Political settlements, elite pacts, and governments of national unity
A growing body of recent scholarship puts political settlements at the centre of the development
process. The political settlements approach focuses on the formal and informal negotiations, bargains,
pacts and agreements between elite actors, as crucial drivers of the locally effective institutions and
policies that promote or frustrate the achievement of sustainable growth, political stability and socially
This paper surveys and clarifies the conceptual field by addressing key questions.
Published on: Mar 4, 2016
Transcripts - Political settlements, elite pacts, and governments of national unity
DLP Policy and Practice for Developmental Leaders, Elites and CoalitionsDEVELOPMENTAL LEADERSHIP PROGRAM Background Paper 10 Political Settlements, Elite Pacts, and Governments of National Unity A Conceptual Study Edward Laws August 2012 “The terminology of ‘political settlements’ has become much used, but often lacks definitional clarity” (Whaites, DfID 2008: 6-7). “Advocating systematic governance reforms without analysing existing political settlements and state-society relations, and how reform might affect patterns of inclusion and exclusion may result in harmful advice.” (AusAID, 2011: 25). www.dlprog.org
DLP Policy and Practice for Developmental Leaders, Elites and CoalitionsDEVELOPMENTAL LEADERSHIP PROGRAMThe Developmental Leadership Program (DLP)is an international policy initiative informed bytargeted research and directed by an independentsteering committee. DLP is supported by a globalnetwork of partners and currently receives its corefunding from the Australian aid program.The views expressed in this publication are thoseof the author(s) and do not necessarily representthose of the Australian Government or any ofDLP’s other partner organisations.
‘One of the most infamous political settlements is the peace deal reached in Sierra Leone in 1999 …’ (Fritz and Menocal, ODI 2007: 59). ‘Australia can support processes and spaces where on-going adjustment of the political settlement is decided inclusively and peacefully’ (AusAID, 2011: 34). ‘Deeper political settlements require shared understanding and acceptance of the rules of the game’ (AusAID, 2011: 35). ‘The structures of the state are determined by an underlying political settlement; theforging of a common understanding, usually among elites, that their interests or beliefs are served by a particular way of organising political power’ (Whaites, DFID 2008: 4). ‘Political settlements define how political and economic power is organized’ (DfID, 2010a: 7). ‘An inclusive peace process aims to achieve a peace agreement that lays a strong foundation for a new political settlement’ (DfID, 2010a: 15). ‘… the political settlement is central to all development…’ (DfID, 2010b: 8).‘Every state is based on a political settlement that represents the outcome (but alsoon-going process) of contention and bargaining between elites …’ (DfID, 2010b: 12).‘Political settlements are presented as spanning the continuum from negotiated peace agreements to long-term historical development, in the latter sense approaching the concept of a social contract. Generally speaking, every political regime that is not in the midst of an all-out civil war over its basic parameters is based on some kind of settlement’ (OECD, 2011b: 9). ‘The historically specific dynamics of political settlements explain the difference in performance between countries with apparently similar endowments or disadvantages. Political settlements have been a feature of state building in all states. Every state is based on a political settlement that represents the outcome (but also on-going process) of contention and bargaining between elites, and between social groups and those who occupy authority within the state and society more widely’ (DfID, 2010b:12).
ContentsExecutive Summary 1Introduction 41. What are the various ways in which political settlements have been hdhahas defined? 62. What is the most useful way to define political settlements? 213. Are political settlements distinct from elite pacts, and other related hjdshfasj concepts such as elite bargains and peace agreements? 264. How do GNUs fit into the conceptual terrain? 305. Conclusions and directions for further research 35References 39
1 Executive SummaryA growing body of recent scholarship puts political settlements at the centre of the developmentprocess. The political settlements approach focuses on the formal and informal negotiations, bargains,pacts and agreements between elite actors, as crucial drivers of the locally effective institutions andpolicies that promote or frustrate the achievement of sustainable growth, political stability and sociallyinclusive development. However, alongside a lack of detailed empirical case studies, the usefulness ofthe framework as it stands is hampered by: • lack of consensus over how to define and understand the key concepts A • A tendency in the literature to conflate key terms, i.e. to treat political settlements, elite pacts/ bargains and peace agreements as interchangeable. It is also unclear how governments of na- tional unity should be defined and understood in relation to these other concepts.This paper surveys and clarifies the conceptual field by addressing the following key questions:How should political settlements be defined and understood?1. Political settlements are not simply a form of ‘social-contract’ between states and societies. ‘States’ and ‘societies’ are not, each, unitary entities or actors. They are always characterised by internal differences and a variety of interests and forms and degrees of power, especially in devel- oping societies with weak institutions. There is, therefore, more or less disagreement, division and exclusion within both. This undermines the plausibility of thinking about a social contract between ‘the state’ and ‘the society’.2. Political settlements are on-going political processes that include one-off events and agreements between elites, but are not defined by them. The ‘process’ aspect of the definition is important for political settlements analysis, by virtue of its focus on the on-going politics and the underlying and evolving longer-term elite power relations and negotiations that influence the shape of institu- tions and how players work within and around them. Equally important are the one-off events and agreements that provide rough markers for tracing the point at which settlements begin, end and change.3. Political settlements are typically the outcome of bargaining, negotiation and compromises between elites. This is sometimes formal, open and public, but often informal or less open, and commonly both.4. However, political settlements are also typically played out across two-levels. They involve not only the horizontal negotiations between elites but also vertical relations between elites and their followers. Political settlements are therefore best understood as ‘two-level’ games.5. Political settlements influence the form, nature and performance of institutions. And institutions can in turn help to consolidate and ‘embed’ political settlements.6. Political settlements are not set in stone. They can and should adapt in response to changes in
2 social, political and economic power relations and to contingent events and critical junctures.7. Political settlements can be more or less inclusive of social or political groups. But the stability of settlements over time does not necessarily or always depend on the degree to which they are inclusive, at least in the short to medium term – though longer term settlements may require progressive ‘inclusion’.How should elite pacts and governments of national unity (GNUs) be defined and under-stood?8. Elite pacts/bargains and peace agreements are one-off events that are part of the on-going political settlement. Retaining a distinction between elite pacts/bargains, peace agreements and political settlements is important for explaining how and why one-off pacts and agreements may fail to produce developmentally positive outcomes in the longer-term.9. GNUs are a type of inclusive elite pact aimed at establishing formal shared institutions and organisations of government. They usually occur during or after a time of crisis, but may not last for very long.The paper concludes by suggesting operational messages for the international community, and areasfor further research to build on the conceptual work undertaken here.Operational Messages • Acquire local knowledge and understanding Donors and others will need to acquire a deep understanding of, and sensitivity towards, local political dynamics if they are to make best use of the settlements framework to guide their activities. • Take a long-term approach to influencing political settlements There are no ‘quick-fix’ policy implications of the political settlements framework. Influencing political settlements will require patience and long-term commitment. • nderstand how elites relate to wider coalitions and their support bases U In using political settlements analysis, donors need to be alert to the ways in which elites may come into conflict with wider coalition members, and must also be aware of the pressures on elites to satisfy their supporters. • Work to broker opportunities for elites to come together Negotiation, compromise and bargaining between elites are central to the formation of durable political settlements. Therefore, external players should look for opportunities to facilitate meetings, partnerships, alliances and coalitions between different elites. • e aware of excluded groups B Donors should be aware of the precise boundaries of a political settlement in order to antici- pate possible challenges from excluded groups, and to identify if it may be appropriate to try to broker more inclusive arrangements.
3Further Areas for Research? • he role of external donors T Can the international community work politically to encourage elite pacts and political settle- ments that are developmentally positive? If so, how? • espect for national sovereignty R Political settlements and elite pacts are political arrangements that are local to specific countries and need to be locally legitimate. Further research is required to give an insight into how, and in what ways, the international community can support, broker or facilitate developmentally positive settlements and pacts without compromising national sovereignty, but yet also remaining true to its normative goals. Does ‘policy dialogue’ offer a starting point? • rade-offs between stability and inclusivity T Are stable political settlements inevitably exclusive of certain social and/or political groups? Can the international community encourage stable and robust political settlements whilst also working to promote social, political and economic inclusivity? • lite pacts and GNUs as components of conflict resolution E What are the conditions under which elite pacts and GNUs help to resolve violent conflict and encourage developmental reform? • NUs and developmental outcomes G What have been the developmental results of GNUs? Other than contributing to immediate peace, or handling national crises, have they also contributed to sustained growth and political stability in the medium term? What is the typical ‘length of service’ of GNUs and what has happened after their dissolution?
