One of the key developments of the first decade of the new century in Latin America was the emergence of leftist presidents who reversed the market reforms adopted at the end of the twentieth century. This phenomenon advanced furthest in Venezuela, Ecuador and Bolivia, where deep frustration with neoliberal economics fuelled the election and re-election of charismatic new leaders. The reversal of market reform occurred first in Venezuela, when voters repudiated economic liberalisation by electing Hugo Chávez in 1998 and then re-electing him by wide margins in 2000, 2006 and 2012. Bolivia followed suit with the 2005 election and 2009 re-election of Evo Morales, whose critique of neoliberalism resonated broadly among voters alienated by the failures of one of Latin America's most radical experiences with market reform. Most recently, Rafael Correa in Ecuador won election in 2006 and re-election in 2009 and 2013 by tapping into popular dissatisfaction with what he termed the country's ‘long dark night of neoliberalism’. In each case these presidents interpreted their repeated and convincing electoral victories as support for the construction of new forms of socialism in the twenty-first century, in contrast to the more moderate, market-friendly approaches of leftist governments elected elsewhere in Latin America, including Brazil and Chile.Footnote 1
But the effort by popular presidents to shift from twentieth-century neoliberalism to twenty-first-century socialism is just one dynamic that contemporary Venezuela, Ecuador and Bolivia have in common. Presidents Chávez, Correa and Morales share more than an antipathy to the measures of economic liberalisation that their predecessors adopted, and although their efforts to reverse such policies have received by far the most attention from scholars and journalists, each president also worked consistently to reverse many of the policies of decentralisation adopted prior to his election. Like most of the region, all three countries introduced important decentralising changes in the 1990s, the combined effect of which was to greatly enhance the stature of sub-national governments relative to the anaemic roles they had previously played. These changes included political decentralisation (introducing elections for sub-national officials), administrative decentralisation (devolving important governmental services) and fiscal decentralisation (introducing transfer systems to share revenue automatically with sub-national governments). In this article I show how these earlier decentralising changes came under attack from Chávez, Correa and Morales, all of whom had concluded at least one term in office by 2012. Whereas their opposition to liberalisation was incontrovertible from the very beginning of their governments, it was only over time that solid evidence emerged of their lesser-known, but potentially equally significant, opposition to decentralisation.
In attempting to shift the analytical focus from the reversal of liberalisation to the reversal of decentralisation, it is useful to acknowledge and interrogate an important asymmetry in the public avowals of these three presidents. While each embraced the critique of neoliberalism and the espousal of socialism as core to his mandate as president, none would admit to being opposed to decentralisation per se. On the contrary, all three presidents used language that appeared to value local communities and celebrate the devolution of power, which each argued had in fact occurred on his watch. So, what I seek to explain here is their hostility to the policies of decentralisation that were adopted before the election of each president, rather than their animus towards all forms of decentralisation or to the idea of decentralisation itself. Indeed, these presidents sought to reverse earlier decentralising changes precisely in order to advance what they promoted as alternative forms of local autonomy. The issue is not whether they tried to reverse earlier policies of decentralisation – the evidence that they did is strong – but rather what these efforts meant for each president and what they reveal about underlying regime dynamics. In other words, fuller accounts of these important new regimes require that we understand what happened not just to liberalisation but also to decentralisation.
Beyond demonstrating how these three presidents challenged decentralisation, this article seeks to answer the more difficult question of why they did so. There is little scholarship on recentralisation, so earlier literature on decentralisation is used for theoretical insights that might illuminate more recent and ongoing changes in the opposite direction. Beginning in the 1990s, scholars generated a great variety of hypotheses to explain politicians' support for decentralisation, a policy trend that swept not just Latin America but much of the developing world. Some focused on democratisation and liberalisation, asking whether democrats supported decentralisation in the belief that it would prevent future authoritarian reversals, and whether neoliberals supported decentralisation in the belief that it would shrink the central state.Footnote 2 Others looked to more structural causes such as urbanisation and the shift to a post-Fordist economy in which firms now demand institutional support from nimble sub-national governments.Footnote 3 Still others looked to political institutions and the partisan and electoral incentives that encouraged individual politicians – either ‘from above’ in the national government or ‘from below’ in sub-national governments – to push for decentralisation.Footnote 4 Appropriately altered, can these same hypotheses now explain the shift in power back to the centre?
This article draws on two of the most salient types of explanations generated by the earlier literature on decentralisation: ideational and institutional. Rarely were these explanations combined in that literature; whereas ideational arguments emphasised normative and principled commitments to decentralisation, institutionalists argued that politicians supported decentralisation independent of its merits and purely as a way to advance political careers and/or defend short-term electoral interests. In fact these explanations can be conjoined to offer a compelling account for why presidents have sought to recentralise. What is lost in terms of parsimony is offset by a fuller and more accurate explanation of the politics of recentralisation. At the ideational level, aspects of the drive to recentralise power in these three cases would remain unintelligible without examining the ideological dimensions of the struggle between officials in national and sub-national governments. Unlike in many other countries in the region, decentralisation in Venezuela, Ecuador and Bolivia was intimately tied to neoliberalism in the 1990s. This close association subsequently enabled Chávez, Correa and Morales to make two arguments in defence of recentralising measures. First, the support of pro-market presidents for decentralised governance made it possible to impugn the motives behind decentralisation and to maintain that each country had experienced the wrong type of decentralisation. Second, the conflation of decentralisation with liberalisation bolstered the claim that undoing decentralising changes was itself a way to undo neoliberalism. At the other end of the ideological spectrum, those who benefited from or who sought to defend market reforms saw recentralisation as a direct threat to those reforms.
While ideological contestation forms the backdrop for the struggle over recentralisation in these three cases, institutionalist approaches are most helpful for understanding the shape and urgency that the drive toward recentralisation has taken. Kathleen O'Neill's work is especially useful in its emphasis on politicians' calculations about the territorial distribution of their electoral support and the likelihood of winning either national or sub-national elections. Just as O'Neill has argued that national politicians decentralised when their electoral prospects were better at the sub-national level than at the national level, I find that national politicians confident of their continued success in national elections pursued recentralisation in the face of persistent and significant sub-national electoral defeats. Many of the most economically dynamic and important sub-national units in Venezuela, Ecuador and Bolivia (both municipalities and states/provinces/regions) were continuously governed by opponents of the president, despite each president's impressive successes in national electoral contests (including not just re-elections but also recalls, plebiscites and referenda). Having successfully overcome most checks on their power by controlling almost all salient national institutions, presidents Chávez, Correa and Morales saw recentralisation as a way to overcome remaining checks in the form of opposition-controlled sub-national governments.
