Secession, Territorial Integrity and (Non)-Sovereignty: Why do Some Separatist Movements in the Caribbean Succeed and Others Fail?

Abstract

Secessionist movements are ubiquitous in the Caribbean, with virtually every multi-island state and territory experiencing centrifugal tendencies. The region thus offers a unique opportunity to examine why some succeed and others fail. By and large, the propensity for secession has not attracted the attention of scholars beyond the region, with small states and territories largely excluded from supposedly ‘global’ analysis on the subject. The article fills this gap by analysing secessionist movements in both sovereign and non-sovereign territories. We find that secession was most likely to occur in the run-up to independence. In the post-colonial period, successes have only occurred among non-sovereign territories—as demonstrated by the fragmentation of the Netherlands Antilles and the administrative separation of St Martin and St Barthélémy from Guadeloupe—which have split from each other while collectively remaining part of a metropolitan state. Non-sovereignty reduces the costs of heterogeneity via the shelter provided by the larger metropolitan power. By analysing hitherto understudied cases, this article thus adds to studies that show how secession is contingent on continued state protections which allow downsizing to occur in an orderly manner, which is in turn consistent with the desire of the international community for geopolitical stability.

Introduction

The post-Second World War period has been defined by rapid expansion in the number of states and a corresponding fall in their average size, despite growth in total world population (Lake & O’Mahony, 2004). This trend has been termed the ‘Age of Secession’ (Griffiths, 2016), defined by ‘the voluntary withdrawal of a political territory from a larger one in which it was previously incorporated’ (Bauböck, 2019, p. 227). There are three dominant explanations for these trends: (1) culturalist accounts emphasise social identity, including language, ethnicity and religion; (2) rationalist accounts emphasise utility maximisation; and (3) capability accounts emphasise the ways in which institutional and geographic contexts enable and constrain individual decision-making (for review, see Hale, 2008, pp. 57–59; Lluch, 2014, pp. 20–24; Siroky et al., 2017, p. 304; Siroky et al., 2020). This article lends empirical support to the latter sub-set of explanations by showing how secession is contingent on continued state protections that allow downsizing to occur in an orderly manner that is consistent with the desire of the international community for geopolitical stability (Coggins, 2014; Griffiths, 2016). Specifically, in the Caribbean, secession was more likely to succeed in the run-up to independence. In the post-colonial period, successes have only occurred among non-sovereign territories that have split from each other while remaining tied to a metropolitan power.

In formulating this claim we make two contributions. The first is that the existing literature tends to focus heavily on the secession of a limited number of relatively large units from larger, powerful, western sovereign states (e.g. Scotland, Québec or Catalunya). Consequently, it suffers from an unstated gigantism—and western-centrism—that excludes many of the world’s smallest states and territories, which are considerably greater in number (Veenendaal & Corbett, 2015).¹ Second, and relatedly, it tends to overlook separatist movements within small non-sovereign territories, including some of those outside the West that are still on the United Nations (UN) decolonisation list. Indeed, we would expect that this type of downsizing is more common as concerns about viability are mitigated by continued attachment to the metropolitan state, thus reducing the costs of territorial rescaling. Analogous arguments can be made about secessionist movements in the EU. In both cases, the point is that the presence of multi-level constellations, which are often created to accommodate heterogeneity, can, in fact, encourage secession (cf. Kelle, 2017; Siroky & Cuffe, 2015).

These empirical gaps are mutually reinforcing: fifteen of the seventeen territories on the UN list are small islands, while independence movements in island territories are common in recently decolonised nations (e.g. the potential ‘Balkanisation’ of Indonesia or Bougainville’s desire to secede from Papua New Guinea). They also undermine the claim to a truly global analysis, both because small states are emblematic of the trend towards smaller political units, but also because we can often identify and analyse the mechanisms at work in any social or political phenomena more starkly in small-scale societies. In this sense, small states are ‘laboratories’ (Payne & Sutton, 2001; Reilly, 2002; Veenendaal & Corbett, 2015) for social science; they are of a similar size to a town, borough or county and can thus be studied with the same intimacy, but they also have to fulfil the obligations and responsibilities of statehood in analogous ways to larger countries. By the same token, small non-sovereign territories, which, despite their often-enormous geographical distance from the metropole and status within that polity equating to that of a mainland county, department, state or region, tend to embody the kind of ‘national’ political and cultural identity redolent of their independent neighbours (see Aldrich & Connell, 1998, 2020). Moreover, in all cases, given their pervasiveness, their experience is also arguably more representative than the handful of mainland secession movements in a limited number of large Western countries, and should therefore be viewed as integral to the wider literature rather than simply an addendum to it.

In this article, we focus on a region—the Caribbean—in which secessionist movements are ubiquitous in virtually all multi-island states. Indeed, it is not an overstatement to say that the region exhibits, relatively speaking, the largest, most sustained, and therefore, most comparable range of such movements at both the national and sub-national levels globally. Furthermore, although at the micro-level Caribbean societies are a mélange of historical and cultural diversity, those with strong secessionist movements tend to be linguistically, ethnically and religiously homogenous at the macro-level, meaning they defy classic assumptions made about heterogeneity in the existing (mainstream) literature. As such, the reticence of existing accounts to take them seriously not only represents a gap worth filling, but arguably skews that literature decisively by ignoring what are potentially the most compelling and most-similar cases. To be sure, whether these movements genuinely desire full statehood or greater autonomy is an open question (Baldacchino & Hepburn, 2012; Clegg, 2012). But that has not precluded the study of similar movements in other parts of the world.

To substantiate this claim, we take the following steps. First, we highlight how the literature on secession would benefit from paying greater attention to small states and territories. Second, we discuss how adding hitherto neglected cases to supposedly global analysis challenges key assumptions and sharpens core questions. Third, we provide an empirical analysis of Caribbean secession, distinguishing between movements in the English-, French- and Dutch-speaking Caribbean. The conclusion reviews the key findings and reflects on the wider implications of our argument about the importance of institutional context in explaining why some movements succeed and others fail.

The Age of Secession

Small states and territories in the Caribbean are especially puzzling because their experience tends to debunk two of the dominant theories for secession: culturalist and rationalist explanations. The former emphasises linguistic, ethnic or religious heterogeneity. But Caribbean states tend to have highly homogenous populations when measured in terms of identity-based difference (Anckar, 1999). To be sure, there are differences between islands that reflect historically constituted ethno-sociological diversity, but this does not challenge that general picture. Likewise, the latter theories struggle because diseconomies of scale mean that small communities have the most to gain from economic integration and forms of semi-autonomous government (Alesina & Spolaore, 2005). As we will show, a capability explanation best accounts for the experience of Caribbean small states and territories as it foregrounds the geographic and institutional contexts in which secessionist movements occur (cf. Siroky et al., 2017). In doing so, it echoes LaFlamme’s (1983) observation decades ago that a common feature of virtually all ‘archipelago states’ is that the peripheral islands are much poorer than the main island and the inevitable consequences of this disparity are centrifugal tendencies and island nationalism (cf. Byron, 1999; Corbett, 2020).