4 IntroductionThe aim of this paper is to provide a critical survey of the conceptual definitions of political settlements,elite pacts, and governments of national unity (GNUs) that can be found in recent literature on thepolitics of development, and to suggest the most useful way of understanding their meaning. 1Although these concepts are recent additions, they fit broadly into an historical tradition of politicalanalysis that emphasises the importance of powerful actors in society as key determinants of politicalarrangements. For example - whilst they did not articulate their arguments through the specificlanguage of settlements or pacts - a number of late 19th and early 20th Century political theorists,including Gramsci (1929-1935), Mosca (1923), Michels (1911) and Pareto (1901), argued that powerrelationships within and between competing elites were central to understanding the form and functionof political regimes. More broadly speaking, much of the history of classical political theory, from Platoand Aristotle through to Machiavelli, Hobbes and Rawls, has been concerned with the problem of whattypes of political arrangements generate political stability – the necessary but not sufficient conditionfor sustainable growth and inclusive social development. Political settlements, elite pacts and GNUsare part of a renewed focus on the significance of the interactions and negotiations between local eliteactors for understanding the conditions for robust political associations and for developmental reform(Di John and Putzel, 2009: 14).It has become clear to many theorists and donor organisations that transferring technical skills, moneyand resources to aid-recipient countries is not enough to promote developmental goals, in the absenceof locally appropriate institutional arrangements (DfID 2010a; DfID 2010b; Di John and Putzel, 2009;Leftwich: 2011; Parks and Cole, 2010; Gleason et al., 2011). In line with a recent focus on the politicalprocesses that underpin development, the political settlements framework, along with associatedconcepts such as elite pacts and elite bargains, sheds light on new avenues for donor agencies and theinternational community to address the underlying causes of conflict, political instability and poverty indeveloping countries. GNUs have received far less scholarly attention, but they are included briefly inthe final section of the paper because there is some conceptual overlap with settlements and pacts.2Despite the recent attention they have received, the usefulness of the terms ‘political settlement’ and‘elite pact’ is hampered by: • lack of clear definition A • lack of consensus over the meaning and use of the terms, and A • A lack of detailed empirical case studies to support the conceptual claims made on behalf of the1 I would like to thank a number of people for their help. Alan Whaites and Joel Migdal clarified important points on the theoretical and historical background of the political settlements framework and directed me towards early works on the topic. Sam Hickey allowed me to cite his draft paper on settlements, and provided sound guidance on early definitions of the concept. Sue Ingram sharpened my summary of her work on the political settlement in East Timor. I would also especially like to thank Tom Parks for reading and commenting on the paper and for ideas that helped to strengthen the last section. Finally, I must thank Adrian Left- wich for reading and commenting on many previous drafts of the paper and for his support, encouragement, advice and guidance throughout. Any errors or shortcomings are of course my responsibility alone.2 This paper focuses only on political settlements at a national level. For a discussion of sub-national political settlements, see Parks and Cole (2010).
5 concepts.3There is some consensus that these terms describe political arrangements between elite figures andthat they are important for analysing the political dynamics of conflict-resolution, state-building, anddevelopmental change.4 But what this paper shows, in the first instance, is that once authors startunpacking the concepts in more detail and then applying them to specific case studies, they divergein subtle but important ways. The term ‘political settlement’ is used to describe a range of differentpolitical phenomena, from peace agreements, one-off events such as the signing of a declaration or billof rights, through to long-term political processes of state-building (Parks and Cole, 2010: 5-6; Evans,2012: 7-8). Moreover, there is a tendency in the literature to conflate key concepts, i.e. to treat politicalsettlements, elite pacts/bargains and peace agreements as interchangeable. Box 1. Key points for clarification i. What are the various ways in which political settlements have been defined? ii. What is the most useful way of defining political settlements? Are they best understood as on-going political processes? Or do they represent a specific, clearly articulated agreement, or a one-off event? iii. Are political settlements distinct from elite pacts, and other related concepts such as elite bargains and peace agreements? If so, what is the most useful way to differentiate these concepts? iv. How do GNUs fit into the conceptual terrain? Do they presuppose prior elite pacts or political settlements or are they constitutive parts of them in some cases?These four questions are addressed in sections 1 through 4 respectively. By aiming for conceptualclarity, the hope is that donors and policy-makers will have more useful conceptual tools to hand foranalysing the political processes at work in developing countries, and for understanding how best toinfluence positively those processes. 3 Laws (2010) looks at the 17th Century revolutionary settlement in England through the political settlements lens. A forthcoming stream of research work from the DLP will be looking at how the developmental achievements in Somaliland have been shaped by the country’s political and economic settlements.4 According to Gleason et al. (2011: 11) state-building involves “fostering the development of strong, well run, democratic states that serve their citizens with transparency and accountability”.
6 01 What are the various ways in which political settlements have been defined?1.1 IntroductionThis section provides a critical survey of recent accounts of political settlements in the literature onthe politics of development, points to areas of divergence, and draws out key points that should beincluded in a robust definition. Box 2. Chronological summary of the key definitions and functions of political settlements Khan (1995) Definition of a political settlement: the overall balance of power in society. Functions of a political settlement: explains the performance of formal and informal institutions. Fritz and Menocal (2007) Definition: a negotiated agreement binding state and society. An on-going process, rather than a one-off event. Functions: provides the foundation for core state functions. OECD/DAC (2008) Definition: The outcome of peace processes. Functions: underpins the social contract. Whaites (2008) Definition: informal, unarticulated understandings between elites. Functions: brings about an end to conflict and prevents its re-occurrence. Menocal (2009) Definition: a common understanding between elites about how power should be organised and exercised. Includes formal institutions and informal agreements. Functions: determines the success or failure of peace-building and state-building. Di John and Putzel (2009) Definition: the outcome of bargaining and negotiation between elites. Functions: constrains and facilitates institutional and developmental change. Brown and Gravingholt (2009) Definition: a two-dimensional concept - the outcome of an historical event and a property of society. Functions: the cornerstone of every social and political order. Leads to institutions that prevent violence. DFID (2009) Definition: starts with a common understanding between elites then expands into a contract between state and society. An adaptable political process, formalised through or grounded upon one-off events like peace agreements. Functions: stabilises peace agreements and establishes state-society relations. Barnes (2009) Definition: a common understanding between elites about how power should be organised and exercised. The outcome of peace processes in war-to-peace transitions. Functions: Defines the nature of institutions and the role and power of political organisations and their members. Khan (2009, 2010) Definition: the ‘social order’; a compatible, viable and sustainable combination of power and institutions. Functions: describes how societies solve the problem of violence and achieve minimum political stability and economic performance. DFID (2010a) Definition: a common understanding between elites about how power should be organised and exercised. Includes formal institutions and informal agreements. Functions: establishes basic rules governing economic relations and resource allocation.
7 DFID (2010b) Definition: ongoing formal and informal bargains between elites. Functions: determines the extent to which countries successfully overcome internal conflict, promote developmental reform and reduce poverty. Parks and Cole (2010) Definition: the arrangements that elites agree to in order to end violent competi- tion over power and resources. Settlements are characterised by elite actors, pursuing their interests, thereby influencing the shape of institutions. Functions: determines the nature of development, gover- nance and security institutions. Gleason et al. (2011) Definition: dynamic, on-going negotiations between elites. Functions: connects states and societies by enabling the state to fulfil core functions and provide key services. Sharan (2011) Definition: a framework for ending hostilities amongst competing elites. Functions: helps to secure peace and the post-conflict containment of failed states and their disorderly elites. AusAID (2011) Definition: the formal and informal institutions, agreements, and understandings, between elites and between elites and the wider society, that underpin a political system. Functions: defines how power is organised and exercised in society. Central to the peace-building and constitution-making process in fragile and conflict-affected countries.Four interrelated themes can be identified in this survey. Some authors side very clearly with one ofthese themes, whilst others draw on a combination of two or more.i. Political settlements describe state-society relations. They are a type of contract between states and societies, the public acceptance of which is central to understanding how states perform core functions and attain public legitimacy.ii. Political settlements are the product of elite negotiations, which shape the character and perfor- mance of formal and informal institutions governing political and economic behaviour (the ‘rules of the game’). They are central to understanding the way that political and social rules operate in a society, and the degree to which those rules support core state functions.iii. Political settlements reflect the overall balance of power in society. They are central to explaining the degree of economic and political stability in a country.iv. Political settlements are the means through which societies overcome violence and political instability. They are central to understanding the elite dynamics behind peace agreements in the aftermath of internal conflict, and arrangements that prevent a return to violence.The overarching theme in all accounts of political settlements is that they are the precondition forpolitics.5 In other words, a political settlement is only in place where competition for power andresources occurs through institutionalised non-violent political processes.1.2 The origins of the concept6The term ‘political settlement’ appears explicitly in the work of the historical sociologist Joseph Melling(1991) in his study on industrial capitalism and the welfare state in 19th century Britain (cited in DiJohn and Putzel, 2009: 4).7 Although no precise definition of the concept is provided here, Melling usesthe term to describe the political processes that shaped the emergence of social welfare policies. Anearly reference to political settlements within development studies can be found in Khan (1995: 71),5 Appropriate forms of politics, in turn, are the preconditions for stability, sustainable economic growth and inclusive social develop- ment. This point, alongside the prior questions of what politics is and why it is so important for development, is discussed in detail in Leftwich and Bavister-Gould (forthcoming, 2012).6 The remainder of this section traces the various definitions of political settlements according to themes, rather than chronological order.7 To the best of the author’s knowledge, Melling (1991) is the first example of the concept of a political settlement being used in an analytical sense.