The more recent phenomenon of recentralisation is not simply the obverse of decentralisation, however, and O'Neill's approach cannot be applied in a mechanical fashion. Instead, the drive toward recentralisation produced two important dynamics that were absent from the earlier pursuit of decentralisation and that are not captured in her approach. First is a high degree of asymmetry; whereas decentralisers in the 1990s endorsed mostly universal changes vis-à-vis sub-national units, recentralisers have sought where possible to impose specific losses on particular, opposition-governed sub-national governments. Second, whereas O'Neill's approach suggested that the supporters of decentralisation could reap its benefits only in the future when they had won sub-national races, the supporters of recentralisation can and do experience immediate gains. From the centre's perspective, recentralisation does not confront the same cross-temporal trade-offs that complicated decentralisation. Not only can recentralisation hobble the political opposition as soon as it is adopted, but it can also, in a more positive mode, simultaneously enable presidents to build and maintain new support coalitions at the centre. Revenues and responsibilities wrested away from opposition-controlled sub-national governments can be, and have been, redeployed to favour regime supporters, amounting to a ‘punish and reward’ strategy that was missing from the earlier dynamic of decentralisation.
From Decentralisation to Recentralisation
In using the literature on decentralisation to shed light on the phenomenon of recentralisation, one of the most prominent hypotheses to consider is that which explains the devolution of power to sub-national governments as a result of the ideational triumph of the market. According to this neoliberal thesis, pro-market reformers endorsed decentralisation for a variety of reasons, including that it would reduce the salience of the central government (along with that of powerful unions organised to influence that level of government), force sub-national governments to mimic the market and compete with each other for private investment, and enable the central government to achieve fiscal balance by offloading significant expenditures onto sub-national governments.Footnote 5
However, the neoliberal thesis was confronted with the empirical reality that market reforms had not actually driven many of the most important cases of decentralisation in Latin America. In Colombia, for example, the Conservative and Liberal parties decided to decentralise in the mid-1980s in response to public protests over inferior governmental services and in the hopes that stronger sub-national governments could help end the internal armed conflict.Footnote 6 In Peru, well before Alberto Fujimori took the country in a neoliberal direction, Alan García's leftist government decentralised in the late 1980s for reasons that had much to do with his party's hopes of winning regional elections.Footnote 7 In the case that had perhaps the greatest resonance in the region, Brazil decentralised in the 1980s not as a way to limit state intervention in the economy but rather in response to a broad demand for devolution as an antidote to authoritarian rule.Footnote 8 More generally, research showed that decentralisation can be used for illiberal ends, including to expand the prerogatives and resources of the sub-national state in ways that increase the aggregate size of the public sector, just as certainly as it can be used for liberal retrenchment purposes.Footnote 9
Although the neoliberal hypothesis was thus correctly dismissed by those looking for region-wide explanations of the turn toward decentralisation, market reform clearly operated as an important driver of decentralisation in the three countries that are the focus of this article. As described in greater detail in the case studies below, the governments that opted for decentralisation (the second Carlos Andrés Pérez administration in Venezuela, the Jamil Mahuad and Gustavo Noboa administrations in Ecuador, and the first Gonzalo Sánchez de Losada administration in Bolivia) were all closely identified with the attempt to reorient economic policy away from statism and toward the market. The ideological orientation was clear: decentralisation was promoted by neoliberal presidents whose visions of a less statist political economy involved devolution to sub-national governments. Beyond my three cases, I hypothesise that, where it was liberalisation that drove decentralisation, we should expect that subsequent struggles over the status of those market reforms will also lead political actors to revisit the decentralisation debate, and that these ideational conflicts will be important for understanding the pursuit of recentralisation.
Coming as it did on the heels of the institutionalist turn in political science, the study of decentralisation also produced a number of arguments about how institutional incentives could account for the shift to decentralised governance. Relative to ideational arguments, these institutional accounts received far more scholarly attention and resulted in more precisely specified theories. Three arguments were particularly influential, but of variable degrees of usefulness for the study of recentralisation in my three cases.
First, Eliza Willis, Christopher Garman and Stephan Haggard argued that internal party dynamics and lines of accountability within political parties were key to understanding decentralisation.Footnote 10 According to these scholars, national legislators who are beholden to sub-national officials within their parties (that is, governors) are more likely to support decentralisation than legislators whose careers depend on pleasing national party leaders, a hypothesis that explains why countries such as Argentina and Brazil were far more decentralised than Mexico and Venezuela in the 1990s. On the one hand, this argument is useful for my purposes because it underscores the tremendous latitude that Chávez, Correa and Morales all enjoyed as the undisputed leaders of new parties which they themselves had built and which declined to give important forms of authority to sub-national officials such as mayors and governors.Footnote 11 On the other hand, the collapse of established party systems in these three cases and their replacement with poorly institutionalised new parties mean that the analytical leverage provided by this party-centric account is limited, beyond the simple point that the presidents' own parties failed to impose checks on their drives toward recentralisation.
A second influential institutional account is that of Tulia Falleti, who emphasised the territorial interests of the institutional actors who initiate the decision to decentralise.Footnote 12 Focusing on the distinction between administrative, fiscal and political decentralisation, Falleti argued that the national executive prefers a sequence of changes that shifts from administrative to fiscal and then to political decentralisation, while sub-national officials prefer the opposite sequence. While the first path of national dominance limits the degree of autonomy that sub-national governments come to enjoy as a result of decentralisation, as occurred in Argentina, the sub-national dominance path produces far greater autonomy, as in Colombia. Applying Falleti's same logic to the case of recentralisation leads to the hypothesis that presidents would prefer first to recentralise political authority (before fiscal or administrative authority) because it was political decentralisation that they found most threatening and tried to defer until the end of the sequence. In contrast, in my three cases presidents have very much dominated the decision to recentralise, but they prioritised fiscal and administrative measures over political recentralisation. Presidents may well prefer to appoint sub-national officials, but cancelling sub-national elections once they have been introduced appears to be a non-starter in contemporary Latin America (except in the wake of outright authoritarian reversals, as in Peru in 1992).Footnote 13 Perhaps sensing that sub-national elections are here to stay, presidents have prioritised instead fiscal and administrative changes designed to reduce the governing authority of those sub-national governments where the opposition continues to win elections.
Kathleen O'Neill's work constitutes yet a third institutionalist approach, one that is especially valuable for the study of recentralisation. According to her argument, when national party leaders believed that sub-national offices would be important to their parties in the future, they sought to make those offices more attractive and powerful through the adoption of decentralisation.Footnote 14 She shows that governing party leaders in Colombia and Bolivia decentralised when they judged their chances of holding on to the national executive branch to be quite slim, while their counterparts in Venezuela and Ecuador, who had grounds to be more optimistic, preferred to keep authority centralised. Lengthy time horizons are a critical component of O'Neill's model because, even as fiscal decentralisation would produce hoped-for long-term gains at the sub-national level, it would impose immediate losses on the national politicians who adopted it and who would have access to fewer fiscal resources for the duration of their control over the national government. In this sense, fiscal recentralisation has proved to be even more appealing than fiscal decentralisation. Supporters of fiscal decentralisation in the national government faced a strong cross-temporal trade-off in that they had to be willing to undergo immediate losses in the expectation of future gains (after they could run for and win control over sub-national governments). But supporters of fiscal recentralisation in the national government, in contrast, could enjoy immediate gains of two types: imposing losses on opponents and redirecting recentralised revenues to their own supporters.