In making the claim that secession is contingent on continued state protections that allow downsizing to occur in an orderly manner that is consistent with the desire of the international community for geopolitical stability, we build on Griffiths' (2016) theory of ‘metropolitan response’. He argues that system-level constraints guide how states respond to secessionist demands, but internal structures, including administrative lines and categories, determine how they contract and the manner in which they do so. The post-Second World War order provided a permissive environment for small state decolonisation and ongoing security, despite limited military capacity due to the twin norms of self-determination and territorial integrity (cf Sharman, 2017). However, the administrative status of breakaway regions is the key factor in determining the success of secession. This is because, while states are loath to part with territory, administrative boundaries create recognisable categories that both focalise secessionist demands and create an evident hierarchy of territory—core versus periphery—to each state. This is especially so in small islands, where the sea inherently demarcates clear legal, political, cultural and psychological borders, which is one reason why regional integration has generally faltered in the Caribbean (see, inter alia: Bishop, 2014; Bishop & Payne, 2010; Clarke, 1976; Payne, 2008). Such boundaries therefore also allow for secession to occur in an orderly manner that is consistent with the desire of an international community of states for system-level stability.

This argument has affinities with Lecours' (2020, p. 4) distinction between static and dynamic autonomy. For Lecours, the former is more likely to precipitate secession than the latter because the political settlement can be adjusted as circumstances change. The upshot is that, in dynamic institutional configurations, communities are less likely to push for full statehood because they have a reasonable expectation that they might win further autonomy in the future. Likewise, Siroky et al. (2020) capture the importance of dynamism in their account of how indirect rule, a system in which local notables who profit from the status quo seek to dissuade their communities from pursuing outright nationalism, mitigates against secessionism. The point is that the institutional and geographic context in which these movements take place is key to explaining why some succeed (rather than simply exist without provoking a decisive rupture).

We test these arguments against hitherto neglected cases of both sovereign and sub-national secession. Secessionism is especially pronounced among small islands, although the implicit aim of many of these movements is often greater autonomy rather than full independence (Baldacchino & Hepburn, 2012). Scholars of the Caribbean have produced a substantial body of literature that speaks to global themes (e.g. Bishop, 2013; Byron, 1999; Clegg, 2012; Lewis, 2013; Payne & Sutton, 2007). However, it has generally been ignored by mainstream political science, which favours large quantitative databases that tend to exclude the smallest states (Veenendaal & Corbett, 2015). This empirical neglect is lamentable, both because these cases are interesting and important in their own right, and also because the existence of small states is emblematic of the global trend towards smaller and smaller political units.

Data and Method

This article draws on empirical examples of secessionist movements in one of the regions home to a high concentration of small states and territories: the Caribbean. Secessionist movements or discourses on increased autonomy have arisen at times in almost all multi-island independent Caribbean states, as outlined in Table 1 below. They have also had varying levels of success and attracted the full spectrum of metropolitan response—from armed repression to referenda. They thus provide considerable variation, but within a region of similarly small island states and territories. Further comparative research could expand our study to other regions of small states—the Pacific Islands region, for example—and shed new light on previous quantitative analysis by including all states. But for now, we take the intermediate, yet still fundamental, step of demonstrating why these cases have much to offer scholars of mainstream political science.

Table 1. Cases of successful and unsuccessful attempts at secession covered

Case	Era	Characteristics
Britain and the ‘Commonwealth’ Caribbean	1962 1962 1966 1967 1959/1973 1976–1986 1998	• Jamaica and Trinidad seceded from the West Indies Federation to become independent, precipitating Federation’s collapse. • Cayman Islands leave Jamaica to become separate Crown Colony. • Barbados became independent from Britain (after collapse of Federation). • Anguilla seeks independence from British Associated State of St Christopher (St Kitts) and Nevis. Separation is legally formalised in 1980. • Turks and Caicos Islands separate from Jamaica and Bahamas respectively. • Tobago begins ongoing campaign for greater autonomy from Trinidad. • Failed secession referendum on Nevis (from St Kitts).
The Dutch Caribbean	1986–1996 2010	• Secession of Aruba from the Netherlands Antilles, and unsuccessful attempt by Dutch government to provoke full independence. • Dissolution of the Netherlands Antilles Bonaire, Curaçao, Saba, St Eustatius, and St Maarten, and creation of new non-sovereign entities vis-à-vis the Netherlands with varying degrees of autonomy.
The French Caribbean	1990s+ 2003+	• Intensified political independentism in Martinique (and, to a lesser extent, Guadeloupe) in partially successful pursuit of greater autonomy. • Separation of St Martin and St Barthélémy from Guadeloupe.

The creation of colonial administrative units that privileged the needs of metropolitan powers rather than the social and political organisation of colonised peoples is a common basis of secession all over the world. The Caribbean is no exception to this pattern, although it is distinctive in certain ways. First, because its colonial experience was totalising—almost everywhere, pre-existing civilisation was effectively wiped out and new societies oriented to serve extractive interests were created by the establishment of plantation slavery—this decisively shaped subsequent political regimes (Bishop, 2013; Bishop et al., 2020). Pre-colonial fragments do remain, with small indigenous communities in Dominica, St Vincent, Trinidad and Belize, and most societies are a highly diverse mix of European, African, Asian and, in some cases, East Indian peoples—the latter’s ancestors arriving as indentured labourers after slavery—such that, at the microscopic level, they are ethnically complex melting pots practising religions including Obeah, myriad variants of Christianity, Islam and Hinduism. This is further reinforced by small-island insularity that exaggerates and reinforces a self-image of difference. As Payne (2008, p. xxxiii) has noted in relation to Carriacou, one of Grenada’s even smaller constituent territories: not only is it ‘typically conscious of its own individuality in relation to Grenada, but … the people of its dependency, the even tinier Petit Martinique, identify scarcely at all with the people of Carriacou’. Nevertheless, despite their micro-level differences, West Indian societies are strikingly homogenous at the macro-level, and this is key to sustaining our methodological agenda. Most are majority Afro-Caribbean, and even when not, they are still recognisably and uniformly creole in social structure and culture; everybody generally buys into ‘national’ identity or, if not, their insular island identity (thus reinforcing our broader argument); and, even in those places where people practise a kaleidoscope of religions, almost everyone is religious and usually celebrates each other’s holidays and rituals.

Second, historically, territories varied markedly in size and economic potential: they were constantly fought over, often changing hands multiple times; they were developed to starkly differing degrees over more than two centuries; and these processes waxed and waned alongside the rise and decline of sugar and the capricious engagement of respective metropoles. Third, and consequently, although they exhibit many generalised patterns reflective of this broadly shared history, individual governing systems embody idiosyncrasies that derive from the contingent exercise of power by local colonial officials, influential plantocrats or metropolitan governments at particular critical junctures. Finally, then, perhaps more than any other part of the world, these longstanding legacies decisively shaped the specific nature of decolonisation, and, later, local and regional politics in the modern era, ultimately determining to a substantial extent the contours in which secessionism would—or would not—play out.

Our analysis is qualitative. We sequentially trace the trajectory of secessionist movements within each of the multilevel constellations apparent in the region. Although we acknowledge that all of our cases have experienced a bewildering array of administrative arrangements over centuries of European colonisation, we primarily focus on the post-Second World War period as the most relevant to the ‘Age of Secession’. We also accept that, while there is considerable variation among the experiences of these islands, our focus is unequivocally on identifying commonalities between their movements. In this sense, we do not aim to offer a holistic explanation of each individual case but rather to identify patterns that hold across them as a group. All are ‘most similar’ because they are small Caribbean islands, but within that category, they are also ‘most different’, and, following that logic, we claim that any commonalities we identify are likely to have more general relevance.