8where the author analyses the performance of institutions by means of the “inherited balance of poweror ‘political settlement’”. The concept was later defined by Khan (2004: 168) as the “distribution oforganizational and political power between competing groups and classes”. Khan’s more recent workon settlements is discussed in sub-section 1.10 below.1.3 Political settlements as a state-society agreementThe concept of the political settlement did not gain significant traction in development studies until itwas picked up by Fritz and Menocal (2007). The authors diverge from the Khan definition by puttingforward an interpretation of settlements as a “negotiated agreement (at least in principle) bindingtogether state and society and providing the necessary legitimacy for those who govern over thosewho are ruled” (ibid.: 27).8 The political settlement lays the basis for core state functions: public admin-istration, security and the rule of law (ibid.: 24, 27). The political settlement is not a “one-off event”but an on-going political process. Settlements can adapt to accommodate changes in the relationshipbetween states and societies that arise due to different needs and emerging demands (ibid.: 27).91.4 What is the significance of elite actors in the political settlements framework?What is conspicuously absent from the definition of political settlements in Fritz and Menocal (2007)is the role of political, economic or social elites.10 Elites and inter-elite negotiations occupy a centralplace in all of the discussions of settlements that followed. By omitting any reference to the ‘horizontal’negotiations between elites that most others regard as central to the political settlements framework,Fritz and Menocal (2007) develop a model that is based exclusively on the idea of a ‘vertical’ agreementbetween states and societies. The focus, in later studies, on horizontal elite negotiations is a key aspectof the settlements framework, providing an insight into the political dynamics that promote or hinderdevelopmental goals. For example, according to AusAID (2011: 13), the political settlement is “anagreement between elites on the processes by which power and resources are shared and decisionsmade”.11 Developing states are typically under the control of a coalition of elite factions, which competeamongst themselves for access to resources and political power (Parks and Cole, 2010: 6). This pointcan be illustrated by the example of elite factions in the Philippines (Box 3). Box 3. The political settlement in the Philippines (Parks and Cole, 2010: 6) Most of the key political and economic decisions in the Philippines are controlled by a small group of elites, consisting of long-established families, newcomers who have gained power through holding prominent positions in government, and elites from the private sector with ties to government. As the economy grew in the post-colonial era, the land-holding elites became involved with the private sector, while a new group of governmental elites exploited their position to secure rights to resources and markets. These elite factions exert their influence on government policy through closed, informal networks that can shift rapidly from alliance to rivalry. The slow pace of developmental reform in the Philippines has been attrib- uted to the self-interested activities of these elites.8 The settlements framework outlined in Fritz and Menocal (2007) was picked up again in OECD/DAC (2008). Whilst the concept is not explicitly defined in this paper, it is used to describe the institutional foundations for state-building in fragile and conflict-affected states.9 There is a dearth of examples in Fritz and Menocal’s paper that would help to illustrate these abstract claims. But presumably, by implication in the definition outlined above, any change to a state’s institutions, service delivery or justice system will reflect a re- negotiation of the underlying political settlement.10 The authors do acknowledge that this is a significant omission of their paper (Fritz and Menocal, 2007: 21 fnt. 18).11 The definition of settlements in AusAID (2011) as an elite agreement could be taken to suggest that settlements are a one-off, static event. However, there is a conceptual drift, as the paper develops, towards a definition of settlements as on-going political processes:. For example: “political settlements evolve over time as different needs, demands and tensions arise” (ibid.: 92). In sec- tion 2 of this paper, I will argue that, although one-off events and agreements are part of the on-going political settlement process, settlements should not be defined as any particular event or agreement.
91.5 How useful is the state-society definition of political settlements?The idea of a political settlement as a negotiated agreement between states and societies is ratherabstract in its construction, and it is difficult to identify any solid empirical or historical cases wheresuch an agreement, or ‘social contract’ between ‘the state’ and ‘society’, can be shown to have takenplace. Moreover, the notion of an ‘agreement’ between ‘state’ and ‘society’ incorrectly assumes that(any) ‘society’ is a unified entity, without a diversity of more or less distinct social, economic and politicalgroups with more or less different interests and, especially, different forms and degrees of de jure orde facto power. Such an assumption also seriously overlooks the fact that not all such groups sharea common relationship with, or have comparable access to, the state, nor do they necessarily enjoycomparable civil, political or economic rights. Developing (and all other) societies typically encompasssignificant social, ethnic, religious and gender groups who are excluded from, or marginalised withrespect to their access to, or participation in, formal institutions and organisations of the state. Forexample, the majority of minority ethnic groups in Nepal – accounting for almost 70% of the popu-lation – are disenfranchised (World Bank, 2011: 90). And although they may have other routes ofinfluence and participation, the widespread exclusion and limited political representation and participa-tion of women in many societies, past and present, is well documented (Phillips, 2011b; Hodes et al.,2011; Tadros, 2011). So, if there are groups in society that lack the capacity to aggregate and articulatetheir interests, or are denied the right to do so, and hence are without ‘agency’, it is difficult to see howthat ‘society’, conceptualised somehow as a single entity, can be party to a negotiated agreement withthe ‘state’. While this raises important questions about the complexity of the politics of institutionaland policy change (and introduces the important notion of legitimacy), the idea of a political settlementas a state-society agreement papers over these divisions, disagreements and exclusions.By recognising the fact of divided interests and excluded groups in all societies - and especially indeveloping societies - donor organisations will be more alert to opportunities to engage with pro-development factions and coalitions, to encourage the inclusion of marginalised groups. For example,efforts to encourage a more inclusive, stable, and developmentally-progressive settlement in Thailandin the late 1990s centred on identifying and seeking support from sympathetic groups in the emergingThai middle class, the academic elite, and members of the business elite in Bangkok. Since the 2006peace agreement in Nepal, the international community has supported a more inclusive constitutionand encouraged political participation on the part of previously excluded ethnic groups (Parks andCole, 2010: 14, 27). A robust definition of political settlements should therefore be alert to the fact ofsocial divisions in developing societies.1.6 Political settlement as a two-level gamePolitical settlements need to be responsive to the interests of the non-elite general public if they are tobe sustainable over time, and in this sense Fritz and Menocal’s (2007) definition is helpful. However, theexample of the transition to democracy in South Africa, discussed below (Box 4), suggests that what isreally at stake in the politics of developmental change is not so much agreements between states andsocieties, but rather the relations and interactions between key elites, and between those elites andtheir respective groups of followers – and that it is this combination of horizontal and vertical relationsthat should be at the centre of the definition of political settlements. If the negotiating process inSouth Africa had been exclusive in either the vertical or the horizontal sense (i.e. if key elites had beenexcluded from or left unsatisfied with the terms of the settlement, or if those key elites had beenunable to retain support for the settlement from their followers) then it is unlikely that the ensuingsettlement and the new democratic order would have been sustainable over any length of time.
10Box 4. The South African transition to democracyThe transition to democracy in South Africa came to involve horizontal negotiations between theapartheid regime, represented by the National Party (NP), and the African National Congress(ANC). By the mid-1980s, the two parties had reached a stalemate. The ruling coalition ingovernment had control over the state, the police and defense force, whilst the ANC enjoyedwidespread popular support amongst the general populace, the trade unions, civic groups, andinternational advocates. Escalating civil unrest, violence and mounting international pressure onthe NP made governance unmanageable. Horizontal negotiations became unavoidable.The negotiating process was highly inclusive along the horizontal dimension, in that all politicalparties across the spectrum, from left to far-right, were urged to participate. Along the verticalaspect, both parties had to satisfy their respective followers and constituencies, as well as thenational interest more generally. To be sustainable, the new political settlement had to addressthe aspirations of the black majority whilst assuaging the fears of the white minority (Nathan,2004: 2-3). The debates over the terms of the new constitution were vertically inclusive in thesense that the non-elite members of political parties, members of the general public, and repre-sentatives from a broad range of civil society groups, were formally consulted throughout theprocess (ibid.: 4). The ANC established a Negotiations Commission that met regularly to conductregional briefings and consultations between the elected National Executive Committee (thebody with formal decision-making authority) and grassroots ANC supporters (Maharaj, 2008:25). On the side of the National Party, a referendum was conducted solely within the whitecommunity, the results of which gave the party leader F. W. de Klerk a mandate to negotiate withthe ANC. As nearly 88 per cent of registered white voters cast ballots, the endorsement had ahigh degree of vertical legitimacy (Wren, 1992).However, the fractious relationship between the ANC and the country’s third largest electoralparty, the Inkatha Freedom Party (IFP), was the cause of considerable violence and instability inthe run up to the 1994 elections. According to Sisk (1995: 95), the violence of the IFP supportersstemmed from their perception of being excluded from the close political relationship that haddeveloped between the ANC and the NP, and their fear of being sidelined from the negotiationsthat preceded the elections.In response to the escalating unrest, a peace deal was established between the ANC and IFP onJanuary 29th, 1991, which established the legitimacy of the IFP as a formal electoral body witha seat at the negotiating table (ibid.: 100). After the first democratic election in South Africa in1994, a Government of National Unity was formed and the IFP President, Mangosuthu Buthelezi,was appointed as a cabinet minister. His inclusion in the government helped to assuage the fearof the IFP leaders and their followers that the traditional and tribal structures of authority wouldbe wholly swept away by the new democratic order (Sisk, 1996: 299). This example underscoresthe significance of the goal of inclusion in the political settlements framework – a point that isdiscussed in more detail in sub-section 1.11.