More generally, the same emphasis on the territorial distribution of electoral support that characterised O'Neill's approach to decentralisation also helps expose the core rationale behind the more recent pursuit of recentralisation. Consider first the national level. Chávez, Correa and Morales have been remarkably successful not just in winning election and re-election (and surviving recall votes in Venezuela in 2004 and in Bolivia in 2008), but also in increasing their representation in important national institutions, including legislatures, judiciaries, comptrollers, electoral courts and regulatory agencies.Footnote 15 Just as importantly, presidential supporters also won elections that gave them majorities in the constituent assemblies that were convened to produce new constitutions in Venezuela (1999), Ecuador (2008) and Bolivia (2009). In each case, new constitutions significantly expanded presidential power relative to other branches of government.Footnote 16
In all three countries, the rapid attenuation of checks by independent or opposition-controlled national institutions has thrown into sharp relief the existence (and persistence) of checks by sub-national governments. In Venezuela, for example, while Chávez's opponents enjoyed little representation in the national assembly, the elections for which they boycotted in 2005, the opposition won important states and municipalities in the 2000 and 2008 sub-national elections.Footnote 17 In Ecuador, when Correa suspended Congress in 2007 and thereby eliminated an important space for the representation of opposition interests, the popular mayor of Guayaquil, Jaime Nebot, emerged as the leader of the resistance to the president. In Bolivia, Morales' landslide electoral victory on 18 December 2005 sidelined Congress by ending the role it traditionally played in selecting the president, even as on that same day his opponents won control over a majority of departments in regional elections.
In a challenging national electoral environment, electoral victories in sub-national governments have provided a lifeline to the president's ideological opponents. Controlling sub-national governments is certainly no substitute for having adequate representation in national institutions, but the shelter provided by sub-national governments was critical for opposition groups – both those who hoped to use them as a toehold to launch national movements that challenge twenty-first-century socialism, and those who simply wanted to pursue alternative market-oriented projects in the territorial units that they held. As an example of the former, the victory of opposition politicians in critical sub-national races in 2008 in Venezuela accelerated efforts to create a unified national opposition movement known as the Mesa de Unidad Democrática. As examples of the latter, the ability of the opposition to hold Guayaquil in Ecuador and Santa Cruz in Bolivia ignited autonomy movements that sought to preserve each jurisdiction's more market-friendly framework.Footnote 18
Just as the hope of winning sub-national races animated the decision to decentralise in O'Neill's model, the fear that the opposition will continue to win important sub-national races has propelled the more recent pursuit of recentralisation. More specifically, opposition-controlled sub-national governments pose three types of threats to Chávez, Correa and Morales. First, while the established party system collapsed in each country with the rise to power of each president, sub-national governments have enabled the survival of these parties, without which it is likely that they would have disappeared from the scene. For example, the Partido Social Cristiano (Social Christian Party, PSC) in Ecuador retained control over the province of Guayas and the municipality of Guayaquil, the Movimiento Nacionalista Revolucionario (Revolutionary Nationalist Movement, MNR) has persisted as a presence in the Bolivian departments of Beni and Tarija, and Acción Democrática (Democratic Action) and Comité de Organización Política Electoral Independiente (Political Electoral Independent Organisation Committee, COPEI) candidates have won gubernatorial elections in Aragua, Miranda, Nueva Esparta and Táchira in Venezuela. Second, in addition to keeping traditional parties alive, sub-national governments have served as important conduits for the emergence of new challengers who have no or few ties to the earlier party system and who may therefore be even more threatening, including such figures as Miranda governor and 2012 and 2013 presidential candidate Henrique Capriles and Chacao mayor Leopoldo López in Venezuela. Third is the size and economic and political significance of the sub-national governments that the opposition has won and held. The opposition has won control of many of the most economically powerful jurisdictions in each country, including gas-producing Tarija in Bolivia, oil-producing Zulia and the industrial state of Carabobo in Venezuela, and agro-export-dominant Guayas in Ecuador.
To summarise the argument that I apply to the specifics of my three cases in the remainder of the article, Chávez, Correa and Morales have pursued recentralisation for both ideational and institutional reasons. As an ideological struggle, recentralisation appealed to the presidents in all three cases because it offered an important arena in which to challenge one of the most important legacies inherited from their neoliberal predecessors: decentralisation. In addition to ideology, I will show how recentralisation appealed to each president's institutional interests, both as a way of responding to the persistence of the opposition's electoral strongholds at the sub-national level, and as a way of redirecting resources toward supporters in new relationships unmediated by sub-national governments.
Venezuela
In Venezuela, decentralisation is connected to liberalisation not just because all significant decentralising reforms were adopted by neoliberal governments, but because these reforms were explicitly designed to help the regime weather the storm created by neoliberalism. In February 1989, Venezuela exploded in violence in response to neoliberal pricing measures introduced by President Carlos Andrés Pérez. One chief result of the Caracazo was that it increased support among the traditional parties for thorough-going decentralisation reforms designed to shore up the political establishment.Footnote 19 Figures within the establishment who had blocked meaningful decentralisation before the Caracazo subsequently relented and endorsed it as one possible way to close the great distance between state and society that the protests had exposed.Footnote 20 As a result, later in 1989 the government introduced elections for sub-national chief executives, including for mayoral offices that had not previously existed and for gubernatorial offices that had been appointed by the centre.Footnote 21
Market-oriented economics and decentralisation also characterised the administration of Ramón Velásquez, who became president after Pérez's impeachment on corruption charges in 1993. As head of the commission that had promoted state reform via decentralisation in the 1980s, Velásquez pushed the decentralisation agenda through a series of institutional, organisational and policy changes. This included the creation of a new Ministry of Decentralisation and support for a new organisation – the Association of Venezuelan Governors – that would bring together the governors to press common concerns. Meaningful fiscal decentralisation occurred in 1993 and 1996, when governors – now directly elected – used their leverage in the national legislature to demand the automatic sharing of revenues from the value added tax and from oil exports.Footnote 22 Velásquez also created the Fondo Intergubernamental para la Descentralización (Intergovernmental Fund for Decentralisation, FIDES), which extended funds to sub-national governments for projects that were not funded by the country's revenue-sharing system (situado constitucional); the latter was also raised from 15 to 20 percent of national government revenues under Velásquez.Footnote 23 Thus, thanks to a rapid series of changes introduced in Venezuela's relatively brief neoliberal interlude, governors were now elected in their own right, enjoyed a direct share in the country's oil wealth, could petition for the services they wanted to control, and participated in a new national organisation expressly designed to serve as their mouthpiece.
