Go global or go home: Comparing the regional vs. global engagement of Brazil and South Africa at the UN General Assembly

Brazil and South Africa have long been regarded archetypical regional powers, commanding more resources than their neighbours, spearheading regional projects and pursuing high-profile global status. Yet, recent years have also evidenced how the engagement with their regions and acceptance as leading players is often ambiguous and incomplete. How does one ascertain that a regional power privileges either the regional or the global stage? Through an original dataset of Brazilian and South African output at the UN General Assembly between 1994 and 2013, we monitor sponsorship patterns and thematic preferences in order to verify whether these countries indulged their regional partners and topics. Our findings suggest that Brazil and South Africa favoured their immediate neighbourhoods, but have gradually engaged their regions in different ways: while Brazilian emphasis on regional peers and themes declined over the years, South Africa developed an increasingly more regionalised UNGA agenda.


Introduction
The importance of regional powers within their respective neighbourhood has become a prolific academic topic in the last decade, due to the growing multipolarity of the international system.Yet, the topic of regional powers remains 'a complex and multifaceted one' (Nolte 2010, 883).Considerable attention has been devoted to the question of how such countries direct their foreign relations, especially given their pivotal position between the region and the world.However, due to the mostly qualitative and case-oriented nature of this research agenda, answers to such questions have been so far incommensurable.
How does one ascertain that a regional power privileges either the regional or the global stage?The lack of a comparable metric across cases has stranded scholarship on idiosyncratic evaluations and hampered further progress.Responding to this query becomes currently pressing, as archetypical regional powers, such as Brazil and South Africa, have undergone significant foreign policy changes.Ambitious regional projects of the past decade have receded due to economic decline and political change, as new governments unceremoniously jettison previous alignments, and refuse to take upon the role of stabilizers even in the face of mounting tensions in their own neighbourhoods.In light of such changes, it is important to assess if and to what extent such players are indeed (or have ever been) deeply enmeshed with their surroundings.
The choice of privileging regional or global linkages has already been approached through different angles (Prys 2010;Hurrell 2010).Scholarship on Brazil and South Africa has also attempted to weigh in on the regional vs. global orientations of their foreign policies (Vigevani and Cepaluni 2011;Malamud and Rodriguez 2014;Krapohl et al. 2014;Alden and Schoeman 2013).Some fuzziness, however, remains as to precisely which arena is emphasised as 'the global' level in these cases: whether that concerns ties with extra-regional partners from the developed world (Krapohl et al. 2014) or from the Global South/BRICS constellation (Steiner et al. 2014), whether it involves status-seeking in high-profile multilateral organisations (Malamud 2011;Cline et al. 2011), or if it consists in conforming to expectations of the international liberal order instead of sticking to local allegiances (Prys 2009).Following Nolte's (2010) criteria for defining regional powers, which include articulation and representation of regional interests in global fora, this article attempts to ground the option between regional and global priorities more solidly.Namely, we inquire how much attention have regional powers devoted to their respective surroundings in comparison to other geographies, as manifested in multilateral organisations.We believe an original contribution is in reach if based on original longitudinal data.
We intentionally avoid the lengthier debate over how to categorise regional powers and their strategies, given the already robust state of such typologies (see Nolte 2010;Destradi 2010;Prys 2010).Accordingly, we take up Brazil and South Africa as case studies, often considered classical examples of regional powers (Flemes and Lemke 2010, 324), sharing similar diplomatic profiles and regional predicaments.Amidst other heavyweight BRICS countries, Brazil and South Africa have been consistently singled out as a more comparable duo (Krapohl et al. 2014;Westhuizen 2016).Yet, for both cases, the question of how central the region is in their priorities remains ambiguous (Lazarou and Luciano 2015;Alden and Schoeman 2015).
Our descriptive research design revisits the canonical interpretations of Brazilian and South African foreign policies in the post-Cold War era, paying particular attention to the relative centrality of regional endeavours over time.Regional vs. global orientations are then contrasted with a novel dataset on United Nations General Assembly (UNGA) draft resolutions.Such data are particularly well-suited, as the UNGA provides a 'record of how the state wants to be seen by others, the international norms it finds acceptable, and the positions it is willing to take publicly' (Mattes et al., 2015, 284).In contrast with prevailing UNGA scholarship, we do not rely on roll-call voting and concentrate instead on cosponsorship of draft resolutions. 1he article begins with a review of the regional-global linkages amidst the literature on regional powers, as well as some clarification of key concepts.After bringing forward our research design, we explore the scholarship on Brazilian and South African foreign policies in order to pinpoint key entanglements with both the region and the world.We then test our argument by evidencing their regional aspirations and attempts to go global while based on their sets of regional supporters and topics of choice in UNGA proceedings.Following the analysis of the results, we conclude with a few closing remarks.