11The difficulties noted above with the state-society definition of settlements are avoided by the defini-tion put forward in a later DfID-commissioned paper by Alan Whaites. Horizontal understandingsbetween elites are at the centre of Whaites’ (2008: 4) definition; political settlements are “the forging ofa common understanding, usually among elites, that their interests or beliefs are served by a particularway of organising political power”. Whaites’ (ibid.) discussion also refers to the importance of thevertical relations between elites and their followers: “elites can rarely take social constituencies forgranted, they must maintain an ability to organise, persuade, command or inspire. Wider societies arenot bystanders in political settlements…”.12 In this sense, he regards political settlements as a two-levelgame.1.7 Political settlements in responsive and unresponsive states (DfID 2009)The horizontal and vertical aspects of political settlements are also at the centre of the discussion ofstate responsiveness in DfID (2009). In ‘responsive states’, political settlements start with a commonunderstanding forged between elites, then over time the settlement expands to reflect a broaderunderstanding between state and society, as the former delivers on basic functions (security, revenueand the rule of law) and accountability. In ‘unresponsive states’, the political settlement is characterisedby informal rules, patronage systems, and tensions between competing, predatory elites. Broadeningthe inclusivity of the settlement so that it includes all major social groups, i.e. strengthening the hori-zontal dimension referred to above, is a key element in the transition from unresponsive to responsivestates. Political settlements in developing countries are likely to feature a combination of responsiveand unresponsive features, as the examples of Cambodia (Box 5), Papua New Guinea (PNG) and theSolomon Islands (Box 6) illustrate.13 Box 5. The political settlement in Cambodia (DFID, 2009: 9) Responsive features • The settlement is relatively ‘robust’, in so far as the governing Cambodian People’s Party has uncontested authority, holding an outright majority in the National Assembly and the Senate. • Progress has been made on establishing key ‘survival functions’ of the state, including secu- rity and macro-economic stability. Unresponsive features • There has been “uneven” progress on many core state functions; particularly on combating corruption and upholding human rights. • isputes between competing elites tend to be resolved through ‘personal channels’ rather D than formal institutional mechanisms. • here is a risk that competition over capturing anticipated rents from oil and gas reserves T could provide a new source of conflict.12 Parks and Cole (2010: viii), DfID (2010b: 17) and Gleason et al. (USAID 2011: 5) put forward a similar two-level game definition of political settlements. Levi and Murphy (2006: 655) argue that effective coalitions are engaged in a “nested or two level game” when negotiating with representatives from other organisations.13 North et al. (2007) articulate the idea of responsive and unresponsive states through the concepts of open and limited access orders. Although they do not refer explicitly to ‘political settlements’, their work fits into the pattern of analysis that the settlements framework brings into focus, by recognising the centrality of elite negotiations in accounting for the stability and responsiveness of states.
12Box 6. The political settlements in PNG and the Solomon IslandsBoth PNG and the Solomon Islands have political settlements based upon a traditional system of kinshipand land-ownership known as wantok. The wantok is the primary source of clan loyalty for individuals;it is a system of affiliation based upon common language, kinship group, geographical area and religiousgrouping (Renzio, 2000: 21-22). In both countries, governance is impeded in certain respects because ofthe absence of a ‘fit’ between formal state institutions and the wantok (Fukuyama, 2007a: 4).The PNG state has struggled to deliver certain core public services – per capita income has risen onlyvery marginally since the country’s independence in 1975, primary school attendance rates are extremelylow, and roads, schools and health facilities are poorly maintained. These difficulties cannot be attributedto a lack of resources; the country has a wealth of timber, minerals, agriculture and energy (ibid.: 2). Thedifficulty stems instead from the fact that the traditional balance and organisation of political power (i.e.the political settlement) that is shaped by the wantok system conflicts with formal state institutions in away that often impedes effective governance and the provision of public services (ibid.: 4, 6-7). PNG hasa Westminster-style electoral system with single-member constituencies and first-past-the-post voting.When transposed onto the wantok system, this structure produces highly unstable ruling coalitions basedupon personality and patronage. Candidates are typically elected with as little as 10 per cent of a district’svotes. Very small changes in voting behaviour can result in different candidates coming to power, so thereis a frequent turnover of leaders (Fukuyama, 2007b: 6). Ruling coalitions, in turn, are primarily concernedwith seizing the largest amount of resources to distribute to their respective support bases – defined bywantok affiliation – rather than providing public goods on a national basis. Politicians have little incentiveto tackle social problems that do not affect their constituency and there is no overarching national reformparty (ibid.: 12).In the Solomon Islands the wantok system has similarly caused difficulties for certain functions of gover-nance. Whilst it has produced key social benefits at a local level – including comparatively low levels ofcrime, drug use and HIV/AIDs – there are elements of public policy and collective action on a larger scalethat are frustrated (Monsell-Davies, 1993: 1). For example, commercial agriculture and foreign investmentare constrained by the structure of customary land tenure (Fukuyama, 2008: 8). Moreover, as in PNG, thefirst-past-the-post electoral system typically results in unstable ruling coalitions that capture resources fortheir own support bases (ibid.: 10).In the Solomon Islands, the Provincial Government Act of 1981 created a two-level government whereauthority is divided between an elected premier – at the national level – and a provincial assembly.According to a Diagnostic Study in 2004, the structure of authority has resulted in an unresponsivepolitical settlement because it has been left unclear how authority is to be divided between provincial andcentral government. For the most part, locally elected officials are unclear of their powers and responsi-bilities, and they see themselves as competing with national MPs that represent their districts. As a result,there is little cooperation between the two levels and so the political settlement as a whole fails to delivera coherent programme of governance and public policy (ibid.: 12). In a paper presented at the AustralianNational University, a coordinator working for the Regional Assistant Mission to Solomon Islands (RAMSI)identified a number of long-term challenges facing the Islands, including inadequate government services,a lack of economic opportunity, and failures in the delivery of law and justice (Wilson, 2009: 9).It has been suggested that one important way in which the political settlement in both PNG and theSolomon Islands could be made more responsive to the needs of the general public would be toencourage the emergence of an overarching sense of national identity that would complement wantokloyalty (Fukuyama 2007a; Fukuyama 2007b; Fukuyama 2008). If such an identity is to emerge, it will mostlikely need to be created and managed through the activity of a class of elites. This, in turn, implies thatnation-building in the Western Pacific may well rely upon appropriate forms of secondary and highereducation, to create leaders with a national consciousness and loyalty to the overarching state (Fukuyama2008: 19, 27).
131.8 Political settlements as adaptable processesIn his paper for DfID, Alan Whaites (2008: 7) suggests that a single political settlement can endureacross numerous changes of government, so long as the structures and institutions of power, definedby horizontal negotiations and understandings between elites, remain more or less intact. But to besustainable in the long-term, a political settlement will have to adapt in response to social develop-ments. States that are responsive towards demands for political or social change are, by implication,founded upon an adaptable political settlement (ibid.: 10). However, settlements are not infinitelymalleable. When there is a fundamental shift, a critical juncture or a “step-change” in the nature ofpolitical power and/or institutions, a new settlement may emerge. Such changes occurred with the fallof communism in Eastern Europe and the end of apartheid in South Africa (ibid.: 7). Box 7. Political settlements can adapt in response to social developments A settlement that excludes an emerging social class will likely face pressure to expand the franchise. The 1832 Great Reform Act in Britain fundamentally altered the terms of the political settlement in place at the time, by introducing wide-ranging changes to the electoral system in England and Wales. The act was designed to give greater representation to the cities of the industrial North, which had been growing steadily in size and economic power. The Act increased the electorate from around 366,000 to 650,000 – from approximately 5% to 7% per cent of the total adult-male population in England and Wales (Cook and Stevenson, 2001: 81). Although the 1832 Reform Act did not significantly broaden the franchise, it initiated a process whereby the political settlement gradually became more inclusive, as more people from the lower classes – and eventually women, in 1918 – gained political representation through further Acts. The 1867 Reform Act extended the right to vote further down the class ladder, adding approximately one million voters drawn mostly from the working classes, thereby doubling the electorate to two million and enfranchising about 16% of the population. The 1884 and 1885 Redistribution Acts tripled the electorate again, by extending the vote to most agricultural laborers – by which point 28.5% of adult males had the vote. Women were excluded from the franchise until 1918, when women over thirty were given the vote. In 1928, the Equal Franchise Act amended this age limit to twenty-one, with the result that 97% of the total population were entitled to vote (ibid.). This example illustrates how political settlements can expand to reflect the changing nature of social relations and class structures once an institutional precedent, or principle, for inclusion has been set.1.9 Political settlements as the product of elite bargainsThe papers discussed above define political settlements as a reflection of a common understandingbetween elite actors. Di John and Putzel (2009: 4) argue that this emphasis on common understand-ings is unrealistic in its suggestion of agreement and consensus. Instead, they define political settle-ments as the outcome of bargaining, primarily between elites but also more generally between stateand society actors, whose interests may be in contention.14 They argue that political settlements do not14 Arguably, Di John and Putzel (2009) overstate their objection to the idea of ‘common understandings’ in previous definitions of political settlements. These definitions can still accommodate the point that, in reaching a common understanding or agreement, elites had to go through a process of bargaining and negotiation.
