Countries at Risk of Instability: Risk Factors and Dynamics of Instability The views expressed in...

59
Risk Factors 1 PMSU Background Paper FEBRUARY 2005 Countries at Risk of Instability: Risk Factors and Dynamics of Instability The views expressed in this paper are the responsibility of the Prime Minister’s Strategy Unit and do not necessarily reflect agreed UK policy Authors: Nick Donovan, Malcolm Smart, Magui Moreno-Torres, Jan Ole Kiso and George Zachariah PMSU Contact: [email protected] Executive summary 1. Stability emerges from the peaceful management of change and internal tensions. Countries that have limited internal capacity and resilience, and lack external stabilising factors are faced with a higher risk of instability. 2. There is no single definition of instability. Describing a country or a region as unstable suggests the presence of political, economic, or social upheaval. Instability can be manifested, inter alia, in coups d’Etat and other types of illegal or unpredictable political succession; breakdown of political, economic, and social institutions; systemic corruption; widespread organised crime; loss of territorial control; economic crisis; wide-scale public unrest; involuntary mass population displacement; and violent internal or international conflict. 3. Political transitions and rapid development can cause instability. The risk often arises from mismanagement or political manipulation, especially where domestic institutions are weak. Strong, authoritarian states that lack domestic and international legitimacy are also at risk of instability. If sources of discontent are suppressed, the risk of instability will likely increase over time. Furthermore, there are varying degrees of instability that can be manifested across and within countries and regions over time. 4. Increasing a country's capacity to manage effectively and adapt to change is at the centre of creating stability. It is also important to address structural factors, such as poverty, economic decline, natural resource dependence, and a bad regional neighbourhood, as well as external stabilisers, such as security guarantees and strong political associations. Instability is the dynamic outcome of various pressures, not a simple causal process. Prediction of the timing and nature of any particular crisis is not possible, as it will depend on a complex chain of events and interactions. It is possible to perform a robust assessment of the underlying risks of instability in

Transcript of Countries at Risk of Instability: Risk Factors and Dynamics of Instability The views expressed in...

Risk Factors

1

PMSU Background Paper FEBRUARY 2005

Countries at Risk of Instability: Risk Factors and Dynamics of Instability

The views expressed in this paper are the responsibility of the Prime Minister’s Strategy Unit and do not necessarily reflect agreed UK policy Authors: Nick Donovan, Malcolm Smart, Magui Moreno-Torres, Jan Ole Kiso and George Zachariah PMSU Contact: [email protected]

Executive summary 1. Stability emerges from the peaceful management of change and internal tensions. Countries that have limited internal capacity and resilience, and lack external stabilising factors are faced with a higher risk of instability. 2. There is no single definition of instability. Describing a country or a region as unstable suggests the presence of political, economic, or social upheaval. Instability can be manifested, inter alia, in coups d’Etat and other types of illegal or unpredictable political succession; breakdown of political, economic, and social institutions; systemic corruption; widespread organised crime; loss of territorial control; economic crisis; wide-scale public unrest; involuntary mass population displacement; and violent internal or international conflict. 3. Political transitions and rapid development can cause instability. The risk often arises from mismanagement or political manipulation, especially where domestic institutions are weak. Strong, authoritarian states that lack domestic and international legitimacy are also at risk of instability. If sources of discontent are suppressed, the risk of instability will likely increase over time. Furthermore, there are varying degrees of instability that can be manifested across and within countries and regions over time. 4. Increasing a country's capacity to manage effectively and adapt to change is at the centre of creating stability. It is also important to address structural factors, such as poverty, economic decline, natural resource dependence, and a bad regional neighbourhood, as well as external stabilisers, such as security guarantees and strong political associations.

• Instability is the dynamic outcome of various pressures, not a simple causal process. Prediction of the timing and nature of any particular crisis is not possible, as it will depend on a complex chain of events and interactions. It is possible to perform a robust assessment of the underlying risks of instability in

Risk Factors

2

a country or region. Close monitoring of these risk factors enables early response to the emergence of conflict.

• This project developed a framework, illustrated in the following diagram, to

assess countries and regions’ stability. It categorises the various forces and structures that determine the level of stability.

Figure 1.1 The Basic Instability Framework

Shocks

Feed-back: Violent Conflict, Political instability, Loss of Territorial Control, Economic Crisis

Risk factors for instability

External Internal

Country capacity and resilience

External stabilising

factors

5. This simple framework can be used to structure country- and region-specific analysis of the overall balance of stability, both at the present time and into the future, by projecting critical trends and developing scenarios. This paper provides some of the background analysis that led to the creation of a framework for examining factors that determine a country’s stability. It focuses particularly on the risk factors and the negative feedback loop. • Country capacity and resilience lie at the centre of country stability. This

determines the extent to which a country can successfully manage risk factors and potential shocks, as well as take advantage of external stabilisers. Country capacity depends on both state and non-state institutions, and should be assessed at an aggregate level and in relation to capability in managing specific risk factors like horizontal inequality.