Regional powers and the regional-global nexus
Regional powers have been defined by the literature around a core of constitutive features, including: a) belonging to a given region; b) having material capabilities superior to the rest of the region; and c) exerting some kind of influence over the region (Destradi 2010).Other typologies and labels also abound, ranging from middle powers to rising powers (Jordaan 2017; Nel 2010).Characterising the expected relationship between a regional power and its region is not a simple exercise, as the latter is never cut-off from the rest of the world and its intruding influence.In addition, scholarship has employed varying theoretical viewpoints to frame the interlinkage between regional and global arenas, from multilevel governance (Kacowicz 2018) to international society (Hurrell 2010).For the purposes of this article the regional-global nexus can be defined as a 'conceptual point of tension on which multidimensional regional and global forces are focused' (Job 2009, 43).Under this perspective, the tension within the regional-global nexus refers to how competing (or complementary) forces, constraints, opportunities and resources at both levels -particularly within the context of institutions -hamper or advance the goals of regional powers.
The pursuance of such aims may or may not lead to strong regional engagement.Some suggest that the region is cast aside once claims of global powerhood come into play, because the latter holds the promise of greater reputational rewards (Krapohl et al. 2014).That falls in line with Hurrell's argument that 'you can be a global player without being a regional power and that it is the discretionality of your involvement within the region that is a potentially important indicator of your global power ' (2010, 21).In other words, when balancing the region-world equation, regional powers should not be assumed to always favour the former.
One way to explore the effects of this bifurcation on the strategies of regional powers is to rely on Prys' understanding of the double-mirrored 'outside-in ' and 'inside-out' dimensions (2010, 496-499).The first consists of the systemic pressures by the international system that downwardly affect regional dynamics.These may legitimise the aspirations of regional powers, via a de facto deputization of responsibilities or even their appointment as intermediaries of the international community.But systemic forces can also go against existing regional integration projects (e.g.concurrent trade agreements) and undermine the path of choice favoured by the regional power.
An inverse rationale is found in the 'inside-out' dimension.Regional powers might act as gatekeepers and strive to become exclusive regional representatives in multilateral fora; or instead be tempted to divert political emphasis and material resources away from regional management and allocate them instead towards the pursuit of a more bountiful global agenda.Although both goals can be blended, a duality still emerges over which domain receives precedence (Prys 2010, 499).The lack of sufficient resources to stake a claim at both levels might become too apparent and undermine achievements on both fronts.Likewise, by playing more actively by the rules of the liberal order, regional powers risk being connoted with the centres of gravity that they might have once rallied against.Such a conundrum can also foster the perception that regional powers are growing apart from regional dynamics and forfeiting their role, thus inciting contestation from regional competitors.Under this framework, a continued commitment to local interests and the disposition to represent them can be considered important tokens of regional prioritisation.

Engaging with regional themes
A regional power's engagement with its region might take on different forms.The literature has so far privileged specific roles as signs of involvement, like paymastering or creating regional organisations (Flemes and Lemke 2010).However, typical regional powers, such as Brazil and South Africa, tend to underperform such roles, due to their own economic vulnerabilities and autonomy-preservation strategies (Vigevani and Cepaluni 2011).Another dimension of regional engagement which has received less attention is the matter of representation.Nolte (2010, 893) understands that one of the features of a regional power is that it be 'integrated in interregional and global forums and institutions where it articulates not only its own interests but acts as well, at least in a rudimentary way, as a representative of regional interests.'Schirm (2010) ties this issue of representativeness to the very notion of leadership/followership.He defines followership as a way of 'supporting the goals and positions of another country which were not shared previously and/or as accepting a relative loss of status and power vis-a-vis the emerging power' (Schirm 2010, 200).Hence, states will follow a potential leader whenever their shared interests are enabled (Scholvin 2013).This implies that, alongside the temptations to 'go global', regional powers also have strong incentives to tend to the interests of their peers, consolidate an inclusive regional agenda and bring it to proper venues.The UN is one of the stages to which regional powers might relay such shared demands, if not as initiators at least as supporters.The General Assembly, in particular, is a prized arena due to its horizontality and thematic openness to matters from regions of the Global South (cf.Peterson 2006;Seixas Corrêa 2013;Graham 2016).
Scholarship on regionalism, regional powers and the UN has focused on varied aspects of this interplay, such as regional cohesion (Montenegro and Mesquita 2017), representation (Lai and Lefler 2017) and actorness (Seabra and Sanches 2019).Few studies, however, have paid close attention to the salience of regional themes on UN output, and even less resorted to metrics apart from roll-call votes.

Research design
Regions are not closed off to the outside world and neither should the efforts in understanding their dynamics remain so.But even when the regional-global nexus is considered, most contributions have only focused on sparse episodes where regional support is required, with an emphasis on campaigns for the temporary UNSC seats (e.g.Schirm 2010, 201-204;Malamud 2011, 9-10;Smith 2018, 128).While these initiatives are noteworthy on their own, we would argue that they do not exhaust the pool of available displays of regional followership in global stages.Other platforms, where intense negotiations and bargaining take place in a more recurrent format, should also be considered.
The predicaments of regional powers can be further explored through their behaviour at the UNGA.A focus on co-sponsorship of UNGA draft resolutions provides us the benefit of (1) a large number of observations, with adequate homogeneity and cross-unit validity for comparisons due to institutionalised regularity; and (2) an opportunity to observe global vs. regional dealings simultaneously.Whenever distributing their preferences, member states choose whether regional or global partners and issues will be prioritised.Stately transactions at the UNGA can thus assist in revealing the preferred 'inside-out' strategies of regional powers.Based on our conceptualization of regional engagement, we focus on two observable indices: cooperation with regional peers and thematic emphasis given to the region.The former consists in the relative predominance of regional partners (as opposed to extraregional ones) supporting the same draft resolutions as the regional power.The latter refers to the content of resolutions, specifically to how often regional themes present in the proposals were endorsed by the regional power.