14reflect common understandings, but more often emerge from “a narrow or forced bargain imposedby authoritarian regimes…compromises between previously warring parties, or…more pluralistbargaining arrangements, as in democratic regimes” (ibid.: 4).The significance of bargaining and compromise in the formation of a political settlement can be illus-trated by the experience of Tanzania and South Africa. Despite its high level of ethno-linguistic andreligious diversity, Tanzania is one of the most stable and cohesive countries within the conflict-pronearea of the East Africa. Notwithstanding two military coup d’état attempts, in 1969 and 1982, nosignificant opposition to the ruling party has emerged to contest state power, despite a series ofeconomic crises, widespread poverty, and external security challenges. One explanation for the stabilityof the Tanzanian political settlement is that, from the outset, the post-independence rulers establishedinclusive state institutions through which traditional leaders and tribal authorities were encouraged tobargain for privileges, power and resources. Instead of looking to circumvent or over-throw state insti-tutions, traditional leaders and elites were given the opportunity to seek compromises with govern-ment (Gutierrez, 2011: 12).15Bargaining and compromise between the ANC and the apartheid regime was a critical element in thesuccessful transition to democracy in South Africa. This was underscored by Mandela (1995: 513),who wrote that: “Both sides regarded discussions as a sign of weakness and betrayal. Neither wouldcome to the table until the other made significant concessions”. During the course of the negotiations,the main protagonists made a series of reciprocal compromises. In return for the release of politicalprisoners, the ANC agreed to a unilateral cease-fire; in return for the repeal of racist legislation by theGovernment, the ANC took the initiative in lifting the sporting and cultural boycott of South Africa; inreturn for the Government’s agreement to set out a road map for the negotiations with definite time-frames, the ANC agreed to lift economic sanctions (Haysom, 2004: 5). Bargaining and compromisesbetween elites – whose immediate political interests may be in contention – should be at the centreof our understanding of political settlements.1.10 Political settlements shape institutionsAccording to Brown and Gravingholt (2009: 1), political settlements have two separate dimensions:“the fixed outcome of a certain historical event and a particular characteristic or property of a society,reflected in the conduct of political actors”. The ‘event’ dimension of settlements describes occasionswhere political participants fundamentally re-negotiate the institutions of their polity; for example, bysigning a peace accord or establishing a new constitution (ibid.: 7). On the ‘property’ dimension, politicalsettlements are characterised by the decisions and behaviour of political actors, which lead to theformation or re-negotiation of formal institutions along the ‘event’ dimension. The meanings of ‘fixedoutcomes’ and ‘properties of society’ are highly elusive and these terms do not give a clear enough indi-cation of how to understand and recognise political settlements. A plausible explanation of what theauthors have in mind is that settlements are defined partly by particular political events and institutions,as well as on-going political processes. I argue in support of this definition in the following section.Parks and Cole (2010: 6) put forward a three-part definition of the key elements of a political settle-ment: (i) a coalition of elite actors; (ii) operating in the pursuit of their interests; (iii) to establish orreshape political institutions, which include formal state institutions and informal arrangements. Thehistorical example of the political settlement in Belgium illustrates the importance of including institu-15 Further to this, the Land Reform Act and the Village Land Act in 1999 decentralised land administration in rural areas, awarding responsibility for recognising land rights and titles to elected village governments with a high degree of local legitimacy. This ‘down- wards’ devolution of power has the potential to make land transactions more transparent and accountable. However, work remains to be done to improve the implementation of the system in practice. A lack of resources and training for local authorities, incoher- ence within the administration structure and an absence of enforcement mechanisms have obstructed the effective functioning of the programme (Pedersen, 2011).
15tional development within the definition of political settlements. Box 8. The evolution of the political settlement in Belgium through the interaction of elite actors and institutions Since its creation in 1831, Belgium has always been multi-national, and composed of Dutch (also referred to as Flemish), French and German speakers (alongside a host of more local dialects). In the 1930s, language laws were introduced that made Dutch and French the official languages of administration in Flanders and Wallonia respectively. However, Belgium did not become a federal state, with formal borders demarcated by the language spoken, until 1993. In the 1960s, a group of Flemish politicians emerged who had been educated entirely in Dutch-speaking schools, and who were consequently highly vocal in demanding strict observance of the language laws. This helped to fuel demands for devolution. The political resources of the Flemish were further strengthened when, in the aftermath of the decline of the post-war coal and steel industries in Wallonia (the heart of French-speaking Belgium), the bulk of socio-economic power and policy- making shifted to Flanders, where Dutch was the official language. According to Swenden (2003), the political and economic power of the Flemish political elites enabled them to establish their territorial view of language politics, against the resistance of minority pockets of French speakers in and around Flanders. In this way, the current political settlement in Belgium has been shaped by the informal interaction and negotiations between elites, on the one hand, and formal institutions, i.e. the demarcation of linguistic autonomy within a federal system of rule, on the other.1.11 Mushtaq Khan on political settlements, power and institutionsThe relationship between political settlements and the performance of institutions is also at the centreof Khan’s work (Khan, 1995, 2004, 2010, 2011, 2012). For Khan, a political settlement only emergeswhen the distribution of economic benefits, supported by both formal and informal institutions, iscompatible with the distribution of power in society, and where this is sustainable, i.e. economically andpolitically viable.16 Khan uses the political settlement as an analytical tool for understanding the gover-nance issues that are necessary for sustaining growth, and to explain why apparently similar institutionscan work very differently in different countries (Khan, 2010: 4).Khan’s definition goes beyond those accounts that look just at agreements, compromises or bargainsbetween elites, by focusing at a deeper level on the compatibility (or lack thereof) between institutionsand the distribution of power and economic benefits amongst powerful groups in society: “A politicalsettlement is a combination of power and institutions that is mutually compatible and also sustainablein terms of economic and political viability” (ibid.). Khan defines power in this context as holding power,which is “the capability of an individual or group to engage and survive in conflicts” against other eliteactors or the state itself (ibid.: 6).16 Acemoglu and Robinson (2008) use the term ‘political equilibrium’ when arguing that political and economic institutions need to be compatible with de facto political power if they are to be sustainable over time. They do not offer a precise definition of ‘political equilibrium’ but they use the term in a way that is very similar to Khan’s (2010) definition of political settlements discussed above, i.e. as the underlying balance of power between elite members of the dominant coalition, which influences the form and developmental performance of economic institutions. Neither ‘political settlement’ nor ‘political equilibrium’ (nor any equivalent concept) was used in their most recent book, Why Nations Fail (2012), where they focus instead on ‘inclusive’ and ‘extractive’ institutions. It could be argued that the language of ‘equilibrium’ suggests a kind of stasis that is at odds with the dynamic character of the political processes and elite power relations which the settlements framework brings into focus. On this point, see Sen (forthcoming 2012).