In addition to its neoliberal birth, decentralisation in Venezuela directly threatened the institutional interests of President Chávez. In 1998, when Chávez won a larger percentage of the vote (56 per cent) than any presidential candidate in four decades, his supporters won gubernatorial elections in only 30 per cent of the states (seven of 23). The disparity in the performance of Chavismo at the national and sub-national levels remained pronounced in the 2000 ‘mega-elections’ that followed the adoption of the new Constitution. With Chávez having successfully defeated a proposal during the constituent assembly by some Chavistas that all governors immediately be subjected to recall votes, gubernatorial elections were held along with presidential and legislative elections in 2000.Footnote 24 Chávez won 60 per cent of the presidential vote and secured a majority that would enable him to control the National Assembly, but the opposition held on to eight states: Amazonas, Apure, Aragua, Carabobo, Delta, Miranda, Sucre and Zulia.Footnote 25 While the sub-national threat to Chávez diminished in 2004, when Chavista candidates won all but two states, it increased once again in 2008 when the opposition recaptured the key states of Carabobo and Miranda, and when pro-Chávez candidates also lost in four of the five municipalities of Caracas and in the Caracas Metropolitan District, where outspoken anti-Chávez critic Antonio Ledezma defeated close Chávez ally Aristóbulo Istúriz.
In this context of important sub-national electoral victories for the opposition, recentralisation has taken a number of forms, beginning in the 1999 constituent assembly. Although decentralisation was inserted into Article 4 of the Constitution as a formal commitment of the state, this gesture belies a series of changes that actually diminished the power of sub-national governments relative to the 1961 Constitution. Most importantly, by shifting to a unicameral legislature, the new Constitution eliminated the Senate – established in 1811 – and along with it a chamber designed to represent the states in the national government. While party discipline diminished the Senate's federal identity for most of the fourth republic, it emerged from this period to play a critical role in the fiscal and administrative decentralisation that occurred in the 1990s. The new Constitution also downgraded state legislatures to state legislative councils and infringed on the autonomy of these councils by stipulating that their exact prerogatives would be determined by a subsequent national law.Footnote 26 On the fiscal side, whereas the 1961 Constitution had established 20 per cent of national income as a floor for fiscal transfers, the 1999 Constitution considers 20 per cent to be a ceiling.Footnote 27
Subsequent to the new Constitution, Chávez also instituted fiscal and administrative changes designed to constrain elected sub-national governments. First, with respect to fiscal changes, Chávez reduced the flow of revenues into FIDES, the most important fund for the financing of investment projects by sub-national governments. His government also consistently underestimated the price of oil in the budget as a means of depressing the size of transfers to the states (which had been increased in 1996).Footnote 28 As a result of these fiscal changes, while sub-national shares of public revenues increased from 17 per cent in 1989 to 28 per cent in 1998, they decreased to 21 per cent by 2004. Also by 2004, the national government had run up an accumulated debt of 12 billion bolivares (approximately US$ 6 billion) with the states.Footnote 29 Second, Chávez also pushed administrative recentralisation through a law in March 2009 that returned to the national government control over the nation's ports and airports, which had been decentralised 20 years before and which the 1999 Constitution had confirmed as exclusive state-level responsibilities.Footnote 30 Pro-Chávez legislators and governors defended recentralisation as an appropriate response to opposition governors who had used their powers to extend concessions to private sector operators.Footnote 31
Venezuela under Chávez also experienced a number of asymmetric measures designed to target and weaken opposition-controlled jurisdictions. For example, when Ledezma won as metropolitan mayor in 2008, Chávez responded a few months later by creating a new government of the capital district, appointing ally Jacqueline Faría as its head, and then transferring 90 per cent of the budget from Ledezma to Faría.Footnote 32 Furthermore, prior to surrendering control over sub-national jurisdictions that his candidates had lost in the 2008 regional elections, Chávez was also widely accused of asset-stripping by transferring vehicles, computers and even buildings from sub-national governments to various national ministries.Footnote 33 Although Chávez refrained from outright political recentralisation by letting the 2008 regional elections take place despite anticipated opposition gains, his administration influenced these elections by banning popular opposition candidates. Specifically, the National Electoral Court banned 272 candidates from running for sub-national offices, including Leopoldo López, the popular mayor of Chacao, who was prohibited from running for metropolitan mayor of Caracas.Footnote 34
If recentralisation in Venezuela offered a way to weaken sub-national governments – the mirror image of the dynamic that O'Neill highlighted in her understanding of decentralisation as a way to strengthen sub-national governments – it also appealed to Chávez as a means of building and maintaining new support coalitions. Put slightly differently, Chávez's attempt to build these new coalitions triggered various changes that partially reversed decentralisation in a number of ways. Fiscally, reducing and delaying transfers to sub-national governments had the effect of expanding the resources Chávez could then devote to expenditures that were closely identified with his government and over which he could claim credit, including both the Bolivarian Circles and the popular housing, literacy and health care missions that he established.Footnote 35 In other words, mayors and governors lost the fiscal means to provide critical governmental services at the precise moment that recipients could turn instead to centrally funded missions. The latter arrangement was not necessarily more prone to clientelism, but the patron in question was clearly Chávez.
While political recentralisation in the form of cancelled regional elections has not taken place in Venezuela, Chávez did claim additional powers to appoint regional authorities, which he could use to reward supporters and to keep them loyal. Specifically, in response to the defeat of the constitutional reform in December 2007 that would have, among other things, given Chávez the right to appoint regional authorities, in August 2008 Chávez used delegated decree authority to nevertheless establish the position of appointed ‘regional vice-presidents’.Footnote 36
Chávez also pursued recentralising administrative changes that expanded his ability to use the bureaucracy to reward supporters. He essentially replaced the logic of decentralisation that dominated in the 1990s with a logic of deconcentration that keeps decision-making authority vested in bureaucrats who respond to national ministries, chiefly the Ministry of Planning and Development.Footnote 37 Likewise, Chávez promoted regional axes of development via the regional development corporations that were established in the 1970s and that are controlled by the central bureaucracy rather than by governors. More sensationally, Chávez was able to claim credit for micro-level administrative outcomes via mobile cabinets (gabinetes mobiles), through which he was able to direct favourable governmental decisions toward preferred recipients (even for decisions that officially remained the prerogative of the sub-national officials who also, awkwardly, attended these cabinet meetings).Footnote 38 Just as clearly, administrative recentralisation enabled Chávez to cultivate new forms of union support, including from workers at Zulia's main port and airport who complained that, under decentralisation, the governor had violated labour rights and collective contracts.Footnote 39 Similarly, when a labour dispute occurred between Caracas Metropolitan Mayor Antonio Ledezma and municipal workers whom he was seeking to fire when he could no longer pay their salaries, Chávez's minister of popular power argued that the workers had been illegally fired and ordered their reinstatement.Footnote 40
But Chávez did more than simply reverse 1990s-era decentralisation measures. Over time he embarked on increasingly radical attempts to fundamentally redesign sub-national institutions, seeking to shift from established forms of liberal representative democracy to more direct forms that organised regime supporters into new governing bodies. While supporters and detractors disagree about the relative independence of these new institutions vis-à-vis the national government, it is hard to deny that they have detracted from the powers of existing sub-national governments in a dynamic that essentially reverses earlier decentralising changes. The communal councils created by Chávez in 2006 are a case in point. Chávez limited what sub-national governments could do not just by cutting or withholding transfers to them, but also by adopting new rules that forced all states and municipalities to spend their transfers on communal councils. Furthermore, whereas previously FIDES funds were split 60/40 per cent between states and municipalities, after 2006 their shares decreased to 42 per cent for states, 28 per cent for municipalities and 30 per cent in direct transfers to communal councils.Footnote 41 Chávez's preference for sub-national institutions that include regime supporters but exclude elected sub-national officials can also be seen in his decision to privilege communal councils over the Local Public Planning Councils (Consejos Locales para la Planificación Pública, CLPPs) that were created in the 1999 Constitution and that included significant forms of participation by local elected officials.Footnote 42
Ecuador
Turning from Venezuela to Ecuador, recentralisation appealed to Correa largely for the same reasons that it appealed to Chávez. For each president, ideological commitments and institutional interests both pointed in the direction of recentralisation. Although recentralisation was characterised by the asymmetric treatment of sub-national units in all three countries (and the harsher handling of opposition-governed units), this asymmetry was particularly intense in Ecuador, where the recentralisation process has often resembled a personal struggle between Rafael Correa and one particularly powerful opposition mayor: Jaime Nebot of Guayaquil. Institutionally speaking, whereas presidents in Venezuela and Bolivia were threatened by opposition electoral victories in multiple sub-national units, in Ecuador it was the persistent dominance of Guayaquil by the PSC that mattered most.