The process of UNGA resolutions
In spite of their complexity, the institutional rites of the UNGA remain fairly stable and predictable.Every session, an item is put on the agenda, as previously discussed by the General Committee of the UNGA.Most draft resolutions are then initiated and drafted by a member state (the 'main sponsor'), and usually subscribed by other supporters ('cosponsors').Every draft then turns into a so-called L-document and is normally tabled in one of six different committees, each referring to a different policy area.
Once a draft is tabled, its trajectory can vary considerably.In an ideal-type scenario, a draft will cruise unmodified through the original committee, before being adopted as a full resolution by the plenary, either through a vote, by consensus, by acclamation, without objection or without any vote (Peterson 2006, 54).However, L-documents frequently receive additional contributions along the way.These can alter either their original sponsors (via Addenda), their content (via Corrigenda and Revisions), or even both (via Revisions).The range of opportunities for member states to engage with the process by sponsoring a draft remains therefore very wide, without any major caveats or preconditions.2

Data and methods
The data for this analysis comprise the total amount of drafts that Brazil and South Africa cosponsored in the UNGA between 1994 (Session 49) and 2013 (Session 68). 3 This period was selected because South Africa was suspended from the UNGA between 1974 and 1994 as part of the international embargo against apartheid.We ended the series in 2013 so as to adhere with the timeframe of traditional interpretations of Brazilian and South African foreign policies, here under analysis.The resulting timespan of 20 years provides a comprehensive overview of these regional powers in the first decades of the post-Cold War era.
We retrieved our corpus of draft resolutions from the United Nations Bibliographic Information System (UNBISNET). 4 In total, we extracted 2555 pieces pertaining to Brazil and 2514 to South Africa.That included not only drafts for which either Brazil or South Africa were listed individually as co-sponsors, but also those supported by groups to which the two countries belonged: GRULAC (Brazil); the African Group (South Africa), the Non-Aligned Movement (NAM, South Africa); and the G77 (both Brazil and South Africa).
Given that many L-documents amounted to subsequent versions of a preceding document, we organised our dataset so as to link original drafts to their complementary additions (revisions, addenda and verbal sponsorship).We refer to this process as 'sequencing'.After sequencing all documents according to their linkage to previous pieces, we obtained a total of 2169 drafts for South Africa and 1988 for Brazil.Lastly, sequenced drafts were paired to their resulting UNGA resolutions.South Africa's 2169 drafts were matched to 2078 resolutions, whilst Brazil's 1988 drafts were paired to 1909 resolutions.and processing, as well as the list of regional groups, member countries and dictionaries of terms for the content analysis.
We employed our data in the following manner: in order to verify whether regional partners are preferred to extra-regional ones, we carried out several comparisons of the frequency of draft co-sponsorship between the two regional powers and their respective neighbours.As outlined in section 3.1, countries can co-sponsor a draft either from the outset or in later stages through revisions or addenda.We believe a more accurate depiction of regional followership is displayed by those states that cooperate as soon as the regional power is involved.Hence, our analysis takes into consideration synchronic co-sponsorships and ignores prior and later adhesions.As for the prevalence of regional themes, we carried out a content analysis of the text of UNGA resolutions that were originated in the collected drafts, searching for keywords referring to the geographical region, its political groupings or member countries.5

When Brazil and South Africa attempt to go global
Brazil and South Africa have regularly stood out amidst the literature on regional powers.The case for their comparison is compelling, either due to a structural 'like-mindedness' (Westhuizen 2016) or to their similar claims of regional leadership, which fuelled aspirations in the world at large.Both, however, have also come to exhibit volatile behaviour that hampered deeper regional integration while facing an equivalent retrenchment when confronted with failed expectations and domestic instability (Krapohl et al. 2014).This section makes the case for exploring their external insertion as a testament of the regional-global duality that often characterises regional powers.