Secession in the Caribbean

In this section, we discuss the evolution of secessionist processes in the three sub-regional groupings and key episodes identified in Table 1: the Anglophone or ‘Commonwealth’, Francophone, and then Dutch-speaking Caribbean.

The Commonwealth Caribbean

In managing its Caribbean colonies, Britain generally established arms-length Crown Colony governments that were essentially a vessel for plantocratic interests, with local governors wielding enormous residual power. After slavery was formally abolished in 1833, little changed substantively for a hundred years: labouring people still worked on plantations in ‘serflike conditions’ (Brereton, 1981, p. 85) and few political rights were granted beyond the narrow franchise enjoyed by a handful of powerful, usually white, families. As Austin (1980, pp. 3–4) once put it, Britain’s approach was decidedly ‘non integrative’ as embodied in odd notions like the ‘self-governing colony’. In contrast to the assimilation of the French islands—discussed below—he adds that ‘there was to be no L’Angleterre d’Outre-mer’. However, this relative autonomy on the part of (some) local actors reflected a dysfunctional mix of British administrative incompetence and negligence, which in turn derived from a general laissez-faire approach to empire, a sense that empires themselves were impermanent, and a dawning realisation that the Caribbean colonies were becoming a financial burden that had outlived their extractive usefulness (Bishop, 2013). Throughout this period, larger colonial administrative units such as the Leeward or Windward Islands were re-consolidated and, within them, smaller islands forced together with larger ones: examples include Barbuda with Antigua, Anguilla and Nevis with St Christopher (St Kitts), Tobago with Trinidad—which meant that Tobago left the Windwards in 1889 to become part of a distinctive colony that Britain had long been perplexed about how to govern (Millette, 1970). Yet London never provided the resources to infuse these reconfigurations with developmental purpose: by the late nineteenth century, as malign neglect intensified, Disraeli described territories that had once generated unimaginable imperial riches as ‘wretched colonies’, ‘millstones round our necks’ and ‘colonial deadweights’ (cited in Williams, 1970, p. 399).

Nonetheless, the Crown Colony system of imperial governance stumbled on into the twentieth century, until the West Indies Federation was established in 1958 with the express desire to gradually decolonise the islands, considered unviable individually, as a collective (Mordecai, 1968). Britain promoted the federal structure as an enabling mechanism which might allow the large grouping of small territories with limited capacities to jointly navigate the waters of sovereign statehood, initially under London’s tutelage. However, nationalist actors in the larger territories of Jamaica and Trinidad saw the Federation as an interim vehicle towards their real goal of unitary sovereign statehood. These were elite perceptions with scarce regard for the populations of the various island territories: the discourse of the press and politicians constructed the smaller islands as potential economic and social burdens. The Federation, if resourced and institutionalised adequately, could genuinely have underpinned regional development and vibrant federal governance. Yet Britain’s commitment was neglectful (Mawby, 2012). Moreover, personal antagonism among the Federation’s top leadership and island nationalist ideologies combined with these other dynamics to bring about Jamaica’s secession referendum in 1961, followed by the withdrawal of both Jamaica and Trinidad and Tobago in 1962, leading to its collapse (Boxill, 1993; Lewis, 1968; Thomas, 1988). There was a short-lived attempt thereafter to rebuild a federal ‘Little Eight’ with Barbados, the Leeward and Windward Islands, but that never really materialised, and Barbados subsequently became independent in 1966 (Lewis, 1965, 1968).

The 1967 West Indies Act sought to provide an interim solution, Associated Statehood, for the remaining members of the Federation—Antigua; Dominica; Grenada; St Lucia; St Kitts, Nevis and Anguilla; and St Vincent, essentially the six independent members of today’s Organisation of Eastern Caribbean States (OECS)—which provided for domestic self-government but left Britain to take care of foreign policy and security. Separate to this group were a number of even smaller islands which generally were not members of the Federation and largely comprise today’s non-sovereign British Overseas Territories (BOTs). The belief was that, while the Associated States were not yet ready for full statehood, it might become a viable option once a degree of economic development had been obtained. They would all ultimately gain independence between 1974 and 1983. During these negotiations, the smaller territories that comprised parts of multi-island Associated States, like Anguilla, Nevis, and Barbuda, protested strongly against the 1967 Act. Anguilla took its grievances furthest: it had been jointly administered as a colony with St Kitts since 1822, with the latter reluctantly agreeing to the union with the proviso that it would not require the redistribution of resources to the smaller island (Phillips, 2002). Anguillans thus came to see themselves as a ‘double colony’ (Petty, 1984, p. 14) and sought devolution prior to the formation of associated statehood. But the mooted Anguilla Council was to receive limited budgetary support and minimal responsibility for island affairs (Clarke, 1976, p. 11).

Anguilla felt too small to go it alone and ultimately preferred the remote, relatively benign, authority of Britain to the closer but harsher rule of St Kitts. When London ignored its demands for autonomy by creating the unitary state of St Kitts-Nevis-Anguilla, Anguillians launched a rebellion on 30th May 1967 in the hope of retaining their link with the Crown. Initially, Britain refused, in part because it did not wish to set a precedent, but also because it felt obliged to uphold the Act which guaranteed territorial integrity to the new Associated States. Under pressure to reunite with St Kitts, Anguilla declared independence unilaterally in July 1967 (Petty, 1984, p. 44). The leader who oversaw this process, Ronald Webster, would later reflect to the New York Times (cited in Treaster, 1988) that the declaration of independence was:

… just a matter of strategy, a way of ‘twisting their arm’. We thought that if we made that move, Britain would be forced to come in and deal with the situation … [and] It actually happened.

Fearful that the island would become a haven for illicit activities, Britain sent in paratroopers on 19th March 1969. After lengthy negotiations, Anguilla become an Overseas Territory in 1980, precisely what its people had desired all along. As a BOT, locally elected officials manage domestic affairs, overseen by a British Governor, with security and foreign relations outside the region also handled by Britain.

Anguilla was not the only example of military force being deployed to end secessionist movements in the region. The 1979 rebellion on Union Island in the Grenadines was led by Rastafarian youth protesting against tourism development. The revolt was short-lived, but because it occurred only months after the Grenadian revolution, it prompted a Barbadian expeditionary force, which reflected both the concerns of Caribbean governments about threats to territorial integrity and those of Washington and London about growing militancy and anything resembling a socialist ideology (Campbell, 1980, p. 48). Secessionist tensions are, though, no longer a strong feature of politics in St Vincent and the Grenadines. One explanation for this is that, from 1984 to 2000, James Mitchell, then Prime Minister, came from the island of Bequia in the Grenadines and focused on peripheral infrastructural development. Moreover, some remaining Vincentian Grenadine islands are controlled largely by external interests: Mustique is a private enclave for the rich, and Canouan’s political dynamics have involved balancing local interests and resort tourism which dominates around half of the territory, rather than the state itself (see Bishop, 2010, 2013).