Although decentralisation at the time of the left turn had not advanced as far in Ecuador as in Venezuela, it was just as tightly connected to neoliberalism. The initial impetus for decentralisation in the 1990s came from the PSC, a pro-market party based in coastal Guayaquil that receives strong support from the country's most important agro-exporting interests.Footnote 43 The PSC failed to win the presidency in the 1990s, but it served as an important coalition partner in Ecuador's highly fragmented party system and used its leverage on several occasions to secure the passage of key decentralisation measures.Footnote 44 Most importantly, in 1997 the PSC pushed through the so-called Law of 15 Per Cent, which stipulated the automatic revenue-sharing of no less than 15 per cent of the government's current revenue. Administrative decentralisation followed later that year with the passage of a law that established the procedures according to which individual sub-national governments could petition for the transfer of administrative responsibilities.Footnote 45
The association of decentralisation with liberalisation deepened in the late 1990s in Ecuador. In 1998, under the administration of Jamil Mahuad, pro-market representatives in the constituent assembly sought to reinforce fiscal and administrative decentralisation by successfully lobbying for the insertion of both measures into the country's new Constitution.Footnote 46 Simultaneously, in a country where Congress had blocked most significant neoliberal reforms in the 1980s and 1990s, the constituent assembly also approved pro-market changes, including the controversial lifting of restrictions on concessions to privately owned utility companies.Footnote 47 Although widespread opposition to Mahuad's proposal to dollarise the economy led to his overthrow in 2000, the neoliberal orientation of the government was sustained through the rise to power of vice-president (and Guayaquileño) Gustavo Noboa. In 2002, Noboa adopted one of Ecuador's most important decentralising reforms to date, which enabled taxpayers to devote up to 25 per cent of their income tax payment to the municipality of their choice. Although all municipalities were covered by this law, the measure was clearly a fiscal victory for Guayaquil, the country's wealthiest city, and a political victory for Ecuador's paramount neoliberal, Leon Febrés Cordero, who had aggressively advocated fiscal decentralisation as mayor of Guayaquil in the 1990s.
Against this backdrop, when Rafael Correa rose to the presidency in 2006, not only did he have ideological cause to oppose decentralisation as a neoliberal imposition, but he also had institutional reasons to fear that it was disproportionately benefiting Febrés Cordero's successor as mayor of Guayaquil and his main political opponent, Jaime Nebot. President Correa found a particularly formidable opponent in Nebot, a PSC politician who ran unsuccessfully for the presidency in 1992 and 1996 but who handily won election on three different occasions as mayor of Guayaquil (2000, 2004 and 2009). After his election in 2006, Correa repeatedly dismissed Nebot as a representative of Ecuador's traditional and corrupt party system, and the mayor in turn offered strong rhetorical and logistical support for Guayaquil's autonomy movement, including sponsorship of the pro-autonomy Junta Cívica de Guayaquil (Guayaquil Civic Board). In March 2008, Nebot convened a large anti-Correa rally in which he and other city officials critiqued what they viewed as the president's drive toward socialism. This rally culminated in demands that the constituent assembly, which was convening at the time, recognise a pro-autonomy referendum that Guayaquil had endorsed in 2000. When the new Constitution made it harder rather than easier for provinces to achieve autonomy,Footnote 48 a majority of Guayaquil voters rejected the Constitution in 2008 even as it won an easy victory nationwide.Footnote 49
Correa responded to the opposition in Guayaquil with a series of (often idiosyncratic) recentralising measures that cut across different issues, including policing, transit, charity foundations and civil registries. Correa's recentralising drive stands out for its attack on the most symbolically significant and distinctive public institutions and civil society organisations in the city of Guayaquil and in the province of Guayas. A case in point is the sui generis Comisión de Tránsito del Guayas (Guayas Transit Committee, CTG), an institution established in 1948 to set transit policy and finance transit investments through a surcharge that residents agreed to pay on their income taxes (and that exists nowhere else in Ecuador). In 2007, Correa sought to take control of the committee's board of directors by successfully filing an injunction with the politically compliant constitutional tribunal against committee members who also belonged to the anti-Correa Chamber of Production and Guayaquil Civic Board.Footnote 50 In 2008, Correa secured the elimination of the CTG altogether. Likewise, by taking away its lottery, Correa reduced the financial basis of Guayaquil's independent Junta de Beneficiencia (Charity Board), an important cultural space that brought together the city's socially prominent families.Footnote 51 Correa has also undermined uniquely Guayaquileña organisations that are of more recent vintage, including the city's Corporación para Seguridad Ciudadana (Citizen Security Corporation, CSC). Created in 2005 to augment and complement the work of the national police force within the city, the CSC received 70 per cent of its funds from the national government until Correa ended the support in 2008, arguing that policing is a uniquely national prerogative.Footnote 52 Each of these measures has had the effect of undercutting the authority of a mayor who has operated as the most articulate and persistent critic of the president on the national stage.
In addition to reaching within Guayaquil and rolling back local practices that have made it distinct, Correa has sought to handicap the opposition through other changes that would reduce this district's oppositional weight vis-à-vis the centre. Most importantly, the president delivered on a pledge he made on the campaign trail in 2006 to support the provincialisation of the Santa Elena peninsula within Guayas. Upon his election, Correa pushed through a law in 2007 that gave Santa Elena its own provincial status, effectively carving up Guayas. Communities in Santa Elena's less developed municipalities had complained that Guayas' provincial government favoured richer municipalities, including Guayaquil. Elevating Santa Elena gave the president an opportunity to pursue his broader campaign against the private sector business interests that dominated Guayaquil and led the opposition to his government.Footnote 53 Even as he supported the division of Guayas, Correa also proposed (unsuccessfully) a new territorial scheme to amalgamate existing provinces into a smaller number of regions in the form of horizontal bands, each of which would straddle the country's three natural zones: coast, highlands and Amazon.Footnote 54 Opponents in Guayaquil accused the president of thereby trying to further divide the provinces of the coast from one another and to dilute the oppositional power of Guayas by subsuming it within a more aggregated unit.
