Democratization 22(6): 957–980.

70.

von Soest

Wahman

(2015b) Not all dictators are equal: Coups, fraudulent elections, and the selective targeting of democratic sanctions. Journal of Peace Research 52(1): 17–31.

71.

Wiebusch

Aniekwe

Oette

, et al (2019) The African charter on democracy, elections and governance: Trends, challenges and perspectives. Africa Spectrum 54(2): 95–105.

72.

Witt

(2023) Forging an African union identity: The power of experience. Global Studies Quarterly 3: ksad052.

73.

World Development Indicators (WDI) (2022) World Bank data. Available at: https://data.worldbank.org