Nevis is the island that has come closest to secession in the post-colonial period. Its leaders sought separation from St Kitts in the late 1970s and early 1980s. The islands are separated by a few kilometres of the Caribbean Sea that can be crossed in a short ferry ride. But, despite the proximity, Nevis, like Anguilla, has always had reservations about administration by St Kitts and its leaders were strongly opposed to becoming independent as part of a unitary state (Byron, 1999; Midgett, 2005; Premdas, 1998). In particular, they were fearful of being ruled by the dominant St Kitts and Nevis Labour Party (SKNLP) which had held power continuously on the islands since 1952, and its erstwhile leader, Robert Bradshaw, who was the first Premier of the Associated State between 1967 and his passing in 1978 after initiating independence talks with Britain. The numbers in the pre-independence legislature meant that the SKNLP could rule without winning any votes on Nevis (or Anguilla). The consequence, Nevisians argued, was systematic—even ‘neo-colonial’—disdain for their economic development (Midgett, 2005, p. 45). As an MP from the island, Simeon Daniel, wrote:

The people of Nevis have grown accustomed to the neglect, spite and disregard for their political, social and economic welfare meted out to them over a quarter of [a] century at the hands of your Government … but we will not be beaten into submission. (cited in Midgett, 2004, p. 56)

By contrast, the SKNLP viewed Nevis as a burden and resented any concessions that granted additional autonomy. This impasse was only overcome after Bradshaw died. In the 1980 election, a new coalition government, led by Dr Kennedy Simmonds of the People’s Action Movement, included MPs from Nevis in key Cabinet posts. The coalition took the country to independence as a federation in 1983 but with a clause inserted in the constitution (Section 113) that provided Nevis with the ability to secede if it obtained a two-thirds majority in the Nevis House of Assembly and a two-thirds majority in a referendum on the island itself (Byron, 1999). The expectation at independence was that Section 113 would be triggered within 18 months.

While the Simmonds coalition government held and Nevis had a voice in Cabinet, its leaders saw no rush to secede. But, from the late 1980s, the coalition began to fray, and Nevisian politicians stated publicly that they were considering the secession alternative guaranteed to them in the constitution (see Byron, 1999). In 1998, the island held a referendum on the subject, which most observers believed would pass the 66 per cent threshold. Yet, the secessionist movement fell a few hundred votes short, winning 61.7 per cent of the vote in a 58 per cent turnout (Byron, 1999). The ‘yes’ vote was nevertheless significant enough to wield considerable influence in the ongoing national debate on constitutional reform (Byron, 1999). Since then, the movement has not gone away. However, successive coalition governments comprising leaders of all the parties in Nevis, changing political dynamics and power configurations, as well as closer intertwining of the economies of the two islands, have served to dampen secessionist tendencies (at least for now). Nonetheless, every Nevisian political party still genuflects to a hypothetical secession, even though it has never been invoked in practice.

A similar story might be told about Tobago. It holds just two of 41 parliamentary seats, its 61,000 people comprise only 4.5 per cent of the Republic of Trinidad and Tobago’s population, and they are overwhelmingly Afro-Caribbean, whereas Trinidad’s 1.3 million reflects a uniquely diverse ethnic mosaic (Thompson, 2021). The marriage of the two islands in the late nineteenth Century was, in a historical sense, one of convenience: according to Bridget Brereton (1981, pp. 153–4), this was purely the result of ‘imperial fiat’ and ‘the outcome of Britain’s anxiety to shuffle off responsibility for an impoverished little colony by tacking it onto a more prosperous one’. As the country’s first post-independence PM, and celebrated historian, Dr Eric Williams (1962, pp. 51–2) described it, Tobago had always suffered from ‘betweenity’ in that it was a ‘no mans’ land’ transgressed by a ‘never-ending free for all on the part of the colonial powers’ in which the ‘bankruptcy of colonialism reached its nadir’. It changed hands frequently, yet no metropolitan power ever committed the resources to developing or securing it, and there were regular upheavals in governance. Its forbidding geographical position—socially and culturally contiguous with, but distant from, the Windward islands, yet closer to, but lacking contiguity with, Trinidad—has always reflected these legacies. As Trinidad increasingly industrialised and urbanised, the smaller island struggled to shake off its history of rural poverty or find reliable methods for generating foreign exchange (Premdas & Williams, 1992). This accentuated the sense of difference and dissatisfaction with the post-independence settlement (Luke, 2007; Ryan, 1985).

By 1976, calls for greater self-government began, and genuinely Tobagonian political parties began to emerge. After splitting from the People’s National Movement (PNM)—which had dominated Trinbagonian politics between the mid-1950s and was the vehicle that saw Williams elected Prime Minister (PM) three times—Tobago MP and former PNM Deputy Leader ANR Robinson established the Democratic Action Congress (DAC). This quickly established a monopoly over the political scene on the smaller island: as Meighoo (2008, p. 119) later put it, when ‘Robinson left the PNM in 1970, the Tobagonians essentially left with him’. Perhaps ironically, Williams had previously described the 1889 decision to declare Tobago a ‘ward’ of Trinidad as a ‘humiliation’, and famously argued in the Legislative Council in 1957 that ‘Tobago had exchanged the neglect of United Kingdom Imperialism for the neglect of Trinidad Imperialism’ and, in turn, had ‘had to pay a price for its union with Trinidad which it never ought to have paid’. Yet he would later be accused of doing little to rectify this situation as PM of the unitary state (Dumas, 2012). In tandem with the economic travails of the 1970s and concomitant fragmentation in the political coalition that had underpinned the PNM’s electoral pre-eminence during the first two decades of independence, Robinson became the country’s first non-PNM and Tobagonian PM at the head of a rainbow coalition under the ‘National Alliance for Reconstruction’ (NAR) banner.

The NAR only served one term, which began with a global collapse in oil prices and ended with a short-lived coup by the Jamaat Al-Muslimeen in response to deleterious social conditions. Robinson nonetheless oversaw the reinvigoration of the Tobago House of Assembly (THA) as a meaningful site of autonomous government and was able to direct a larger share of national budgetary resources to the island to facilitate infrastructural development. This was historically fortuitous: Tobago had previously been a self-governing, rather than Crown Colony, with its own bicameral legislature established in 1768 (Williams, 1962, p. 130). Although the franchise only extended to a handful of plantocratic families—just 102 people from a population of 12,000 in 1857 (Brereton, 1981, p. 154)—the assembly itself lived on beyond the colonial era as a space where distinctly Tobagonian politics could find expression.

Since the early 1980s, when the THA Act was passed, local politics has taken place within this framework and the island has not seriously pursued secession from Trinidad. In part, this is because it has continued to seek, with some success, greater autonomy: a new THA Act was passed in 1996 which further expanded the House’s powers and responsibilities, its economic influence and its staffing capacity. Moreover, even in national elections, it tends to vote in step with Trinidad: in 2010, both Tobago seats were won by the Tobago Organisation of the People (TOP) as part of an opposition rainbow coalition (see Bishop, 2011); in 2015, similarly, when familiar patterns reasserted themselves and the PNM led by Dr Keith Rowley (himself a Tobagonian, although representing a Trinidadian constituency) won office once more, it also took both seats in Tobago, and again in 2020. Rowley’s first administration drafted a new bill in 2018 which envisaged even greater decentralisation, but this was stymied by insufficient support in parliament. In sum, despite the desire for further autonomy, Tobagonians generally wish to remain within the unitary state of Trinidad and Tobago, but have long remained dissatisfied with the current system and have sought continued devolution (Dumas, 2012).