1990-2003: Opening to the region and to the (developed) world
During the Cold War, Brazil's regional ties remained largely underdeveloped, due to mutual suspicions and introverted economic agendas of the military governments.With the end of bipolarity, South American countries sought greater dialog as a means of both strengthening their nascent democracies and pursuing economic development adequate to globalisation (Dabène 2009).For Brazil, this was materialised in a steady approximation with former competitor Argentina, which then culminated in the creation of the Mercado Común del Sur (Mercosur) in 1991, along with Paraguay and Uruguay.
The 1990s were therefore a period of regional rediscovery.Even though domestic instability prevented presidents Collor andItamar (1990-1994) from developing long-term agendas, their diplomatic endeavours were still relevant for regional (i.e. the Rio Group) and global (i.e. the Eco92 summit) affairs.Meanwhile, the Cardoso years (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002) were mostly characterised as a 'search for credibility' (Soares de Lima and Hirst 2006) and a pursuit of 'autonomy through participation' (Vigevani and Cepaluni 2011).Cardoso was invested in assertive regional initiatives, such as South American presidential summits and regional crisis management.Mercosur was also strengthened via the inclusion of a democratic clause in 1998 and intra-bloc trade grew vigorously in its early years.At the same time, an emphasis on global stages was explained by the need for an economically vulnerable country such as Brazil to abandon previous isolationist stances and constructively adhere to international regimes.The prevailing tone was one of 'opening'.Both towards the region, as manifest in overcoming past suspicions and fostering integration through trade; and towards the globe, by embracing a collective stance on new themes of the globalised agenda (e.g.environment, human rights) and adhering to the normative frame of the international liberal order.Brazil's re-orientation on nuclear non-proliferation, in particular, is a significant emblem of harmonization with regional (ratification of the Treaty of Tlatelolco in 1994) and global (ratification of the Nuclear Proliferation Treaty in 1998) regimes.This dual opening was not considered mutually exclusive.Mercosur was in fact labelled as 'open regionalism', reaping gains from economic liberalisation, but with few sovereignty costs.By and large, it was regarded as a means to improve Brazil's position as a global trader (Vigevani and Cepaluni 2011).Brazil's 'inside-out' strategy therefore perceived regional overtures as a springboard towards global insertion.The South American context of the time contributed to this alignment.After long military dictatorships and failed importsubstitution paradigms, Brazil and its neighbours were attracted to the same set of liberal values involving democracy, human rights and economic liberalisation.Interest convergence did not result mainly from hegemonic co-opting and payoffs; rather, Brazil's 'consensual' brand of hegemony relied on framing ample regional interests under the guise of its own goals (Burges 2015).Hence, regional cohesion was a by-product of symmetric regional preferences for liberal values, not solely of Brasilia's agency.

2003-2013: Diversification and globalization
The later years of Cardoso's presidency were marred by international financial crises, which damaged both Mercosur relations and faith in the US-led liberal order.Under such strain, foreign policy under Lula da Silva (2003Silva ( -2010) ) underwent powerful reorientations.The ideological makeup of South American democracies changed, as several left-wing presidents came to power, with the institutional landscape evolving accordingly.Although Mercosur continued to expand, its commercial rationale was sidelined in favour of political cooperation.
For Brazil, the main regional project became the Union of South American Nations (Unasur).
Privileging political coordination over commercial aims, the bloc gathered all 12 South American countries and became an important arena for mutual assistance on areas such as defense, infrastructure and crisis management.The initiative was regarded as Brazil's grand attempt to structure the South American space around Brasilia (Sanahuja 2012, 34).Enjoying political stability and economic growth, Brazil at last appeared to display paymaster-traits through large investments in neighbouring countries, even though it is unclear how sufficient they were to elicit actual followership (Lazarou andLuciano 2015, Burges 2015).
Brazil also began to pursue a stronger identification with new centres of power, by means of 'autonomy through diversification': the country valued partnerships with the Global South, not necessarily to the harm of relations with developed countries, but consistently seeking to reduce asymmetries and leverage its bargaining position (Vigevani and Cepaluni 2011).Such orientation was manifested in the balanced mix of South-South (BRICS, IBSA) and North-South (G20) groupings of choice.Consequently, Brazil acquired a global profile, with a growing diplomatic footprint in Africa, Asia and the Middle East.For some geographies, such as Atlantic Africa, this approximation was even canvassed by initiatives with an UN-point of origin (Abdenur, Mattheis and Seabra 2016).
Regional prioritisation was yet again not detrimental to a global profile, and the country now claimed more explicitly to speak for its neighbours in global stages (Mesquita and Medeiros 2016).Brazil's leadership of the UN Mission for the Stabilisation of Haiti (MINUSTAH), for instance, afforded both regional pre-eminence and status-boosting within the UN system.Yet, Brazil's profile in the late 2000s also grew more autonomous from regional projects, reflecting either weariness from the recalcitrance of its would-be followers or the will to 'go global' alone, considering itself too big to be tied down by regional schemes (Malamud and Rodriguez 2014;Steiner et al. 2014).
The enabling of regional interests was nevertheless partial.Claims by weaker neighbours were at times considered, either through structural investments or political support and crises mediation.However, the ideological fragmentation of the region rendered it more difficult to find a unifying agenda.In contrast to the homogeneous pro-Washington Consensus scene of the 1990s, South America was now crisscrossed by competing projects and institutions, ranging from the liberal Alianza del Pacífico to the anti-imperialist Alianza Bolivariana para los Pueblos de Nuestra América (Alba) (Gardini 2011).
Personal disengagement with foreign affairs also made Rousseff a relatively less active president.Lastly, having inherited a country with a more globalised profile, she was compelled to devote substantial attention to extra-regional partners: visits to South American countries dropped from 43% (Cardoso) and 36% (Lula) to 35% of total presidential travels (Mesquita 2019).