16A sustainable political settlement is one that creates “minimum levels of economic performance andpolitical stability” for a society to function over time (ibid.: 4). To say that a political settlement is acombination of power and institutions that are ‘compatible’ means that powerful groups are derivinga sufficient share of economic benefits for them to comply with the institutional order. When elitesdo not comply with the institutional order, the settlement has broken down. Elites will typically striveto shape the character and evolution of institutions in order to create the distribution of benefits thatthey desire (ibid.).To say that the combination of power and institutions in society is ‘economically viable’ means thatthe political settlement is sufficiently productive to avoid economic crisis. To say that the combina-tion of power and institutions is ‘politically viable’ means that conflict and violence do not reach levelsthat threaten core institutional and political arrangements (ibid.: 20-21). Where there is compatibility,sustainability and viability in the structure of power and institutions in society, a political settlementwill be able to reproduce itself over time. In the absence of a political settlement, or in the event thata political settlement breaks down, a society is likely to be characterised by violence or anarchy, andeconomic development is highly unlikely (ibid.: 8).According to Khan, institutions and the distribution of power in society are dependent on one another:institutions affect the distribution of power by creating economic benefits for different groups, whilstthe distribution of power affects institutions because powerful groups will work to create the institu-tions that will deliver them favorable economic benefits (ibid.: 22). Once this combination of powerand institutions is compatible, viable and sustainable (as defined above), a political settlement hasemerged. However, robust political settlements are not static. They must be characterised by flexibledistributive arrangements in order to respond to “ongoing social mobilizations”, which will continu-ously alter the distribution of political power (ibid.: 7). The kind of incremental and gradual changethat occurs as a compatible, viable and sustainable settlement responds to shifting power dynamics insociety is very different to the sharp disruption that can occur when a settlement breaks down due toelite dissatisfaction with their share of benefits (ibid.: 4-5).In developing countries, the distribution of power underlying the political settlement is shaped heavilyby informal institutions – primarily patron-client rules governing the allocation of economic benefits.Developing countries typically have in place what Khan refers to as a ‘clientelist political settlement’,where elite power is based largely on incomes that are generated outside formal institutions andoutside the productive, capitalist sector of the economy (ibid.: 8, 53-58).17 In clientelist political settle-ments, there are two key factors that can explain differences in the performance of institutions. Thefirst is the power of excluded factions relative to the power of the ruling coalition, and the power oflower-level supporters. This reflects the significance of vertical and horizontal relations in the politicalsettlements framework, discussed above. The greater the power of excluded factions, the greater thevulnerability of the ruling coalition, which will affect the kind and duration of economic policies it willbe willing to undertake. The greater the power of lower-level factions, the weaker the capacity ofthe ruling coalition to enforce institutional rules upon them (ibid.: 9). The second factor that accountsfor differences in the performance of institutions is the technological capabilities of entrepreneurs insociety and their relationship to the ruling coalition – a factor that is illustrated clearly by Khan’s discus-sion of the political settlement in Thailand, summarized in Box 9 below (ibid.).Khan’s analysis has important implications for the goal of creating ‘good governance’ in developing17 By contrast, advanced countries typically have ‘capitalist political settlements’, where the economy is dominated by the formal pro- ductive sector (Khan, 2010: 8). Khan (ibid.: 49) describes two further types of political settlements. ‘Pre-capitalist’ political settlements are characteristic of feudal societies where land rights and military rights are aligned with the distribution of political/military power in society. ‘Political settlements in crisis’ are characteristic of societies where formal institutions have mostly collapsed and informal economic activity is based primarily on the threat or practice of violence.
17countries. It suggests that the creation of the formal institutions characteristic of economically advancedcountries (for example, democratic elections and accountability mechanisms) is an unrealistic short-term aim, because the power of local elites – who can act as ‘spoilers’ in the process of sustaining suchformal institutions – is typically derived from informal institutions (ibid.: 4-5, 127). Given that a robustpolitical settlement is a combination of institutions and power that are supportive of one another, anyattempt to bring about institutional changes must be compatible with this structure, or it will be resistedby elites (ibid.: 5). Resistance can take the form of legal intervention, non-compliance, withdrawals ofinvestments, the organisation of strikes, and violence (ibid.: 31-32). Institutions are more likely to beeffective and sustainable if the distribution of benefits under those institutions are not contested bygroups with holding power (ibid.: 6). Conversely, conflict is liable to break out when groups feel thattheir holding power is not adequately reflected by their share of benefits (ibid.: 7).Alongside a host of detailed case studies, Khan has used this framework to explain the performance ofeconomic institutions in Thailand. Box 9. The clientelist political settlement in Thailand between 1960-1970 (Khan, 2010: 76-78) From the 1932 revolution until the student uprising of 1973, the Thai political settlement was char- acterised by competition and conflict between various military-bureaucratic factions. Generals developed a practice of seeking out business contracts through ‘kickbacks’ and taking on appoint- ments to the boards of companies. Between 1960 and 1970, the military-led ruling coalition did not acquire the characteristics of a developmental coalition, primarily because factionalism within the army meant that generals built their own competing economic power bases. These features of the political settlement in Thailand help to explain the shape of economic institutional performance at this time, and also why a more sophisticated industrial sector - as experienced in South Korea and Taiwan, for example - did not emerge. The kind of institutions required for industrial policy (e.g. based on technology acquisition) would have been difficult to impose on a system where capitalists were virtually always the clients of a powerful general. On the other hand, this clientelist political settlement was highly conducive to other types of growth-enhancing formal and informal institutions. The capital base of the emerging capitalist class was enhanced by the informal system of granting favourable government contracts for construction, imports and exports, and sectors such as logging. Entry into the market by indi- vidual capitalists was relatively easy provided they could pay for informal protection and support from the authoritarian ruling coalition. This meant that capital accumulation and start-up support were quickly accessible, leading to rapid growth, particularly in technologically straightforward manufacturing industries.1.12 Political settlements and inclusivityThe concept of ‘inclusivity’ is a key feature of the literature on political settlements, and is said to accountin large part for the stability and legitimacy of states (Barnes, 2009: 3-4; DfID, 2010a: 7, 14; Parks andCole, 2010: 8; Whaites, 2008: 18). When excluded groups have sufficient numbers and resources,they are likely to challenge the terms of the political settlement from which they are excluded, and itsdistributional implications. For example, DfID (2010b: 13) traces the outbreak of civil war in Ugandaand Cote d’Ivoire to the fact that major groups were systematically excluded from power.18 Similarly,18 The recent experience of coalition politics in Uganda is also discussed in detail in Kjaer and Katusiimeh (2012). The authors note
18Tadros (2012) identifies the spearhead for the Egyptian uprising in 2011 in the socially, economicallyand politically excluded youth groups, citizen groups and civil society activists who - although theylacked elite representation or a coherent organisational structure - were able to form a sufficientlystrong collective coalition across ideological, class and social divides to destabilise the political settle-ment under President Muburak. Conversely, comparatively peaceful states such as Ghana often havecomparatively inclusive settlements, which are exemplified by power-sharing mechanisms at the level ofthe state (in Ghana there is an informal convention that the president generally includes people fromthe major regions and ethnic groups in government) (DfID, 2010b: 14).19However, it is important not to make a straightforward link between the inclusivity of political settle-ments and the stability of states or countries.20 Formal power-sharing provisions have been a featureof the political settlement in Nigeria since 1979, when the ‘federal character’ principle became a featureof the constitution, requiring that appointments to public service institutions reflect fairly the linguistic,ethnic, religious and geographic diversity of the country (DfID, 2010b: 14). Despite this inclusivity, inrecent years the country has experienced outbreaks of civil unrest on a large scale. For example, inJanuary 2012, Nigeria was brought to a virtual standstill over strikes concerning government corrup-tion and the removal of fuel subsidies, which caused prices to rise from US$1.70 to at least US$3.50 agallon (The Economist, 2012). Thousands of protestors took to the streets and violence broke out inLagos and Kano - evidence that some aspects of the stability of the vertical dimension of the politicalsettlement were breaking down. There are two further case studies worth mentioning here, both of which provide a more carefulappraisal of the relationship between inclusivity and the stability and durability of settlements: theGlorious Revolution in late 17th Century England, and the transition to democracy in post-1958Venezuela. These examples suggest that, in some cases, the political exclusion of a particular groupin society can be a key factor in explaining the stability of a particular settlement, at least for a certainperiod of time (DfID 2010a: 14; Parks and Cole 2010: 23) Box 10. The ‘Revolutionary Settlement’ in 17th Century England The close of the 17th Century witnessed the ‘Glorious Revolution’ in England, when the Catholic Monarch James II was displaced by the Protestant Dutch Prince, William of Orange. When William III came to the throne, the constitutional details of the revolutionary settlement, codified in the Bill of Rights in 1689 and the Act of Settlement in 1701, stipulated that no Catholic could ever take the English throne. The exclusion of the possibility of a Catholic lineage to the throne contributed to the stability of the settle- ment in place under William, by providing reassurance to key elite figures in the Anglican Clergy and amongst the landowning elite that their interests would not be threatened by a Catholic monarch, as they had been under the reign of James II (Laws, 2010). that the power and influence of lower-level groups within the ruling coalition in Uganda has been increasing in strength in recent years, causing further fragmentation and disrupting governance (ibid.: 13).19 The IMF and World Bank have historically appointed their respective Presidents according to an unwritten agreement that the former will be European and the latter will be nominated by the United States. The recent appointment of the US-sponsored Jim Yong Kim to the head of the World Bank, over the Nigerian finance minister Ngozi Okonjo-Iweala, has stirred controversy, inviting comments from certain sectors of the media that the system of appointment is based on national affiliation rather than on merit (Rushe et al., 2012).20 For example, AusAID (2011: 32) assumes that “the political settlement will be strengthened if it is broadened to include a wide spectrum of society”. Later on, the authors offer a more nuanced position, stating that “Narrow political settlements between elites can play an effective role in maintaining stability and allowing for development in the short-run” (ibid.: 37).