Barbuda is the smaller island in the state of Antigua and Barbuda, and many of its people aspire to follow the Anguilla and Nevis examples. Its leaders sought a similar arrangement to Anguilla during independence negotiations with the UK, arguing that:

Under the former slave master … at least our bellies were full … Under British colonialism we were neglected; but at least when our people made representations they were listened to and some effort was made to accede to our demands. The land was ours and there was minimal interference with our way of life. Under Antiguan colonialism, the neglect remains, our bellies have become empty, suppression and repression have added to our unhappy lot. (cited in Lowenthal & Clarke, 1980, p. 301)

However, Britain was strongly opposed to separating Antigua and Barbuda. Indeed, the Anguilla experience even appeared to harden its view about the creation of micro-polities (Lowenthal & Clarke, 1980). Secessionist tendencies nonetheless persisted under the long tenure of the Antigua and Barbuda Labour Party’s Vere and Lester Bird. A change of government in 2006 provided an opportunity for Barbudans to increase their autonomy via the 2007 Barbuda Land Act. But, after Hurricane Irma forced the evacuation of the island in 2018, the Act was repealed, and the national government sought to reassert its control of Barbudan affairs (see Corbett, 2020). This has reignited the desire to secede, but also rendered secession harder to envisage due to the scale of devastation wrought by the hurricane and waning local control over land and resources. Barbudan activists concede that they are unlikely to become fully independent, but they retain hope that one day they might be reconnected with the UK. Failing that, they aspire to increase their autonomy in a similar way to Nevis, but this remains difficult to envisage within the unitary state of Antigua-Barbuda: St Kitts-Nevis, by contrast, is a federal state, the only one in the Anglophone Caribbean.

In sum, a demand for greater independence or self-governing autonomy by the smaller island of a multi-island state is a feature of most Anglo-Caribbean cases. These movements were vocal around independence but, concerned about the viability of very small states, Britain was intent on avoiding fragmentation. Anguilla managed to realise its desire for continued (or, rather, reconstituted) metropolitan dependence, as did both the tiny Cayman Islands and the Turks and Caicos Islands (TCI) which today are also BOTs. Britain formally attached the Caymans to Jamaica in 1730 but there had been a less formal attachment since the Treaty of Madrid ceded both to Britain (from Spain) in 1670. They gained a certain amount of autonomy over time, especially as of 1959, when the Governor of Jamaica appointed an Administrator to them. However, formal separation only came in 1962: as Jamaica prepared for independence, Caymanians indicated a preference for remaining with Britain, and so the islands were designated as a separate British colony. The TCI, further away, had a more unstable history of attachments. They were governed from Nassau in the 1700s, and allowed to separate from the Bahamas in 1848. They were then governed from Jamaica as of 1873, and this status continued until 1959 when they were also designated as a separate Crown Colony. In 1965, they again fell under the administrative jurisdiction of the Governor of the Bahamas until the Bahamas became independent in 1973, before becoming a standalone British dependency.

The broader lesson, then, is that it was easier for a small territory that had been jammed together with a larger unit, to separate before the latter became independent, and claim the extant option of remaining as a British dependency. Elsewhere, aside from Anguilla, amongst those states that achieved independence or greater degrees of self-government in the aftermath of the collapse of the West Indies Federation, the smaller island acquiesced, often reluctantly. In all of these places, the desire for greater self-governing autonomy has not gone away, but secessionists have had to settle, at best, for circumscribed forms of devolution. At the micro-level, a range of factors—e.g. population size; geographical distance; economic resources; support from diasporas or expatriate communities; the evolution of local political dynamics; domestic institutions, including the extent to which they are reconfigured to reduce tensions; and the role played by regional and international organisations—influence the everyday waxing and waning of autonomy struggles. But, in the final analysis, the key macro-level lesson from the Anglophone Caribbean is that the timing of the secessionist attempt, and the lengths that a territory is prepared to go to achieve it, are the key determinants for success.

The French Caribbean

There are deep-seated paradoxes at work in Martinique and Guadeloupe. As conventionally measured, they enjoy the highest material living standards in the region, yet there is widespread dissatisfaction with the integrative post-colonial settlement that has produced this outcome (Bishop, 2013). They embody a strong creole political and cultural identity that overlaps contiguously with the wider Caribbean, yet they are largely isolated from the region in terms of language and shared governance (although this is slowly changing as Martinique and Guadeloupe have explored their expanded regional cooperation options in recent years). Perhaps most important for our purposes, there is deep apathy towards French and European politics as evidenced by consistently poor turnout in elections and often-strong electoral support for avowed political independentists. Yet, despite some tinkering with the nature of the decentralised political settlement that has endured since departmentalisation in 1946—meaning decolonisation by integration into France (see Murch, 1968)—substantive support for real independence, and even meaningful autonomy, which is difficult to envisage within the contours of the ‘one and indivisible’ French Republic, consistently founders at the polls. However, within Guadeloupe, there has been a significant degree of secessionism in the Iles du Nord (Northern Isles), which have sought their own autonomy from the administrative territory into which they had previously been placed.

So, what explains this picture? As in the Anglophone Caribbean, the distinctive contours of French colonialism not only produced a general—albeit contrasting—pattern of assimilation, but also significant differences between Martinique and Guadeloupe themselves. This was further complicated by ongoing upheavals in France and its two distinct personalities, which affected them differentially at key historical junctures: i.e. the ‘good France’ of the Enlightenment, Revolution, emancipation, republicanism and resistance, and the ‘bad France’ of the Bourbons, slavery, the Napoleonic Empire and Vichy (Hintjens, 1992). The territories were settled decisively in the early-to-mid seventeenth Century, and, although they represented the same kind of plantation societies as their British counterparts, aside from occasional episodes of temporary external occupation, they have remained to all intents and purposes—culturally, linguistically and politically—French. Moreover, when France finally abolished slavery in 1848, twelve years later than Britain, the people of the French Caribbean were simultaneously made French citizens—albeit in a restrictive and problematic fashion (see Church, 2017)—decisively shaping the assimilationist contours of their future relationship with the metropole.

After the Second World War, these territories (along with French Guiana on the South American mainland and Réunion in the Indian Ocean) were immediately and fully integrated into France as Overseas Departments (DOMs): in part due to the assimilationist logic of French Republicanism and in part the activism of local figures, notably the celebrated intellectual and politician Aimé Césaire of Martinique (cf. Wilder, 2015). Departmentalisation, though, has always been contested. This was partly assuaged by Mitterrand’s decentralisation reforms during the 1980s, in which Departments across France were corralled into 22 ‘Regions’. The DOMs, unusually, became ‘mono-departmental regions’ which meant that, rather than combining them together as a single region, they each retained both a departmental General Council and a Regional Council. This institutional state of affairs remains to the present day in Guadeloupe, while in 2015, Martinique and French Guiana, in accordance with referenda held in 2010, merged their administrations into single territorial representative assemblies with executive cabinets, the Collectivité Territoriale de la Martinique (CTM) and the Collectivité Territoriale de la Guyane (CTG).