Beyond these explicitly asymmetrical measures, Correa also pursued a number of other formally universal changes that nevertheless have generated an especially negative impact on his chief opponents in Guayaquil. For example, as part of his anti-privatisation drive, Correa insisted that Ecuador's new Constitution prohibit utility concessions. This prohibition affected Guayaquil alone as it had been the only sub-national government that responded to the 1993 Ley de la Modernización del Estado (Law of the Modernisation of the State) by offering concessions to private enterprises.Footnote 55 The new Constitution put into question the 2001 decision by municipal regulatory authorities in Guayaquil to sign a contract that transferred responsibility for water treatment, transport and delivery to Interagua, a joint venture headed by Bechtel. In a similar fashion, since Guayaquil generated a more energetic response to Ecuador's optional approach to decentralisation, according to which sub-national governments could petition for responsibilities if they so chose, recentralisation has affected it disproportionately. Guayaquil was, for example, the only jurisdiction negatively affected by Correa's decision to recentralise the civil registry, and it was one of only two cities affected by the recentralisation of airports. Correa rejected Nebot's petition for the transfer of the port to the city in 2007, after which, in 2008, ports were removed from the list of services that could even be the subject of a transfer petition.Footnote 56
Beyond laying siege to Guayaquil as the most important electoral redoubt of the opposition, recentralisation also enabled Correa to reward various sets of supporters, ranging from ordinary voters to recipients of high-level patronage positions. In the fiscal sphere, two key episodes of recentralisation produced significant new revenues for the national government. First, in April 2008 the president's majority in Congress endorsed a seemingly technical change that had the effect of significantly cutting the size of revenue transfers to sub-national governments which had been mandated in the Law of 15 Per Cent in 1997. Specifically, Correa demanded that additional oil revenues produced by the 2003 completion of a new heavy crude pipeline be considered ‘capital income’ (ingresos de capital) rather than ‘current income’ (ingresos corrientes), and thus not subject to revenue-sharing. Second, in 2008 Correa also abolished the right of Ecuadorians to direct 25 per cent of their income tax payments to their municipalities.Footnote 57 This change appealed to the president not just because it served to rebuke Guayaquil, the city that benefited disproportionately from the measure, but because it meant that all income tax revenues would now accrue to the national government at a time when Congress no longer operated as an effective horizontal check on presidential spending. These measures of fiscal recentralisation eased constraints that otherwise would have made it difficult for Correa to pursue popular changes in Ecuador's conditional cash transfer programme, the bono de desarrollo humano (human development bond, BDH). In his first term, Correa more than doubled BDH payments for the elderly and the disabled and broadened coverage to include a wider set of recipients.Footnote 58
Given its weight as an electoral district and considering that it re-elected as mayor the president's leading critic in 2009, Guayaquil was especially critical to Correa's efforts to win the allegiance of voters who otherwise supported local opposition figures. His strategy was to expand the presence and prerogatives of the national government in Guayaquil and to thereby dispute Nebot's claims of being uniquely responsible for positive local outcomes. One important manifestation of this strategy was Correa's use of his executive powers over the bureaucracy to create a new Ministry of the Coast (Ministerio del Litoral) in 2007. Although this change cannot strictly be characterised as recentralisation because it did not take powers away from sub-national governments, it does represent a concerted attempt to expand national government programmes to attenuate the significance of locally provided services. By consolidating the offices of the subsecretariats of all major ministries (health, culture, education) within one ‘Super Ministry’ in Guayaquil, the new ministry was designed to promote the coastal penetration of the national government and to increase its agility in responding to local demands, particularly from marginalised communities.
Located in the landmark ex-Banco del Progreso building opposite Guayaquil's city hall, and including an office on the third floor for Correa himself, the new ministry signalled the president's willingness to take his battle to Nebot's own turf. Referring to the collapse of the Banco de Progreso during Ecuador's 1999 financial crisis, Correa inaugurated the new building by declaring that ‘this building, which was formerly the sign of such opulence and exploitation, now becomes a symbol of the services that we will be providing the citizens of the coast’.Footnote 59 The new ministry had the added benefit of helping Correa find a cabinet-level position for Ricardo Patiño – a close adviser who was forced out as economic minister in July 2007 over charges that he had manipulated bond markets – as well as jobs for the 800 civil servants assigned to work there.Footnote 60
Bolivia
In most core respects, recentralisation in Bolivia parallels the Venezuelan and Ecuadorian cases, and the pursuit of recentralising measures by President Evo Morales provides additional support for the significance of both ideational and institutional motivations. At the same time, examining these motivations in the Bolivian case requires greater differentiation between municipalities and departments as different levels of sub-national government. More specifically, in contrast to decentralisation to municipalities, which played an important role in the strengthening of Morales' Movimiento al Socialismo (Movement Toward Socialism, MAS) in the 1990s, it was decentralisation to the departmental level along with the victory of opposition candidates in the four eastern departments of the Media Luna (Beni, Pando, Santa Cruz and Tarija) that triggered sharply recentralising proposals from the president in Bolivia.
As in Venezuela and Ecuador, decentralisation also had neoliberal parentage in Bolivia, but in two different stages. First, Bolivia's adoption of municipal decentralisation in 1994 was heavily promoted by President Gonzalo Sánchez de Losada, chief engineer of the ambitious market reforms that were implemented a decade earlier. Critics on the Left were suspicious of the 1994 Ley de Participación Popular (Law of Popular Participation), which was designed behind closed doors and without serious public debate, and which met with strenuous objections from unions.Footnote 61 These critics saw decentralisation as an attempt by the neoliberal state to increase its presence in and control over indigenous communities where the state had previously been absent.Footnote 62 Decentralisation and the privatisation of state assets in critical sectors of the economy (telecommunications, utilities and petroleum) were the two most important legacies of Sánchez de Losada's first government (1993–7). While decentralisation was thus intimately associated with liberalisation in Bolivia, it also had unintended consequences as a measure that actually helped end neoliberalism by facilitating the emergence of Evo Morales' MAS party. Although Sánchez de Losada hoped that his own MNR party would benefit disproportionately from municipalisation,Footnote 63 in line with O'Neill's argument, in fact it was the MAS that used stronger municipal governments to greater effect.Footnote 64
In contrast to the first stage of decentralisation in Bolivia, which empowered municipal governments, the second stage focused on departmental governments and arrived not at the zenith of the neoliberal period but at its very end. Advocates of more power for departmental governments were sorely displeased when Sánchez de Losada opted to direct decentralisation to the municipal rather than to the departmental level in 1994.Footnote 65 The sense of betrayal was particularly acute among the pro-market actors who dominated the Pro-Santa Cruz Committee, which shared Sánchez de Losada's liberal orientation and which saw in departmental decentralisation the best chance to check growing demands on the national government to turn away from the market. In a context of increasing polarisation, pro-market forces in the east secured two major decentralising concessions in 2005. First, they forced President Carlos Mesa to agree to hold elections for departmental prefects, positions that had previously been appointed. Second, when Mesa fell from power in June 2005, they secured from the interim government of Eduardo Rodríguez a set of decrees that devolved to departmental governments over half of the revenues produced by a new direct tax on hydrocarbons. Although a greater percentage (34 per cent) was assigned to municipalities than to departments (19 per cent), the decrees were seen as a major victory for the departments given the far fewer expenditure responsibilities that were assigned to departments relative to municipalities.Footnote 66 In the east and among the MAS, departmental decentralisation in 2005 registered as a major victory for the advocates of neoliberalism.