1994-1999: The pains of transition
The role of Pretoria within Southern Africa as well as in the broader continent is marked by both a deep historical legacy and the corresponding political will to surpass it.Following decades of international isolation, 1994 witnessed a wholesome effort to push beyond the apartheid stigma and win over new external credentials.Accordingly, as the face of this new era, Nelson Mandela (1994Mandela ( -1999) ) made clear from the start that South Africa would fully cooperate with the international system and its institutions, while keeping Africa at the forefront of its priorities.This multipronged predisposition entailed putting into practice the much-publicised 'six pillars' agenda, which included, among others, the promotion of human rights, democracy, justice and respect for international law, both in the world and in the continent.
Amidst the centrality attributed to Africa, South African authorities were thus pushed to assume the moral compass of how regional relations should be conducted (Geldenhuys 2010).This stance was 'stressed in part to recognise Africa for its assistance and suffering during the liberation struggle, as well as an attempt to appease regional fears of South Africa's continued hegemony' (Graham 2012, 415).In other words, there was already an imbedded recognition of what the size and ambitions of a regional power like South Africa could incite in its neighbourhood.However, despite early caution, these years ended up being characterised by a largely ambiguous foreign policy outreach, in which the premium on human rights and democracy often incurred in initiatives that either contradicted some of its tenets (i.e. during the South-African led intervention in Lesotho in 1998, with no international mandate) or underscored the limits of its continental appeal (i.e. when South Africa failed to secure enough African backing to censor Nigeria's human rights violations in 1995).
At the same time, efforts put behind the Southern African Development Community (SADC) largely faltered due to polarisation with other members, as South Africa was often pitted against Zimbabwe in different political and security issues.In order to circumvent these hurdles, sizeable investment was made in terms of also becoming a privileged South-South interlocutor before the remaining international community.The hosting of high-profile conferences (i.e.UNCTAD, NAM) during this period as well as the adherence to international normative treaties (i.e. the NPT) attested to the duality of South Africa's external positioning (Landsberg and Monyae 2006;Serrão and Bischoff 2009).It soon became clear the intent to become a 'major player but one that will not act unilaterally or in ways that demonstrate its not inconsiderable power in the region' (Bischoff 2003, 189).Unconcerned with whether it had to choose between the region and the world, South Africa quickly came to recognise the possible drawbacks of a more pushful stance in its neighbourhood.

1999-2013: Consolidation and contestation
Following the post-apartheid transition, the presidencies of Thabo Mbeki (1999Mbeki ( -2008) ) sought to more clearly present South Africa as an active proponent in the world and in Africa.
Attempting to bridge both stages, the country exhibited more overt continental leadership claims, ranging from participating in G8 summits involving Africa, a prominent role in developing and financing the New Partnership for Africa's Development (NEPAD) and in supporting the creation of the African Union (AU).Calls for the reform of both global and regional governance architecture were combined with a renewed emphasis on South Africa's identity as a developing state committed to elevate Africa's stance in the world and promote new economic and political paradigms, wrapped under the African Renaissance narrative.This bridging role was visible at the UN-level: South Africa's first-ever term as a UNSC nonpermanent member (2007-8) culminated with a resolution (1809) institutionalising the relationship between the AU and the UN on peace and security.Meanwhile, NEPAD was also included in UNGA discussions on development (Mashabane 2018).
Overall, Mbeki's foreign policy was considered 'expansive and ambitious.It secured a place for South Africa at the highest table, though not always without its detractors in Africa' (Sidiropoulos 2008, 111).Indeed, these initiatives did not necessarily merit automatic regional acceptance, but 'acquiescence on leadership' at most (Flemes 2009, 149;cf. Scholvin, 2013).Contestation by secondary powers (i.e.Nigeria; Angola; Ethiopia) on such issues as the campaign for a permanent African seat at the UNSC was evident over the years and showcased the limits of South Africa's inside-out approach.Mbeki's recurrent 'quiet diplomacy' towards Zimbabwe, neighbourly fears of economic takeover and accusations of Western deputization were also routinely brought up (Prys, 2009).But despite such misgivings, South Africa managed to deliver on some fronts.Participation in multiple regional crisis mechanisms, like in the DRC, substantiated a more continental purview.However, under Jacob Zuma (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018), and despite more efficient action within the SADC perimeter, South African foreign policy also became more overtly ideological (Adebajo 2018, 13-14) and the favourable vote on the intervention in Libya in 2011 or a more forceful bid for the leadership of the AU eventually contributed to a decline in Pretoria's image throughout Africa.
Yet, South Africa's regional status also kept on being more readily recognised by the international community than by African states.Indeed, involvement with either the NAM, the Group of 77, the European Union or even BRICS, was primarily based on South Africa's perceived capacity to advocate African interests in global venues, even if only symbolically (Alden and Schoeman 2015).As Smith argues, this utilitarian approach has served a dual purpose: on the one hand, it is 'what gives South Africa access to international groupings and leadership positions that are far beyond its reach were it to be judged purely on its own merit'; on the other hand, through such groupings, 'South Africa lobbies for support from its African counterparts, on the basis that it represents them in these fora ' (2018, 128).South Africa's regional powerhood might then comprise more a 'product of international needs for African representation on the global stage, together with its own ambitions, rather than any regional consensus on South African leadership' (Alden and Schoeman 2015, 241).