Nonetheless, the mono-departmental region was a crucial development, encouraging the emergence of a distinctly ‘national’ brand of local politics, along with the rise of non-metropolitan political parties and personalities (Miles, 2006). In Martinique, especially, it facilitated the rise of Alfred Marie-Jeanne’s Mouvement Indépendantiste Martiniquais (MIM) which espoused a broadly leftist agenda similar to the Parti Progressiste Martiniquais (PPM) of Césaire, and came to dominate the electoral landscape throughout the 1990s and 2000s until the PPM took power once more in 2010. The MIM won elections, often by wide margins—and enjoyed substantial resources for local development from France and the European Union’s (EU) structural funds (see Daniel, 2002, 2009)—yet, paradoxically, by participating in this way it effectively legitimised the continued departmentalisation to which it was supposedly opposed, weakening its claims to a genuine nationalism (Giraud, 2005). Guadeloupe has not had a similarly effective strain of electoral independentism; its politics have long been dominated by local outposts of mainland parties (Daniel, 2002). Due to a unique set of historical legacies—its experience of both the French Revolution and the Vichy regime were quite different to Martinique, for example, and heavily influenced the contours of its subsequent social settlements and political culture (Bishop, 2013)—it has a ferociously militant trade union movement which can mobilise large numbers of people to protest the status quo, through which anti-establishment antagonism is often channelled. This was particularly evident in the lengthy shutdown of 2009, during which massive street demonstrations took place regularly (Bonilla, 2015).

Through Article 73 of the French Fundamental law of 1958, the DOMs have always had the right to ‘adapt’ laws to local circumstances (Reno, 2004). However, changes have tended to be constitutional and symbolic rather than substantive and material (Mrgudovic, 2012). As Daniel (2009, p. 130) has argued, on the part of both Paris and local elites, there is ‘a functionalist and circular determinism’ which always posits institutional innovation as the solution to the islands’ existential developmental predicaments (Daniel, 2009, p. 130). The rupture that independentists called for in the 2000s was never seriously contemplated, and legal tinkering to produce a limited degree of greater autonomy along with extra public money for assuaging the social challenges that, in large measure, derive from the economic pathologies inherent in the departmentalisation model itself, has been the result (Bishop, 2013). Reno (2004, in press) and Daniel (2002, 2009) have described this as a ‘deliberate’ development strategy of ‘resourced dependency’. Even when new settlements have been proposed—after 1999 the DOMs considered switching from Article 73 (which governs their specifically departmental status) to a new constitutional Article 74, which would have provided more autonomy and combined the General and Regional Councils into the single Overseas Collectivities (COM) (Clegg, 2012)—these proved extremely controversial despite the earlier discourse of independentism. Indeed, the referenda that took place in December 2003 saw ‘no’ votes for Article 74 in both Martinique and Guadeloupe, albeit narrowly in the former (50.48 per cent) and resoundingly in the latter (72.98 per cent) due to the fact that ‘not even vocal advocates of the reform were comfortable with the undefined implications’ (Miles, 2006, p. 643). Electorates were concerned about possible reduced access to the social benefits derived from French state budgetary support. Yet just four months later, in March 2004, Martinican voters returned Marie-Jeanne’s MIM to the Regional Council with its first-ever absolute majority, a striking achievement under a proportional system.

In Guadeloupe, independentism has little electoral purchase, so local identity has not been translated into a mainstream political programme and instead plays out through the often-dramatic levels of agitation undertaken by the unions and, as of 2009, the Liyannaj Kont Pwofitasyon (LKP)—in creole, meaning ‘Alliance Against Profiteering’—which comprises myriad social movements and unions under a single umbrella. This fills a vacuum in conventional politics: as social demands flare up, they are usually resolved by the provision of greater metropolitan resources, since substantive political independence could neither solve the social crisis nor enjoy traction amongst the population at large. Ironically, the claims which are often made carry with them an inherent integrationist logic: as demonstrated vividly in the 44-day general strike of 2009, which spilled over into Martinique and French Guiana too, they encompassed demands for a broader distribution of the island’s wealth, much of which is today derived from French and European fiscal transfers, including employment subsidies, tax cuts and so on. In this sense, attempts to rectify deep-seated social pathologies serve to reinforce the economic status quo that gives rise to them (Bishop, 2013). It should be noted that the alignment of various French Caribbean political parties on both the Left and the Right with French political parties has resulted in an unprecedented level of ministerial representation in French state governments since 2009—there have been three Ministers of Overseas Development and one Minister of Sports from Guadeloupe in the Sarkozy, Hollande and Macron governments. Likewise, French Guiana’s Christiane Taubira was Minister of Justice for a time during the Hollande administration.

The 2009 crisis ultimately led to another round of constitutional tinkering, and a familiar paradox reasserted itself: in January 2010, new autonomy proposals were resoundingly rejected in Martinique, with 79 per cent voting against (albeit with a 55 per cent turnout). But, shortly afterwards, a majority of 68 per cent assented to ‘the less significant reform of merging the Conseil Régional and Conseil Général into a single body’ [i.e. the aforementioned CTM] (Clegg, 2012, p. 435). Just two months later, the MIM were comprehensively routed in the regional elections by the PPM due to the former’s implication in the poor management of the 2009 crisis. During its term in office, Serge Letchimy’s PPM campaigned vigorously for Martinique’s greater autonomy in its development policies and elaborated a highly visible regional engagement profile, spearheading negotiations with both Paris and the OECS for associate membership of the body. This was formally realised in 2016.

In 2015, the CTM came into being, and the MIM once again took office: Marie-Jeanne became its first President, oddly at the head of a list of right-wing parties, defeating a left-wing list headed by Letchimy, the outgoing (and final) Regional Council President. Again paradoxically, the pace of regional engagement slowed considerably under this supposedly independentist administration. In Guadeloupe, the left-wing party, Guadeloupe unie, solidaire et responsible (GUSR) won the Regional Council in 2015, defeating the mainland Parti Socialiste which had controlled it, with Victorin Lurel serving almost interrupted, since 2004. The new administration of Ary Chalus reopened the suspended negotiations with the OECS in 2017 and Guadeloupe became an Associate Member in 2019. In sum, despite avowed secessionist pressure, especially in Martinique, Overseas France has not seriously contemplated actual independence. Instead, the islands have used the 2000 Loi de l’Orientation de l’Outre Mer (LOOM), and the 2009 Loi de l’Orientation du Developpement Economique de l’Outre Mer (LODEOM), to push the boundaries of autonomous action and cultural identity in a constant process of negotiation with the French state on forms of regional engagement and the expansion of their control over development funding.

However, some smaller units have sought to secede from Guadeloupe. In the 2003 proposal, resoundingly defeated in Guadeloupe proper, both St Barthélemy and St Martin (which shares an island with Dutch St Maarten) voted by huge margins—96 per cent and 76 per cent respectively—to cease being Communes of Guadeloupe in order to become COMs themselves under Article 74 of the Constitution. In addition to the distance factor, they exhibit distinct historical and sociological characteristics, retain very local identities and concerns, and aspired to greater autonomy in economic development matters. In both cases, this act of secession—which took effect in 2007—simultaneously offered both better institutions and greater autonomy with the establishment of Territorial and Executive Councils and an elected President, while also embedding them more deeply as dependent units within the French state (and, by implication, insulating them against any future Guadeloupean moves towards independence). They were combined as a single legislative constituency for the purposes of electing one Deputy to the National Assembly in Paris and they independently elect one Senator each. Interestingly, they opted for different types of relationships with the EU: St Martin chose to retain ‘Outermost Region’ status (see Europa Publications, 2020) and the full benefits of economic and social financing that this implies, while in 2012 St Barthélemy became an EU Overseas Country and Territory (OCT), not governed by Community Law, nor membership of the European Single Market, and with no structural funding, but greater economic autonomy (the same status enjoyed by the BOTs before Britain left the EU in 2020). Their cases attest to the opportunities for diverse nested arrangements that exist in the complex multilevel governance systems of France and the EU.