• Risk factors for instability can arise from internal processes and factors within the country, or as consequences of the actions or inaction of other countries or the international community. Risk factors are generally ‘structural’ and need to be addressed through long-term policy measures and investment.

• Shocks comprise more proximate and unpredictable risk factors that can trigger instability at any moment in time. They impose huge pressures on country capacity and resilience when they occur, and necessitate contingency planning.

Risk Factors

3

• External stabilising factors provide support to country capacity and resilience. External stabilising factors can also set incentive frameworks that can foster stability in a country.

• The feedback loop of instability into the risk factors can - once crisis or conflict emerges - drive a vicious cycle of instability by further eroding country capacity and resilience.

6. In order to meet the challenges posed by countries at risk of instability it is necessary to understand what drives political instability, poor economic performance, and violent conflict. While each country is different, each conflict unique, and each economic crisis distinctive, researchers have uncovered enough common factors to justify setting out the evidence on the interrelated factors which cause and sustain violent conflict, economic decline and stagnation, poor governance and political instability.

Evidence What drives long term changes in levels of income? 7. Both longer term factors like trade and geography, and medium-term factors such as political and economic institutions (the ‘rules of the game’ such as constitutional regimes and the rule of law, and property rights) determine long-run economic success and failure. These political and economic institutions are shaped by the actions of elites. 8. Neither differences in culture and ethnicity, nor economic policies alone explain why some countries have very low levels of GDP. While good policies are important for growth, without strong political and economic institutions, long-term changes in income levels are not achievable. What drives political instability? 9. Full democracies are least likely to undergo significant political upheaval and instability. Countries undergoing political transitions face high risks of political instability and violent conflict. 10. Political institutions are important in managing a country’s risk of political instability. The constraints placed upon a ruling elite (for example, checks and balances provided by legislatures, bureaucracies, and judiciaries); and the degree of public participation in a polity are very important in determining political stability. 11. Bureaucracies are important political institutions that can help to preserve stability. By maintaining political neutrality in the formulation of policy options and the execution of political decisions, they can act as a check on the arbitrary exercise of power. A professional civil service can also add a level of continuity when governments change. By contrast, in a “neo-patrimonial” system, factions capture the state and use its organs for distributing patronage. A politicised bureaucracy can become just another link in a corrupt chain, and a source of sinecures.

Risk Factors

4

12. Economic development often precedes democracy but is not sufficient for it to take root. A study conducted looking at the period 1950-1990 showed that no democracy with a per capita income above US $6,000 (in 1985 prices) has had its democratic system collapse. However, economic growth neither leads necessarily to democratisation nor to improvements in other aspects of governance. 13. Part of the reason for this is the impact of natural resources. For example, petro-wealth can retard democratisation, by increasing the resources of autocracies, it reduces the need for ruling elites to be responsive to the needs of their citizens. What drives violent conflict? 14. Most violent conflicts are now intra-state rather than between states. This is an extreme form of political instability. 15. The most important factors associated with a high risk of intra-state violent conflict across many countries are: poverty (low levels of GDP per capita), fragile political institutions, and a recent history of violent conflict. Poverty may have a direct effect on the risk of conflict, but its main impact is likely to be indirect, reflecting the fact that poorer countries tend to have lower state capacity to mediate social and political conflicts and to respond effectively when they turn violent. Similarly, fragile political institutions are less able to cope with inter-group conflict and to restrain the actions of elites. 16. Other factors found to increase the risk of violent conflict, include regional conflict zones (bad neighbourhoods), the size of a country’s diaspora, the presence of natural resources, ethnic dominance and inter-group inequalities. The presence of oil is associated with the onset of war – particularly wars of secession. The presence of drugs and gems seem to prolong existing wars rather than trigger new wars. As is the case with diasporas, these natural resources probably play a role in financing ongoing conflict. 17. Again, the action of elites is thought critical in this process. Many commentators subscribe to a top-down understanding of conflict. Here elites draw on or manipulate two dimensions in order to further their interests: first, group identities, be they ethnic, religious, regional etc; and secondly, groups’ grievances (which may be related to poverty, especially one group’s position relative to other groups in that country). 18. About half of conflicts now break out in countries with a recent history of conflict. Civil conflict breeds its own conditions. In part, through a criminalisation of the economy, it weakens the institutions that can mediate social conflict. It can also strengthen factional identities based on grievance, which elites can exploit. 19. The length of civil wars increased dramatically from the 1950s to the early 1990s but has decreased markedly since the end of the Cold War. A large role has almost certainly been played by the notable increase in efforts at mediation, preventative