In addition to representing a neoliberal victory on the eve of his election to the presidency in December 2005, decentralisation to the departments posed a sharp institutional threat to Morales because those elections confirmed Bolivia's political polarisation along a single east–west territorial cleavage. In the 2005 presidential race, when Morales won 53 per cent of the vote, he failed to carry any of the four eastern departments, which were won by Podemos candidate Jorge Quiroga. In his re-election four years later, Morales carried all departments save the three lowland departments of Beni, Pando and Santa Cruz, where Manfred Reyes won. Elections for departmental authorities have likewise demonstrated strong opposition to Morales in the east. In elections for prefect in 2005, the MAS won three western departments but lost in all of the eastern departments. In contrast to all of his predecessors, who could appoint and dismiss prefects as they wished, Morales in his first administration thus confronted a majority of prefects who were members of the opposition. Although the MAS picked up an eastern department, Pando, in the 2010 departmental elections, Morales' opposition held on to the remaining three lowland departments. The distribution of this electoral support helps explain the president's strong interest in recentralising changes, along the lines suggested by O'Neill.
Morales' centralising preferences can be seen in the tense stand-off that took place over the question of departmental autonomy in the years between 2005 and 2008. The president responded to the eastern opposition by campaigning against the referendum on departmental autonomy that took place on the same day (2 July 2006) that Bolivians elected representatives to the new constituent assembly. The referendum, which would require the assembly to insert departmental autonomy into the Constitution, passed in all four eastern departments but in none of the five western departments, creating significant uncertainty about how the assembly would decide the question of autonomy. During the many months of combative interactions between eastern and western representatives that dominated the assembly in 2006 and 2007, Morales consistently favoured centralist designs and rejected measures that would empower departmental governments.Footnote 67 The clash between Morales' centralising vision and the eastern prefects' decentralising vision ultimately contributed to the stalemate of the assembly, the abandonment of the assembly by the opposition and the approval of the draft of the Constitution favoured by the MAS in Oruro in December 2007.Footnote 68 While the polarised political environment surrounding the constitutional revision complicates analysis, Morales clearly preferred a Constitution that would diminish the governing authority of his departmental opponents.Footnote 69 In the time that has elapsed since the adoption of the new Constitution, Morales' government has adopted implementing legislation that keeps governors on a tight fiscal and administrative leash (the 2010 Ley Marco de Autonomía y Descentralización, Framework Law on Autonomy and Decentralisation) and initiated judicial proceedings against all four eastern governors (Ernesto Suárez of Beni, Leopoldo Fernández of Pando, Rubén Costas of Santa Cruz and Mario Cossio of Tarija).
Another concrete manifestation of Morales' institutional interests in proposing recentralisation can be seen outside the context of the constituent assembly in the struggle over the distribution of revenues generated by the 2005 tax on hydrocarbons (impuesto directo a los hidrocarburos, IDH). As argued above, the 2005 decrees that directed over half of the revenue from this tax to sub-national governments were a major victory for officials in these governments, particularly those who opposed Morales on ideological grounds and who feared that he could deny revenues to politically non-compliant departments. The effects of the 2005 law under President Mesa which increased oil taxes from 16 per cent to 32 per cent, together with the increase in international gas prices, generated an increase of US$ 1.3 billion between 2004 and 2007, or 10 per cent of GDP. According to Weisbrot and Sandoval, the four Media Luna departments as a group receive the lion's share of the hydrocarbon revenue, 30 per cent as compared to 19.7 per cent for the other five departments – whose combined population is 79 per cent greater than the Media Luna.Footnote 70 As awareness dawned of the boon that the 2005 decrees represented for his opposition, Morales proposed changes in October 2007 that would increase the share of hydrocarbons for the national government and reduce the share for departments, municipalities and universities. In response to widespread protests, Morales then agreed to restore the shares for municipalities and universities but not for departmental governments, which are the bedrock institution of the opposition and which experienced a 70 per cent cut in departmental resources.Footnote 71 Despite a boycott by legislators representing the Media Luna, the legislature passed this recentralisation of revenue in November 2007 – the most significant setback for the eastern opposition in Morales' first term as president.
While recentralisation facilitated a dual strategy in all three countries – enabling presidents to punish their sub-national opponents while simultaneously redirecting recentralised resources to their supporters – the two strategies were especially well integrated in the Bolivian case. Relative to Venezuela and Ecuador, the connection between recentralisation and centrally provided income support in Bolivia is even tighter; Morales explicitly defended his proposed cuts in revenue transfers to sub-national governments (from the IDH tax described above) as necessary to fund the so-called Renta Dignidad (monthly ‘dignity’ payments for older Bolivians). As Morales explained to the nation in October 2007, the nationalisation of the gas industry the previous year cut deeply into the dividends formerly paid by private gas companies, which had been used since 1996 to fund Bolivia's BonoSol (Solidarity Bonus). In addition to renaming this transfer programme to more tightly connect it with his administration, Morales proposed expanding eligibility to Bolivians aged 60 and above and increasing the benefit from approximately US$ 225 to US$ 300 per year. Acknowledging the overwhelming popularity of the Renta Dignidad proposal, opposition governors argued that they were not opposed to the new programme, but felt that the president should finance it out of the national coffers rather than sub-national transfers.Footnote 72 Despite this opposition, MAS lawmakers endorsed the new legislation in November 2007.
As is the case with Chávez in Venezuela, Morales pursued deeper institutional reforms not just to weaken existing opposition-controlled sub-national governments, but also to favour new sub-national entities more likely to support the president. The critical arena for Morales was the constituent assembly in 2006 and 2007, which devoted significant attention to the question of whether and how to extend autonomy to four distinct sub-national entities: municipalities, departments, regions and a new unit, the Territorio Indigena Originaria Campesina (Native Peasant Indigenous Territory, TIOC). ‘Region’ here refers to a jurisdiction within an existing department that could petition for autonomy from that department. TIOCs refer to an indigenous district that can come into being and enjoy autonomy at any scale of government. In the lengthy struggle over the new Constitution, Morales consistently sought to privilege regions and TIOCs – neither of which yet exist as real governments – and to downplay existing municipal and departmental governments.Footnote 73 Morales' constitutional preferences reflected the hope that pro-government TIOCs and regions could be encouraged to form within the opposition-controlled departments of the eastern lowlands, in effect constraining the opposition governors ‘from below’.