In sum, the decolonisation by integration model almost-inevitably produces tensions and secessionist pressure: Martinique and Guadeloupe enjoy high aggregate standards of living that mask economic dysfunction and deeply exclusionary patterns of development (Bishop, 2013). Yet, although independence is regularly posited by certain political elites—and, in Martinique, has enjoyed periods of electoral traction—the populace remains extremely reticent about substantive change to the governing order. Where shifts have occurred, these involve significant—in the French constitutional sense—reforms, but these are inherently unable to assuage the pathologies afflicting the social order. Nonetheless, the establishment of the COM in Martinique and the greater freedom given to the President of the Regional Council in Guadeloupe has permitted a long-overdue engagement with the OECS countries with which these territories are geographically, and, in creole terms, culturally contiguous. Where secession has occurred decisively is in the cases of St Barthélemy and St Martin: yet, ironically, in seceding from Guadeloupe—in part because, like the smaller Anglophone islands, they were only ever governed from the larger territory due to colonial administrative convenience, but also because they have very different social panoramas and tourism enclave economies—they have reinforced their political integration into France (and, to differing extents, the EU). Put differently: they became more autonomous while simultaneously becoming less sovereign. So, in contrast to the Commonwealth Caribbean, although the timing of the secessionist attempt is important—these territories profited from the wider evolution of constitutional reform possibilities as encompassed in the LOOM of 2000—it is less important than the enduring non-sovereign relationship with a metropolitan power which provides a framework for a reconfigured and, like much else in Overseas France, paradoxical form of intensified non-independence that offers slightly greater autonomy than that which preceded it.

The Dutch Caribbean

The six Caribbean islands (Table 2) that continue to be part of the Kingdom of the Netherlands were colonised by the Dutch in the mid-seventeenth century. After being ruled as colonies for three centuries, in 1954 the six islands jointly gained autonomy as the newly established country of the Netherlands Antilles. Together with the (European) Netherlands and Suriname, this archipelagic unit became one of the three constituent countries of the Kingdom of the Netherlands. The 1954 Charter for the Kingdom, which continues to regulate its relations to the present day, stipulates that each of the countries enjoys domestic autonomy, while the Kingdom government is responsible for matters such as defence, foreign affairs, and nationality. The Kingdom, therefore, has a unique, quasi-federal structure that continues to create controversies and conflicts among its constituent units (Veenendaal, 2017).

Table 2. Dutch Caribbean Islands

Island	Population	Location	Political status
Curaçao	150.000	Southern	Kingdom Country (since 2010)
Aruba	120.000	Southern	Kingdom Country (since 1986)
St Maarten	45.000	Northeastern	Kingdom Country (since 2010)
Bonaire	20.000	Southern	Special Municipality (since 2010)
St Eustatius	3.000	Northeastern	Special Municipality (since 2010)
Saba	2.000	Northeastern	Special Municipality (since 2010)

While its structure has remained the same, since 1954 the Kingdom’s membership has changed dramatically. In 1975, Suriname became an independent state, reducing the number of Kingdom countries to two. In the subsequent period, the Dutch government strongly advocated for independence of the Netherlands Antilles, hoping to relinquish ties with its last former colonial territories. However, similar to many of the BOTs, rather than severing ties with the former colonial power, the islands primarily wanted to cut their links with each other. Over a period of twenty-five years, the Netherlands Antilles completely broke apart, and since 2010, when the entity was formerly dissolved, all of the islands have a direct (but different) constitutional connection with the Netherlands.

The Netherlands Antilles was arguably doomed to fail from its inception. The archipelago consisted of two-and-a-half islands in the north-eastern Caribbean (Saba, St Eustatius, and Dutch St Maarten) and three in the southern Caribbean (Aruba, Bonaire, and Curaçao) that are located almost 1000 kilometres apart. During colonial times, they were governed from the most populous island of Curaçao, and the colony was tellingly called ‘Curaçao and subordinates’ (Dutch: Curaçao en onderhoorigheden). Archival research reveals that the relationship between them was fraught with tensions from the very beginning, with the five other members (and in particular the second- and third-largest territories of Aruba and St Maarten) resenting their dependence on Curaçao (Oostindie & Roitman, 2014). In turn, reflecting the pattern of archipelagic units around the world, Curaçao loathed its duty to take care of the other, economically less viable, islands, which were commonly referred to as ‘bultu di buriku’, or ‘donkey’s load’ in local Papiamentu language.

While the structure and problems of the Netherlands Antilles strongly resemble those of the West Indies Federation, the difference is that none of the islands eventually became independent. In line with our broader theory, their enduring non-sovereign status has, seemingly paradoxically, facilitated secessionist movements on the islands. Aruba, which was the first island to secede from the Netherlands Antilles, is an exemplary case to illustrate this. As the second-largest island, it had the most troublesome relationship with Curaçao, and there were longstanding desires for separation and Aruban autonomy. However, in the 1970s the Netherlands made clear that secession would only be allowed if Aruba became a fully independent state. A 1977 independence referendum resulted in overwhelming support for independence (95%), and in 1986 Aruba obtained its much-desired ‘status aparte’ on the condition that it would become independent by 1996 at the latest. However, soon after obtaining this status, Aruban politicians made clear that they would not be forced into accepting independence. The island thus ultimately obtained what it wanted: separation from the Netherlands Antilles while retaining (direct) constitutional ties with the Netherlands.

The separation of Aruba preluded the complete fragmentation of the Netherlands Antilles. Having become increasingly prosperous due to its blossoming tourist economy, St Maarten organised its own status referendum in 2000, resulting in 70 per cent support for becoming a Kingdom country. This, in turn, sparked similar referenda on the other four islands, during which Curaçao also voted to become a Kingdom country, Bonaire and Saba voted for the rather vague option of ‘Direct ties with the Netherlands’, and St Eustatius was the only island to vote for remaining in the Netherlands Antilles. Complex negotiations between the islands and the Dutch government eventually resulted in the dissolution of the Netherlands Antilles on the symbolic date of 10 October 2010 (‘10-10-10’). Curaçao and St Maarten became autonomous Kingdom countries, while the other three islands were integrated into the European Netherlands as public entities or ‘special municipalities’ (Veenendaal, 2015). As a result, the Kingdom now consists of four constituent countries, while the integration of Bonaire, St Eustatius and Saba (the ‘BES’-islands) into the European Netherlands means that this has now become a trans-Atlantic country itself.