Analysis: Going global without the region?
Both Brazil and South Africa assign a considerable role to the UNGA.Brazilian authorities have routinely called for greater centrality to the organ as a forum to give voice to marginalised views, including from their own regions (Seixas Corrêa 2013).As for South Africa, it has led a 'self-conscious Africa-centred UN policy, at least rhetorically, where it has attempted to ground all of its engagements with the world body under the framework of its ambitions regarding Africa' (Cornelissen 2006, 47).
Yet, previous work on Brazil and South Africa's profile at the UNGA has relied solely on data from voting records.Recent studies have focused on Brazil-US relations (Amorim Neto 2012), on Brazil-Africa connections as tokens of South-South solidarity (Seabra and Sanches 2019), the consistency of South Africa's track-record at the UNGA with its major foreign policy tenets (Graham 2016), or the performance of both states in contrast to other BRICS countries in fostering regional cohesion (Montenegro and Mesquita 2017).No effort has been made so far towards exploring the possibilities that co-sponsorship of drafts at the UNGA entail.
Hence, the following sub-sections compare Brazil and South Africa's regional vs. global attention.In the case of Brazil, we contrast the broader Latin American Group (GRULAC) at the UNGA with the South American neighbours, while in the case of South Africa we concentrate on the African Group and on Southern African countries.Also, given that our literature review partitioned these countries' foreign policies in two phases each, our data will also be grouped per decade, when convenient, in the 20 years analysed.

Brazil
Brazil's footprint at the UNGA displays an increasing trend over the years.Overall, the country's co-sponsoring activity consistently augmented, reaching its peak in Session 65 (2010), with a total of 160 drafts.On average, Brasilia co-sponsored 128 drafts per session.In order to verify cooperation with regional peers, we first observed which countries were the most frequent co-sponsors of Brazil's 1988 sequenced drafts.For the whole series, Brazilian drafts warranted 74 co-sponsors on average.As shown in Figure 2, Latin American partners are clearly preferred (except for Mexico).Brazil's five most frequent partners are Argentina (1381 co-sponsored drafts), Chile (1378), Costa Rica (1312), Guatemala (1305) and Peru (1267).As witnessed in Table 1, Brazil stuck to this regional preference over time.Source: elaborated by the authors.In each session/year, the most frequent co-sponsor stands on the left.Grey cells correspond to regional partners and white cells to extra-regional ones.
These results refer to the total of Brazil's UNGA output, including drafts that it cosponsored as a member of groups.In order to better capture the profile of the country, we must disaggregate these totals and verify how much of it is due to collective affiliation and how much it is owed to autonomous agency.As shown in Figure 3, Brazil's participation via groups in the UNGA is restrained as it only belongs to the G77 and the GRULAC.The GRULAC displays scarce activity in comparison with the G77: the former only sponsored 23 drafts in 20 years, while the latter supported 734.In every year, Brazil's autonomous co-sponsorship 24 was nearly double its collective activity.It is also noteworthy that at no point did Brazil cosponsor drafts on behalf of any of the two groups.This low performance within the GRULAC could be interpreted as tentative indication of ineffectual regional engagement.However, such results need to be refined as Brazil's cosponsoring with other groups can still elicit local followership on their own.In order to fully answer this question, we need to verify whether most of Brazil's UNGA output took place with or without the support of regional peers.If we consider South American countries and the GRULAC as our regional litmus scenarios, how often did Brazil secure the backing of most or all of its neighbours behind a given proposal?Figure 4 displays in separate plots the amount of Brazilian-sponsored drafts which mobilised none (0%), less than half (1-50%), more than half (51-99%) and all (100%) of countries from these two settings.The plots differentiate between drafts produced autonomously and within collective caucuses.Each plot compares the amount of drafts for the first ten sessions (1994 to 2003), and the last ten sessions ( 2004to 2013), with respect to the percentage of support amassed.The results indicate the temporal tendency in terms of backing within each regional selection.If we compare the first ten with the last ten sessions, we notice that between 1994 and 2003, majoritarian regional support (when more than 50 percent or all of the members of a regional group joined in a draft) was more common.From the 909 drafts Brazil cosponsored, 560 (62 percent) were supported by over half or all of South American countries (plots A3 and A4); and 50 percent when accounting for GRULAC members (B3 and B4).But in the following decade, these figures dropped to 52 and 45 percent, respectively (averaging values for the same plots).This trend is portrayed in the plots, which demonstrate that the proportion of drafts engaging with none or very few of Brazil's neighbours increased visibly (plots A1, A2, B1 and B2).In other words, Brazilian co-sponsorship with out-of-region partners grew considerably in later years.
Two caveats need to be highlighted.On the one hand, even though the number of drafts yielding complete consensus grew in the case of South American neighbours (from 374 to 428, as displayed on plot A4), that was only achieved through the effect of the G77.In fact, with the exception of work within the G77, regional support decreased across the board.On the other hand, even though total engagement with the full set of Latin American and Caribbean countries was also small (B4), that is probably due to the exit of Mexico from the G77 in 1994.If we were to exclude Mexico from the pool of regional peers, all of the G77's 734 drafts would yield 100 percent Latin American support.
In short, the data on Brazil's preferred co-sponsors show that Latin American countries have remained the most frequent partners over the course of 20 years, consistently ranking higher in the list of recurring collaborators.Nonetheless, this overall affinity is not matched by any sort of shared Latin American drive.The GRULAC, for one, has collectively endorsed remarkably few drafts, thus implying regional cooperation at the UNGA is scarce.By turning to data on regional support and contrasting Brazil's 1994-2003 profile with that of 2004-2013, we detect a more successful streak in securing followership from the narrower South American coterie than from the wider GRULAC.Yet, any signs of South American followership also seem to be more a result of G77-inner work.Coupled with the rise in drafts that engage none or less than half of its neighbours, these figures suggest the interactions of Brazil with its region at the UNGA have become less significant over the years.
In order to observe our second index, thematic emphasis given to the region, we carried out a content analysis of UNGA resolutions derived from preceding drafts cosponsored by Brazil.We elaborated a thematic dictionary containing 62 keywords -including the names of Latin American and Caribbean countries, institutions and geographical subregions -and applied it to a total of 1824 resolutions.In total, 19 percent of those resolutions presented terms corresponding to our list, with the keywords warranting a total of 272 mentions.
Figure 5 summarises the findings along two series: keyword frequency and percentage of resolutions containing at least one related keyword.The former visibly decreased, indicating that mentions to the region in resolutions initially co-sponsored by Brazil became less abundant.Inversely, the percentage of resolutions presenting a keyword was higher in the 1990s and faced oscillations in the following years.On average, however, the mean share of relevant resolutions was the same in both decades.Although the share of relevant resolutions varied considerably, word frequency clearly diminished in later years.This confirms that Brazil's increasingly autonomous profile is also reflected in the thematic content of its output.The results lead us to conclude that Brazil paid less attention to regional themes over the years.