Risk Factors

5

diplomacy, and peacekeeping undertaken by the UN, regional organizations and powerful states. What drives economic decline and stagnation? 20. There is broad agreement among economists that economic growth requires the appropriate level and mix of “factor inputs” (labour, physical and natural capital and technology). However, these are only proximate drivers of growth. Good quality, political and economic institutions are needed to ensure the level and mix of inputs is appropriate and used productively. Institutions both set the incentives elites face and, ideally, constrain their behaviour so they act in the interest of the majority. 21. Structural constraints can directly hinder a country’s economic performance. Being landlocked, having bad disease environments and being situated on difficult terrain are poor starting conditions. Existing low levels of income can hinder a government’s ability to gather taxes and invest in infrastructure, education and healthcare. Lack of access to protected markets also hinders growth. Structural constraints also affect economic development through their interaction with institutions: trade openness improves the quality of institutions (bringing in new ideas, creating new elites, etc.). Poor geography and poverty can be partly overcome by good institutions as occurred in Botswana. 22. The presence of ‘point sourced’ natural resources (resources such as oil, minerals and even coffee and cocoa where patterns of production or distribution lead to concentrated revenue flows) contributes to many instances of economic stagnation or decline. Yet the presence of natural resources can also have a positive direct effect on growth, though in many countries this tends to be outweighed by:

• indirect resource misallocation effects, whereby foreign demand for the resource drives up a country’s exchange rate which crowds out domestic manufacturing and agriculture as exports become less competitive and imports cheaper. This can be exacerbated by attempts to protect domestic industries, which reduces trade openness. Investment is also reduced as the primary commodity dependent economy becomes more volatile. Investment in education drops when there is less need for an educated workforce, as occurs in natural resource-dependent economies;

• indirect political economy effects, where elites compete for control of the revenue flows from point-sourced natural resources (competition which sometimes can turn violent); entrepreneurial activity is directed towards rent-seeking rather than more productive activities; and, most important in the long-term, elites (deliberately or inadvertently) shape institutions to preserve their economic and political power rather than acting for the wider interests of the populace.

23. These mechanisms interact with each other, for example, in the presence of strong political institutions (such as Botswana or Norway) where policy decisions can be taken to mitigate the problem of resource mis-allocation.

Risk Factors

6

24. Economic stagnation or decline is often a consequence of political instability. Weak institutions allow elites to behave in ways that serve their narrow interests. Competition between elites for power often means that sound economic management is subjugated to short-term political interests. 25. External shocks can trigger this damaging political competition. The reverse is also true - spoils politics can lead to a hollowing out of the state, increasing its vulnerability to economic shocks. 26. In addition to politically-motivated economic policy ‘mistakes’, donors themselves have inadvertently proffered misleading advice by being insufficiently flexible over the forms of institutional reform they will support. Donors may also have induced instability through the timing of their support.

Common Themes The centrality of elites and institutions 27. A recurring theme of the evidence on economic stagnation and decline, violent conflict and political instability is the centrality of the inter-play between the behaviour of certain elite groups (elite groups are not monolithic) and the institutions that set the incentives governing this behaviour. In particular, at the core of many of the problems facing a large number of countries is the challenge of managing the inevitable competition for resources, wealth, and power by political and economic elites. While greed may play a role, it is not necessary to impute motives to groups of the elites. The ‘tyranny of the ruled’ may be at work. Elite groups may have to satisfy their vertical patronage networks.1 28. This competition for resources, wealth, and power takes place in all countries. With good institutions (rules of the game) in place, it can be a positive force, for example, if it ensures that those in power are either forced to innovate or surrender power to those with new ideas or more effective ways of keeping supporters content. Robust institutions help to regulate this process. For example, constraints on ruling elites through checks and balances, and autonomous bureaucracies, help to temper the effects of mistakes, greed and exclusionary politics (although the price of these constraints is sometimes stagnation). Similarly, property rights (in their widest sense – not just formal western notions of property) help to create a investment climate conducive to economic growth, or, at a minimum, provide a barrier against the behaviour of rapacious elites. 29. Unfortunately, in many countries, this competition is primarily destructive. Limited resources often mean that losers make absolute losses. Competition becomes conflict, albeit in most cases managed non-violently. 30. Often resources are misallocated in inefficient but politically-necessary projects. Such situations are characterised by periods of macro-economic instability. For example, the need to maintain patronage networks may mean that oil wealth may be spent by politicians on prestige projects rather than on investments that are more economically efficient.

Risk Factors

7

31. Absolute economic decline can occur when either this competition becomes excessively wasteful or an entrenched elite senses its grip slipping and seizes resources. At its worst, an elite group can hollow out the state in a quest for resources. In Zaire, Mobutu’s regime looted the state apparatus to a point where it collapsed economically, failed to provide basic services, and lost control over large swathes of territory.