The fragmentation of the Netherlands Antilles represents the most significant constitutional change in the non-sovereign Caribbean in recent decades. Clearly, this wave of secessions was only possible because it did not presuppose full independence, and, in some cases, even reinforced dependence, albeit on the metropolitan power rather than the dominant member of the non-sovereign group. So, the island populations knew that they could break away from intermediary governing structures that were increasingly perceived as illegitimate while ultimately retaining—and even reinforcing—their non-sovereign status. The option of complete independence was included in all status referendums, but received only 15 per cent support on St Maarten, 5 per cent on Curaçao, and less than 1 per cent on the three smallest islands. The results of opinion surveys carried out in the 1990s (Oostindie & Verton, 1998) and 2010s (Veenendaal & Oostindie, 2018) indicate a similar pattern, and also demonstrate that inhabitants of all islands strongly value Dutch military protection, financial support, administrative supervision, and the Dutch passport and right of abode in the Netherlands (and the wider EU).

This does not mean that relationships within the Kingdom are always harmonious, however. Similar to most non-sovereign territories (Aldrich & Connell, 2020), island populations complain about continuing Dutch political interference into their domestic affairs, defective representation in metropolitan institutions, and a generally arrogant or ignorant attitude on the part of Dutch politicians and the broader public towards the islands (Mulder, 2018; Veenendaal & Oostindie, 2018). The traumatic legacy of colonialism and slavery reinforces negative feelings towards the (European) Netherlands, reflected in recurrent accusations of ‘recolonisation’ or neo-colonialism. As a result, rather than ‘the best of both worlds’ (Baldacchino, 2010; Rezvani, 2014), non-sovereignty is seen by many as a painful trade-off between the pragmatic benefits and the more ideational downsides of the link with a former colonial power. And as in virtually all other non-sovereign territories, this ‘head-versus-heart dilemma’ (Veenendaal & Oostindie, 2018) ultimately results in a victory for the pragmatic benefits of association.

Intriguingly, despite—perhaps even because of—the dissolution of the Antilles, the tension between separatist island nationalism and the ultimately pragmatic choice for non-sovereignty, has persisted on the BES islands. Since their integration into the European Netherlands in 2010, these islands have had to: introduce unpopular Dutch laws regarding same-sex marriage, abortion, and euthanasia; make a difficult and expensive currency transition to the US dollar; and face a significant influx of Dutch civil servants and citizens (Veenendaal, 2015). The sixteen Dutch ministries have jointly established a new shared service organisation, the Rijksdienst Caribisch Nederland (RCN), with headquarters on Bonaire (which is adjacent to Curaçao, just off the northern coast of Venezuela). This immediately created feelings of frustration and disillusionment on the two islands in the northern Leewards, Saba (population: 2000) and St Eustatius (population: 3000), which once more felt that they had become dependent on some larger, faraway island. These tiny islands, which are the smallest non-sovereign territories in the Caribbean, have since 2010 called for more equal treatment by the Dutch government and a breakup of the BES-unit, which, as they perceive it, reproduced many of the weaknesses of the defunct Antilles. For its part, the Dutch government has signalled its responsiveness to these calls, and has since adopted a more differentiated policy towards each of the three islands.

In sum, the experience of the Netherlands Antilles strongly supports our claim that secession is nowadays much more likely to happen in non-sovereign territories. In fact, since the constitutional link with the metropolitan state eradicates most of the challenges that secessionist movements in independent states are facing, there are virtually no limits to secession, with even the tiniest islands obtaining a different status than their neighbours. While a majority of the Dutch politicians and population would still like to relinquish all ties with the Caribbean islands, international legal provisions mean that they have mostly had to accept these developments. In addition, a potential transition of the islands towards independence now seems less likely than ever: most are not only too small to become independent, but the economic and political challenges of their independent neighbours (and, in particular, Suriname) serves as a constant powerful reminder of the pitfalls of making this transition.

Conclusion

What conclusions can be drawn about the overall conditions driving separatist tensions and trends from the Caribbean experience? A common feature of each of the cases we discuss was the tendency of colonial governments to group dissimilar—even far-flung—territories together for administrative convenience. A key reason was the attempt to overcome diseconomies of scale that resulted from duplicating governance structures on each island. But in most cases, attempts to generate administrative efficiencies exacerbated local political tensions. This was the case with Anguilla and St Kitts-Nevis, Tobago and Trinidad, St Barthélemy and St Martin in relation to Guadeloupe, and the creation of the Netherlands Antilles across more than 900 kilometres of Caribbean Sea. Related factors contributing to centrifugal pull were distance and the logistical challenges of air and sea communications and governance. In many cases, the peripheral populations expressed dissatisfaction with the governance process, alleging political and material neglect and inadequate economic support from the administrative centre. In almost all cases (Trinidad and Tobago, St Barthelemy, St Martin and Guadeloupe, the Northern and the Southern Caribbean Netherlands Antilles) this accentuated cultural differences and/or strongly differentiated local identities.

However, these similar conditions have produced varying outcomes, and this is where the capability explanation has the greatest empirical purchase. The reason is that diseconomies of scale that are a consequence of very small population size reduce the perceived viability of separatist agendas. For this reason, the success of secession among independent states has stalled since the break-up of the West Indies Federation, whereas among both French and Dutch territories multilevel national constellations have facilitated local aspirations for increased autonomy, while also offering the political and socio-economic benefits of regional and national citizenship of a member state of the EU (Ferdinand et al., 2020). For other Caribbean territories, secessionism has been more likely to result in increased island autonomy and institutional devolution, as in Nevis, Barbuda between 2006 and 2018, and Tobago. In these cases, decentralisation, autonomy and even separation tend to remain part of the ongoing political discourse, and substate nationalist movements and their agendas continue to evolve over time, influenced by changing leadership and external and local circumstances (Lluch, 2014). But none of them have resulted in full statehood.

The implications of these patterns are threefold. First, we add to studies that highlight how multilevel constellations, which are often designed to accommodate secession, can in fact serve to facilitate it. But, as Bauböck (2019) argues, we were only able to observe these lessons by paying attention to both sovereign and non-sovereign territories’ pursuit of status change within a wider constellation of polities. Second, our focus on institutional context supports Griffiths’s (2016) argument that existing administrative arrangements structure secession because they allow downsizing to occur in an orderly manner that is consistent with the desire of an international community for system-level stability. International actors have intervened to quell secessionist movements and guarantee regional stability, albeit this decision ultimately led to the successful secession of Anguilla from St Kitts and Nevis. Moreover, as the liberal international order has continued to fracture in recent years, and may potentially become less permissive towards small states in future, this insight carries important policy implications for would-be secessionists, their national governments, and the donor community alike (see Bishop et al., 2021). Finally, we have demonstrated the importance of studying small states and territories for mainstream political science (cf. Veenendaal & Corbett, 2015). Not only does their continued omission represent a clear and indefensible empirical gap, but because they are emblematic of the trend towards downsizing that is so integral to the ‘Age of Secession’, their experience allows us to reconsider existing studies in a new and penetrating light. By studying secession in small islands, we learn something about all states and territories, too.

Acknowledgements

The authors would like to thank three anonymous reviewers and John Connell for their helpful feedback on earlier iterations of the manuscript. Jack Corbett would also like to thank the British Academy (SRG1819\191216, supported by the Department for Business, Energy and Industrial Strategy) for supporting this research.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Even supposedly ‘global’ quantitative analysis ignores the very smallest states and territories. For example, Coggins’ (2014) criteria for secession only include movements that lay claim to more than 100 km² of territory. This excludes many of the cases discussed below, which, arguably, are some of the most interesting cases of potential or actual succession (in a region where large numbers of under-researched states experience centrifugal pressures).