Conclusion
This article has argued that the turn to the left in Venezuela, Ecuador and Bolivia also triggered a less observed turn toward recentralisation. Marked by the reversal of many of the decentralising policies that were adopted in the 1990s, this pursuit of recentralisation weakened the governing authority that elected sub-national governments are able to wield in a variety of dimensions. Most importantly, in the fiscal arena presidents Chávez, Correa and Morales used a number of tools to reduce the revenues under the control of sub-national officials, whether by giving other actors a share of the common pool that used to be shared with sub-national governments alone (Venezuela), directly cutting the percentage of specific taxes that used to be transferred to sub-national governments (Bolivia), or determining that formerly shared revenues are no longer subject to sharing with sub-national governments (Ecuador). In addition to limiting sub-national fiscal capacity, presidents also cut into the range of administrative roles that sub-national governments can play, as seen in the return to the centre of control over ports, airports, roads and highways. In none of the three countries were sub-national elections cancelled, although each president made moves that encroached on political decentralisation, either by using electoral and judicial institutions to disqualify sub-national candidates (Venezuela) or remove sitting sub-national officials from office (Bolivia), or by redrawing sub-national borders to limit the political influence of the most powerful sub-national executives (Ecuador).
Why did presidents in Venezuela, Ecuador and Bolivia opt for recentralisation so soon after the adoption of decentralisation? To explain the pursuit of recentralisation in these three countries, I have argued for the relevance of ideational factors in general and ideological struggles over market reform in particular. Although ideational accounts were dismissed as a causal factor in the earlier search for general theories of decentralisation, ideology should be taken seriously in the study of recentralisation in distinct national contexts. In my three countries, it was the plainly neoliberal sponsorship of decentralisation that set the stage for recentralisation once the neoliberal era ended in the course of the left turn. Not only did recentralisation emerge as a mechanism to challenge the neoliberals who had decentralised, but it also offered a way to put a stop to neoliberal behaviours at the sub-national level, as in the cases of a pro-market mayor in Ecuador who used his decentralised powers to offer concessions to the private sector, or pro-market governors in Venezuela who nullified labour contracts. If ideological motivations propelled the pursuit of recentralisation in these three countries, this leads to the more general hypothesis that, where decentralisation was closely associated with neoliberalism, subsequent movements against market reforms will generate pressures to recentralise. Chile might be a case in point, where contemporary critics of market-driven educational reforms under Augusto Pinochet have pushed to recentralise many of the responsibilities over schools and universities that were decentralised under the dictatorship.
Considering the counterfactual here, however, suggests that it would be a mistake to overstate the importance of ideological motives. Even if decentralisation had not been so closely associated with neoliberalism in these three cases, presidents Chávez, Correa and Morales would likely still have sought to reverse it so long as their opponents continued to win such important sub-national electoral victories. More generally, while recentralisation emerged as a way to challenge neoliberalism in Venezuela, Bolivia and Ecuador, and may well play a similar role elsewhere, it does not always conform to this logic. Indeed, in other countries in the region, where decentralisation was not so closely aligned with the neoliberal project, the push for recentralisation has come from neoliberals themselves. In Peru, for example, the most important recentralising episode took place under Alberto Fujimori, whose autogolpe on 5 April 1992 led not just to the closure of Congress but also to the dismissal of elected regional officials (who were replaced by appointed officials for the duration of his government). In Argentina and Brazil, market reformers such as Carlos Menem and Fernando Henrique Cardoso sought fiscal recentralisation as a way to discipline profligate sub-national officials and to restore macro-economic balance at the national level. What seems clear is that, while there is certainly no uniform relationship between neoliberalism and recentralisation in Latin America, ideology likely matters if we want to understand the full import of recentralisation, which acquires a different ideological valence depending on national context, sponsorship and timing.
In addition to substantiating the ideological contours of contemporary struggles between national and sub-national officials in my three cases, this article demonstrates that some of the same institutionalist hypotheses developed to explain decentralisation can also help explain recentralisation. In particular, applying the logic of O'Neill's hypothesis to recentralisation rather than decentralisation brings into focus the tremendous appeal of the former for its sponsors in the national government, who face none of the cross-temporal trade-offs that confronted their decentralising predecessors. Whereas decentralising politicians had to wait for the future to reap the gains of their decision to decentralise, recentralising politicians immediately reap the benefits of recentralisation while they still control the national government. In one of the clearest and most consistent patterns to emerge from the comparison of Venezuela, Ecuador and Bolivia, recentralisation is marked by a dual logic in that presidents can use it to harm political opponents while simultaneously benefiting political supporters. In a punitive vein, amidst the attenuation of checks at the national level, recentralisation appeals to presidents anxious to ensure that checks from opposition-controlled governments do not meaningfully limit their remarkably broad scope for action. In another difference with decentralisation, the pursuit of recentralisation has been highly asymmetric, whether in the form of asset-stripping following sub-national electoral defeats in Venezuela, cuts in revenues from hydrocarbon taxes that disproportionately affected eastern departments in Bolivia, or the highly idiosyncratic changes in Ecuador that have targeted Guayaquil alone.
Recentralisation is not just about punishment, however. In each case it has enabled presidents to insert themselves in a positive fashion into relationships with voters and supporters that no longer need to be mediated by independently elected sub-national officials. Whether directly (Bolivia and Venezuela) or indirectly (Ecuador), for example, fiscal recentralisation has helped presidents establish and finance popular new programmes (missions in Venezuela and cash transfer programmes in Bolivia and Ecuador) that create new connections with supporters and that are not channelled through sub-national officials – in contrast to the Bolsa Família in Brazil, for example, in which mayors are assigned important roles.Footnote 74 Administrative recentralisation has brought with it attractive patronage opportunities, both vis-à-vis high-level bureaucratic appointees and the lower-level employees of governmental offices that have now been returned to the centre. Furthermore, recentralisation as a mechanism to reward supporters has not just involved reversing 1990s-era fiscal and administrative decentralisation, but has also led to even more fundamental proposals that shift power to new types of sub-national entities (communes in Venezuela, TIOCs and regions in Bolivia) and away from existing sub-national governments (many of them dominated by the opposition or with the potential to be won by the opposition).
While the pursuit of recentralisation did not meet with equal results in all three countries – for reasons that are beyond the scope of this article and are examined elsewhereFootnote 75 – cross-national commonalities in presidential attempts to recentralise are striking. Indeed, the similar logic at play in the pursuit of recentralisation in each country suggests that it likely deserves to be considered an important characteristic of the drive toward ‘twenty-first-century socialism’ in Latin America, and not a random or incidental feature. In fact, while scholars continue to debate just how ‘illiberal’ these three presidents really are,Footnote 76 with some noting that essentially market-based systems have survived ‘twenty-first-century socialism’ intact, what seems clear in contrast is that sub-national governments have faced deeper, almost existential threats that promise to fundamentally revolutionise the architecture of government. Although further research is necessary, the reversal of decentralisation may in fact be a more regular feature of these regimes than the reversal of liberalisation itself.