South Africa
South Africa's activity followed an expected trend in the sense that, after having re-joined the UN in 1994, it became a considerably productive member state.As shown in Figure 6, 2002 (session 57) stands out in this regard (164 drafts supported).Still, soon afterwards, South Africa's contribution entered a downwards phase, with the mean number of drafts per session stabilising on 126.  in Africa.If we were to restrict our focus to this general data, we would find tentative grounds to validate strong regional solidarity, given how Pretoria collaborated more intensively over the years with such countries as Senegal (1500 drafts), Morocco (1466), Egypt (1458), Algeria (1455) and Namibia (1454).More importantly, we can observe in Table 2 how South Africa increasingly moved towards its African peers.However, in order to account for the distorting effect of different groupings, we disaggregate these results.As shown in Figure 8, South Africa's footprint is evenly split between associating with larger caucuses (1045 drafts in total) and going at it alone (1124).
Between sessions 55 (2000) and 61 (2006), South Africa co-sponsored far more drafts outside of groups than collectively, though this tendency has been reverted since 2006.From the 3 groups it participates in, the G77 is the most active.The African Group and the NAM yield similar amounts, despite co-sponsorship via the former becoming more robust in the late 2000s.

Figure 8: Total amount of drafts co-sponsored by South Africa per session and group
Source: elaborated by the authors.Some drafts were sponsored by more than one group: G77 andNAM co-sponsored 22 drafts between 1994 and1996; NAM and the African group co-sponsored 1 draft in 2008.The 'On Behalf' series, represented by an X, indicates the number of drafts South Africa submitted representing one of the groups; in total: 7 drafts on behalf of the African Group, 39 on behalf of the G77, and 37 on behalf of the NAM.
We must also verify whether neighbouring countries joined the initiatives subscribed by Pretoria.When comparing the African Group and Southern Africa, it is important to bear in mind that the likelihood of engaging any African partner will always be greater as there are over 50 countries belonging to the former group, while only around a dozen belong to the latter.Source: elaborated by the authors.The total amount of members from African Group varied due to new entries or suspensions.Likewise, our list of Southern African countries mirrored the changes that SADC faced in its composition over the years As Figure 9 shows, an important temporal distinction stands out.Previously, the bulk of South Africa's activity outside of groups mobilised less than half of the countries of Southern Africa and the African Group (plots A1, A2, B1 and B2).In fact, for most of the time, South Africa only secured support from 100 percent of its African peers when acting through the African Group, the G77 or NAM (A4 and B4).But as previously demonstrated in Figure 8, collective co-sponsorship eventually became more common for South Africa than individual activity.Likewise, South African-backed drafts that merited zero regional sympathy became rarer with time (A1 and B1).This implies greater cohesion with the region, as a larger number of drafts warranted support from 100 percent of Southern Africa and the African Group (A4 and B4).In that sense, there is tentative evidence of a shift in South Africa's profile over time, bringing it closer to Africa.In the late 1990s-early 2000s, the country accelerated its autonomous participation at the UNGA and the gap between the number of African and non-African co-sponsorships widened in favour of the latter.From the mid-2000s on, in contrast, most its output implied more frequent co-sponsoring with regional peers.
As for the salience of regional themes, we assembled a dictionary with 83 keywords related to African countries, institutions and subregions, and applied it to a corpus of 1993 resolutions originating from South Africa-backed drafts.In total, 21 percent of those resolutions had relevant keywords, amounting to 422 mentions.
Figure 10 depicts both keyword frequency and share of resolutions containing at least one keyword.The initial years were particularly active for South Africa, with a high count in both word frequency and share of Africa-related resolutions.In the following years, however, the two measures decreased.The early 2000s witnessed a short-lived spike, particularly concerning word frequency.However, from 2009 onwards, we observe a robust and continued rise in both indices again, indicating that a larger portion of South African output is focused on regional matters.only did Pretoria's efforts in later years have a greater regional emphasis, but even during the 2000s, when activity outside groups was greater, thematic attention to Africa was still present.