19Box 11.Venezuela: Exclusive Pacts in the Transition to DemocracyIn 1958, in the aftermath of the collapse of the military regime under Marcos Pérez Jiménez and inpreparation for the transition to a democratic order, the three largest political parties in Venezuela - thecentre-left Democratic Action (AD), the centre-right Social Christian Party (COPEI), and the DemocraticRepublican Union (URD) - entered into a series of formal pacts and agreements to establish the politicaland economic institutions of the new democratic era. This included the Punto Fijo pact and the Declara-tion of Principles and Minimum Program of Government.The Declaration of Principles laid out a national economic development strategy based primarily onupholding property rights and on protecting foreign and domestic private capital accumulation. Theagreement featured a program of state compensation to landowners whose tenure would be threatenedby new agrarian reform policies, and subsidies for domestic industries. Both of these were to be paid forwith assets derived from the state’s oil wealth. Inclusive social development was to be pursued throughhousing programs for the poor, new labor regulations, trade union rights, the right to freedom of associa-tion and progressive policies on public health, education and social security (Karl, 1987: 85).The Punto Fijo pact established a power-sharing arrangement whereby the parties agreed to divide uppower and state resources broadly in proportion to the share of votes each received. Importantly forour purposes here, the Communist party (the PCV) was specifically excluded from the pact (althoughthe party was a signatory to the Worker-Owner Accord, an agreement concerning workers rights andcontract negotiation mechanisms).This act of exclusion is regarded as a key factor in accounting for the comparative stability of the pactsand coalitions that characterised the early years of democratic politics in Venezuela. The marginalisationof the far left provided reassurance to industrial and business elites, the Catholic Church and key elementsin the military that their respective interests would not be undermined by a revolutionary transforma-tion of Venezuelan society and politics (Kornblith and Levine, 1995: 46). In this way, the exclusion of theCommunist party helped, temporarily, to stabilise horizontal elite relations in the new political settlement.In terms of vertical elite-follower relations, the Punto Fijo settlement also appeared reasonably stable, solong as the country’s oil wealth continued to deliver the finances to support public services and progres-sive social development policies. During the 1970s, revenues derived from the boom in oil prices enabledthe state to meet the terms of the Declaration of Principles, and along with low domestic tax ratesthere was improvement in the provision of basic public services in health, housing, education and welfare(McCoy & Smith, 1995: 122; Leftwich, 2000: 185).However, in the 1980s Venezuela experienced a sharp economic downturn. The mismanagement of theeconomy by successive governments, a stark decline in oil revenues and rising levels of public debt, povertyand inequality put severe strain on the stability of the political settlement and the public legitimacy of thePunto Fijo regime.These problems were exacerbated by IMF austerity programmes, to which the countrybecame committed in the late 1980s. Despite the progress made in social welfare, social security benefitswere still the preserve of well-organised urban workers with affiliation to the AD-controlled WorkersConfederation of Venezuela (CTV). Those unorganised rural peasants who had migrated to the largecities between the 1950s and 1970s were mostly excluded from the welfare system, living in vast shantytowns and contributing to escalating levels of crime, poverty and social exclusion (Lapper, 2005). Thescale of public unrest was underscored in February 1989 when, in response to proposals for economicrestructuring that included cuts in petrol and diesel subsidies and a rise in public transport fares, massiveprotests and rioting in Caracas left several hundred dead.(Continued overleaf)
20 Box 11 Continued In the mid 1990s Hugo Chávez and his supporters entered the political arena, forming the Movement for a Fifth Republic and putting Chávez forward as a presidential candidate. Chávez accused the Punto Fijo regime as being economically inept, corrupt and exclusive of the lower stratum of Venezuelan society (Faria, 2008: 525). Chávez came to power in 1999 on the back of a campaign that promised to eliminate state corruption, introduce constitutional reform and provide a new deal for the poor (Lapper, 2005). His election victory marked the end of the Punto Fijo power-sharing arrangement. In the short-term, the transition to a democratic order in Venezuela was stabilised and consolidated through politically exclusive elite pacts and agreements. However, the settlement was both economically and politically unsustainable in the long-term, as made evident by the economic upheaval of the 1980s and the dramatic rise to power of Chávez in the 1990s. Had management of the economy been better, and had the governments pursued more consistent and inclusive welfare policies, the outcome may have been different. But it was not so, and the key point here is that the political dynamics of inclusion and exclusion, and how they relate to the short and long-term stability of political settlements, are both highly complex and context specific. It is therefore important not to assume that inclusive political settlements are necessarily any more or less stable than exclusive settlements.What seems to be of key importance is the relative political power of the elites that are included in thenegotiations, pacts and agreements that make up the settlement process, in comparison to the powerof those elites that are excluded (Hickey, 2011: 20; citing Khan, 2010).21 In some circumstances, it maybe helpful to substitute ‘inclusive’ for ‘inclusive enough’, as the goal for political settlements.221.13 ConclusionThis section has provided a survey of the definitions of political settlements from recent literature onthe politics of development. The following progress has been made towards the most useful way ofdefining political settlements:• Political settlements should not be defined solely as a contract between states and societies• olitical settlements are two-level games; they involve horizontal negotiations between contending P elites, and vertical relations between elites and their followers• olitical settlements should not necessarily be defined as a reflection of a common understanding P between elites; they are defined typically by bargaining and compromise• olitical settlements are not set in stone (i.e. they are never truly ‘settled’); they can adapt in re- P sponse to changes in power relations• olitical settlements influence the form, nature and performance of institutions P• olitical settlements can be more or less inclusive; but the stability of settlements is not necessarily P dependent on the degree to which they are inclusive.The following section builds on these insights to develop a robust definition of political settlements.21 Hickey (2011: 21-22) also makes the point that the meaning and significance of the goal of inclusion in the settlements framework varies depending on the type of inclusion that is being referred to. A settlement may exclude certain groups from political represen- tation – i.e. it may be ethnically or regionally exclusive - but nevertheless perform well in terms of economic growth, and therefore be inclusive in an economic sense. According to Parks and Cole (2010: 9) this has been the broad trajectory of Cambodia’s political settlement over the last decade. The highly complex and contextually specific relationship between practices of democratic inclusion in relation to the stability of political settlements in multi-ethnic countries is also illustrated in Mihailescu (2008). This paper looks at the transition to democracy in post-communist Slovakia and Romania. It argues that the inclusion of ethnic Hungarian minorities in the opposition political par- ties in those countries helped to contain the potential for violent conflict.22 World Bank (2011: xvii) makes a similar point in the context of discussing inclusive coalitions: “Coalitions are ‘inclusive enough’ when they involve the parties necessary to restore confidence and transform institutions and help create continued momentum for posi- tive change; and when there is local legitimacy for excluding some groups …”.
21 02 What is the most useful way to define political settlements?2.1 IntroductionIn this section, I argue that political settlements should be defined as on-going and adaptable politicalprocesses that include specific one-off events and agreements. The two stages of this argument areillustrated, firstly, by a comparison of taxation institutions and developmental reform in Costa Rica andGuatemala, and the experience of the United Nations state-building project in East Timor, and then,through the example of the revolutionary settlement in 17th Century England.2.2 Political settlements are on-going processes; they are not identical to institutions and/or one-off eventsThe long-term development or ‘process’ dimension of political settlements helps to explain why similarforms of political organisation and similar institutions can have starkly different outcomes in differentpolitical contexts. Dynamic, shifting and on-going power relations between elites, and between elitesand their followers – which underpin but cannot be reduced to specific institutions – are part of thereason why countries with similar institutions and similar endowments experience different develop-mental outcomes. This point is illustrated in Box 12 below, which looks at the way in which the politicalsettlements in Costa Rica and Guatemala influenced the performance of tax institutions and develop-mental reform. The point of this case study is to show that if political settlements are identified toonarrowly with specific, formal institutional details and one-off agreements, the framework will lose sightof the underlying power relations between elites, which determine partly the contrasting performanceof similar institutions in different contexts (Khan, 2010; Parks & Cole, 2010: 5-6).Box 13 looks at the experience of state-building in East Timor under the United Nations TransitionalAdministration project. The political settlements framework – with its focus on elite power relations,and relations between elites and their respective groups of followers – helps to explain why theformal state institutions erected under UNTAET were unable to secure long-term peace and stability.Again, the point of including this case study here is to illustrate that political settlements should not bedefined exclusively in terms of formal state institutions. To do so weakens the explanatory power ofthe framework, by not focusing attention on the underlying vertical and horizontal power relations thathelp to explain the developmental performance of different states.
22Box 12. Tax performance and the political settlement in Costa Rica and Guatemala (Di Johnand Putzel, 2009: 11-13; citing Yashar, 1997).After the 1948 independence revolution, Costa Rica consolidated a democratic regime, which was subse-quently characterised by political stability and relatively rapid economic growth. From very low levelsin the 1960s, both tax collection as a percentage of GDP and income tax collection rose steadily tothe point where Costa Rica achieved the highest tax take of any Central American country in 2005.Despite its very similar institutional, historical, economic and geographic context, Guatemala’s politicaland economic trajectory over the same period was starkly different. The country experienced muchslower economic growth, the rule of its military authoritarian regime persisted through the second halfof the 20th Century, and it was riven by political violence. According to Di John and Putzel (2009, 11-13,citing Yashar, 1997), what accounts for the pro-development trajectory of Costa Rica, in comparison toGuatemala, is the emergence of a political party that transformed the nature of the political settlementunderpinning the state.The Partido Social Démocratico (PSD) came to power in Costa Rica in 1951 by gaining political controlof the countryside. In addition to weakening the power of land-holding elites, the PSD undermined theoligarchic elite by nationalising the banking system and dismantling the army. By challenging traditionalelites in this way, the PSD created the political space in which to press for political and economic reform,including redistributive policies, land reform, and the creation of an inclusive welfare state financed bydrastic increases in tax takes and income tax.The different experience of Guatemala in this period starts with its military regime stamping down onpopular demands for democracy and social reform throughout the 1940s and 50s and introducing along-term ban on political parties and trade unions. Guatemala did not see the rise of any party, equiva-lent to the Costa Rican PSD, capable of challenging the power of its oligarchic elites and their hold onthe state and economy. This was due, first, to the fact that the Guatemalan economic elites were lessdiversified than in Costa Rica, with power centralised more in large landowners and less in financial andmerchant groups. Second, the military regime in Guatemala appropriated a number of large estates andlent its support to the practice of labour coercion on the part of the landowning elite. The landowningelite were co-opted into a political settlement that was not designed for social welfare provision butin maintaining the status quo. As the military became powerful landowners themselves, there was littleprospect of imposing higher taxation on this economic sector (Yashar, 1997: 4-7).Box 13. The political settlement in East Timor: illustrating the distinction between formalinstitutions and the political settlement (Ingram, 2012)In 1999, after 24 years of Indonesian occupation, the people of East Timor voted resoundingly in favour ofindependence. The election results were followed by a savage campaign of organised violence, disposses-sion and destruction on the part of pro-Indonesian militias, with the complicity and direct involvement oflocal Indonesian military and police. In the face of massive international condemnation, Indonesia withdrewall claims on East Timor and the United Nations Transitional Administration in East Timor (UNTAET) wasdeployed to stabilise and administer the territory, up to the point of formal independence.Ingram (2012) uses political settlements analysis to explain the failure of UNTAET to create the condi-tions for lasting peace and stability in the country. By focusing exclusively on achieving certain formalelements of the state and government, UNTAET did not pay due attention to the consequences of itsdecisions for the internal political settlement, with the result that widespread violence broke out again in2006.(continued overleaf)
23Box 13 ContinuedUNTAET was widely regarded as a success at the time of East Timor’s independence in 2002. Amongstits list of achievements, it could count the restoration of security and the containment of the militias, thereturn and re-integration of many Timorese who had been displaced across the border, and the buildingof a police and military force from scratch. The mission had assembled much of the formal apparatusof government and state from the ground up, including voter registration, a peacefully elected Constit-uent Assembly, a national civil service, a physical infrastructure for government, and the groundwork fornational economic management and public finance.Despite the considerable achievements of UNTAET, Ingram (2012: 2) argues that the project built “thewrong peace” by failing to engage with the underlying political settlement. The legacy of Indonesian occu-pation had created “corrosive” relationships between key elite figures and their respective followers.By not recognising and dealing with the unstable and volatile aspects of the underlying political settlement,the formal institutions and state and government structures put in place under UNTAET exacerbatedtensions between elites and between elite interests and the wider society. In other words, the politicalsettlement was both horizontally and vertically unstable.From the outset, few provisions were made to include the Timorese people in executive decision-makingor administration within the UN Transitional Administration. UNTAET ploughed ahead with the technicalexercise of constructing the apparatus of the state and government with little evident insight into themanner in which those institutions impacted on local interests and the character of elite relations. Further,UNTAET had little understanding of indigenous political organisation, and did not engage extensively atthe local, village level. Where it did engage, it did so in a haphazard, ad hoc fashion that for the most partfailed to facilitate constructive interaction (ibid.: 9).Understanding the local context was of paramount importance, as the years of occupation and exilehad created deep social, political, economic and linguistic cleavages. The protracted resistance strugglehad exposed and intensified fierce ideological divisions between former comrades in arms, and theunited resistance movement, which formed in 1998, masked serious differences and grievances across itscomponent political elites. Other divisions subsequently opened up. Those individuals who had left thecountry during the time of occupation and now returned to East Timor with wealth, education and thelanguage competencies required for the new administration were deeply resented by those who hadstayed behind and found themselves economically and professionally marginalised in the new politicalorder. The younger generation that had led the resistance through the decade of the nineties was alsoresentful of the impositions of the older generation of leaders (ibid.: 4-5).The process of constructing a Constitution and parliament further exacerbated the underlying tensions.A Constituent Assembly was formed to draft the Constitution, but the way in which the assembly waselected narrowed the basis of popular representation and therefore local legitimacy. The Assembly subse-quently voted to transform itself into the first parliament at independence. Horizontally, this createdconflicts of interest for the members of the Assembly, as it meant that they had to deliberate over acritical policy choice that directly impacted on their own interests and prospects. Vertically, it was prob-lematic as the citizens of East Timor had not elected this body to govern them into the future. Voters hadbeen denied an opportunity to hear the policy agenda of the parties, with the result that the governmentlacked democratic legitimacy from the outset (ibid.: 12-13). Moreover, the President, Xanana Gusmao,was a bitter political opponent of the Prime Minister, Mari Alkitiri. The ‘dual leadership’ system of theConstitution had the effect of creating two rival power centres, with the potential to destabilise the state.UNTAET faced considerable pressure from the Timorese leadership and the UN Security Council tocomplete its mission as quickly as possible. In the rush to construct the formal structure of an indepen-dent state and to finalise the Constitution, the mission achieved the core goals that it had been set butdid not create a stable platform for the state (ibid.: 16).
242.3 Political settlements should be defined partly by particular events, agreements and institutionsThe case studies discussed above make clear that political settlements should not be consideredidentical to formal institutions or one-off events. However, if - on the other hand - political settlementsare characterised exclusively as on-going political processes, and are not identified in part by theirembodiment in formal institutional details, agreements, pacts, events, and so on, then a risk is that theysimply become synonymous with ‘politics’ more generally. ‘Political settlement’ would simply be anotherword for all of “the pervasive and unavoidable (and necessary) activities of conflict, negotiation andcompromise involved wherever and whenever human beings in groups have to take decisions abouthow resources are to be used, produced and distributed” (Leftwich, 2011: 3). The concept would loseanalytical precision because it would fail to pick out and identify any specific feature(s) of politics.Second, if settlements are not embodied in or expressed through formal institutional arrangementsand particular events, then for practical purposes it would be very difficult to specify the point atwhich one settlement ends and another begins.23 If external aid agencies and donors want to use theconcept as a tool for analysing political arrangements in developing countries, then it would seem tobe important that settlements have a beginning and an end-point (even if those boundaries are more-or-less vague and contestable). For this reason, it is important to retain a place for institutional detailsand one-off events, pacts and agreements, within the definition of political settlements. These eventsafford some historical precision to political settlements analysis, as rough ‘markers’ to distinguish onesettlement from another. Box 14. Illustrating the need for one-off events in the settlements definition The revolutionary settlement in 17th Century England, referred to above, was the outcome of a long and complex process of parliamentary reform. It is a truism to point out that the social and political chain of cause and effect which led to the displacement of James II in 1689 could - in principle - extend as far back as recorded history. In order to use the political settlements framework to analyse and understand the Glorious Revolution with some precision, we need to identify key events as markers, which signify roughly the point at which the settlement under James II came to an end and was superseded by the ‘revolutionary settlement’ under William III. There are a number of critical moments in the revolutionary story which provide such markers; historians converge on the central importance of the birth of James’ male heir in June 1688. The prospect of a Catholic dynasty in England galvanised political opposition to James’ rule across the Anglican clergy, the commercial class, key stakeholders in the landowning elite, the military, and a broad spectrum of the general, non-elite population. By the following year, a coalition of elites, which became known as the Immortal Seven, had negotiated the invasion of England by William III, and a new political settlement had begun to unfold. The birth of James II’s male heir was an event, or a critical juncture, that occurred during the course of the political settlement in place during his period of rule, and which provides a rough marker for tracing the genesis of the revolutionary settlement that was to follow shortly thereafter (Laws 2010).23 As Brown and Gravingholt (2009: 6) point out, “The longer the timeframe adopted … the harder it is to pinpoint the nature and boundaries of a settlement and the less precise the notion becomes”.