Conclusion
The narrative over regional powers in the international order has been seriously curtailed as of recently, due to a generalised retrenchment in the capacities at their disposal to materialise ambitious external priorities.More often than not, 'troubles in the backyard can be burdensome' (Nolte 2010, 884).Our study brings to light a far more complex picture, by revealing similarities and differences between Brazil and South Africa when it comes to relying on the region as a springboard for multilateral endeavours abroad.The inside-out dimension comprised the focal point through which we sought to move forward the existing debate.By confronting prevailing interpretations of these countries post-Cold War foreign policies with their UNGA behaviour, where both regional and global themes and allegiances are available to member states, we were able to gauge the relative salience of regional matters, relying on a comparable metric instead of interpretivist approaches.
Concerning the relative centrality of regional peers, we observed that Brazil has indeed privileged its immediate neighbours in Latin and South America as preferred partners.
However, it has also began to steer away from the region -or any other grouping for that matter -over the years, resulting in a more autonomous profile.This might be endogenously explained as a reflection of an enduring concern with not being tied to any particular group, in tandem with its autonomist drive and a sense of overall regionalism fatigue (Malamud and Rodriguez 2014), or exogenously as a consequence of growing ideological disunity in Latin America (Gardini 2011).In the case of South Africa, we find that the country worked closer with other African partners as it reconnected with the international community after 1994.
Yet, in contrast to Brazil, two changes became visible: not only did South Africa's engagement with the region and the overall continent grew more salient with each passing session, but it also began to privilege collective associations as the main approach at the UNGA.This falls squarely in line with previous assumptions over the inescapability of the global stage for the projection of South Africa's African credentials (Alden and Schoeman 2013;Smith 2018).Even though both countries delivered less regional prioritisation than one would expect in light of the rhetoric often attached to their regional powerhood claims, South Africa succeeded in coming ahead of Brazil.
As for the thematic emphasis given to the region, the same divergence between both regional powers emerges.Brazil's engagement with Latin America in this domain was very thin and decreased over time.South Africa, for its part, managed to put up a stronger showing than Brazil, tackling regional matters more frequently.Indeed, the institutionalisation of relations between the UN and regional initiatives seemed to be of significant importance for Pretoria.Yet, as we compared only two cases, similar exercises with other countries might be required to ascertain whether regional emphasis was indeed inexpressive or, in contrast, above UNGA average.When trying to capture regional dynamics in multilateral fora, it is important to bear in mind that some regions might demand and offer more leadership-followership opportunities, depending on the venue.As Prys (2010) pointed out, leadership is embedded within regional contexts, and as certain international organisations become more engaged with a specific region, they might also become preferred stages to incite followership.This is arguably the case for Africa: as the continent often warrants peacekeeping missions, development initiatives and incites humanitarian concerns, the UN becomes a natural platform for the exercise of African leadership, perhaps even more so than for other regions. 6iven how regions are not all equally porous to global institutions (Kacowicz 2018), we may thus end up learning more about South African regional powerhood than about Brazil when focusing on the UNGA.
This article provides two key findings to the overall debate on regional powers.First, some distinctions between Brazil and South Africa became apparent, showing that, for all the similarities the literature underlines, their profiles and outlooks towards the region in multilateral fora have come to take on distinct paths over the years.Second, by focusing on co-sponsorship as a profiling tool -as opposed to the often-employed roll-call voting records -we delved into a much richer data source concerning preferences and associations displayed by countries at the UN.This approach has proven useful for tapping into the duality of the regional-global nexus, which should remain central to the research agenda on regional powers in the coming years.
Our conclusions must be pondered against inherent limitations.Firstly, the connection between regional actors (and particularly organisations) and the UN can vary diametrically, ranging from cooperation to hostility (Kacowicz 2018).We should therefore not expect full symmetry in the translation of regional emphases onto UNGA behaviour for all cases at all times.Secondly, for the sake of parsimony, we did not distinguish between initiating a draft resolution and merely co-sponsoring it, even though theoretically the former could result in a superior token of regional prioritisation.
Ultimately, we took upon the challenge of evidencing the implications of the interplay between regional and global power projections.If regional powers are truly the 'nodes between the global and the regional power hierarchies' (Nolte 2010, 889), then regional powerhood should not just be defined by what it is accomplished within the region, but also by what such countries are able to bring into the world stages.Our data provide a more indepth outlook of how states behave in international multilateral fora when attempting to go global.The same strategy can be replicated with other would-be regional powers which claim to carry the region with them abroad but either stay too close to home or go about alone instead.

Figure 1 :
Figure 1: Number of drafts co-sponsored by Brazil and their share in total UNGA output

Figure 3 :
Figure 3: Total amount of drafts co-sponsored by Brazil per session and group

Figure 4 :
Figure 4: Regional support by South American and GRULAC countries to drafts with Brazilian co-sponsorship

Figure 5 :
Figure 5: Keyword frequency and share of Brazil-sponsored resolutions mentioning Latin

Figure 6 :
Figure 6: Number of drafts co-sponsored by South Africa and their share in total UNGA

Figure 7
Figure 7 builds upon the set of 2169 sequenced drafts and shows which countries

Figure 9 :
Figure 9: Regional support by countries from Southern Africa and the African Group to drafts with South African co-sponsorship

Figure 10 :
Figure 10: Keyword frequency and share of South African-sponsored resolutions