32. When political institutions fail to provide a means of satisfying the needs of elites and their supporters by peaceful reform, this can lead to political instability, coups d’Etat, or violent civil conflict. The systematic discrimination in favour of certain groups in society (growing horizontal inequality) that can fuel economic stagnation may also create grievances.

The importance of natural resources 33. ‘Point sourced’ natural resources warp the political economies of countries. The evidence is that such resources often make governments distant from their citizens, usually fuel economic stagnation, sometimes fund ongoing civil wars and can lead to wasteful competition or violent conflict for the riches flowing from the mines, wells or plantations. Good institutions, such as those found in diamond rich Botswana, can regulate the actions of elites and allow resource wealth to be used wisely. The global number and intensity of conflicts can be, and has been, reduced 34. In the long run, a twin-track approach to economic development and the creation of robust, democratic political institutions is the best conflict prevention tool. 35. If the international community is faced with a choice as to where to allocate its scarce resources, the high risk of civil war in post-conflict countries justifies a focus on preventing conflicts recurring. While capacity is low in these situations, there is a greater opportunity to reshape institutions, and the international community may have more leverage and points of engagement. 36. In the short term, conflicts can be shortened and reduced in intensity through preventative diplomacy, mediation, peacekeeping, and cutting financial flows to protagonists.

Risk Factors

8

Contents 1. Introduction 2. Historical economic development of different regions 3. Factors related to the instability of political institutions 4. Factors that cause and sustain violent conflict 5. Factors specific to the creation and sustaining of poor economic performance Appendix 1: Case Studies

Risk Factors

9

1. Introduction 1.1. The purpose of this paper The aim of this paper is to explain how various risk factors affect a country’s stability, in order to be better able to identify and help prioritise policy responses.

1.2. Instability

There is no single definition of instability. Describing a country or a region as unstable suggests the presence of political, economic, or social upheaval. Instability can be manifested, inter alia, in coups d’Etat and other types of illegal or unpredictable political succession; breakdown of political, economic, and social institutions; systemic corruption; widespread organised crime; loss of territorial control; economic crisis; wide-scale public unrest; involuntary mass population displacement; and violent internal or international conflict.

Political transitions and rapid development can cause instability. These risks often arise from mismanagement or political manipulation, especially where domestic institutions are weak. Strong, authoritarian states that lack domestic and international legitimacy are also at risk of instability. If sources of discontent are suppressed, the risk of instability is likely to increase over time. Furthermore, there are varying degrees of instability that can emerge across countries and within countries and regions over time.

1.3. The approach of this paper

Each country is different, conflict unique, and economic crisis distinct. However, researchers have discovered enough common patterns for us to attempt to map out the inter-related factors that cause and sustain violent conflict, economic decline and stagnation, poor governance and political instability, and governmental loss of territorial control.

This paper highlights factors which, on average, across countries associated with political instability, violent conflict, and economic decline and stagnation. Where possible, this paper sets out whether there is a causal relationship, and by what mechanism one factor affects another. This paper has drawn on large-sample statistical studies, and the results of systematic multi-country case studies.2 This area of study is rapidly developing. A large amount of new scholarship is emerging on the links between violent conflict, economic development, regime type, state failure, and national reconstruction in the era of rapidly increasing globalisation. There is much research to be done and a great need for more work on the implications of new studies for policy. Policy makers also need more and better quality information and analysis tools for decision-making. This paper is intended then as a further step along the road to a better understanding of the issues, and not as a definitive solution to such complex problems.

Risk Factors

10

1.4 Methodology This paper has drawn on a number of quantitative, econometric studies. Such studies typically explain only 25-50% of the variation in the onset, duration or prevalence of conflict across countries and time. While useful, such statistical studies suffer from a number of limitations, including many factors which:

• cannot be quantified accurately or sensibly; • cannot be easily rated on a scale; • do not lend themselves to regression analysis;

Other constraints include: • incomplete or inaccurate data for many countries (particularly those in conflict

or lacking the resources to collect such data regularly and accurately); • widely differing interpretations of events in historical examples; • a lack of data from sub-national regions in countries that have large regional

disparities (e.g. in poverty, security, institutions, ethnicity, etc.). The benefit of large-sample statistical studies is that they are a key technique for making systematic comparisons of factors across countries and time. Case studies, while often richer for the qualitative information they provide, are often not directly comparable, because their authors use different methodologies. This paper has sought, as far as possible, to tease out the main factors from the large sample studies, to use evidence from case studies to explore more contested findings, and to help understand the mechanisms behind the statistical correlations. The table below shows the three most common factors which emerge from the five major statistical studies reviewed: the level of GDP (the higher the level of GDP, the lower the risk of contributing to conflict); the newness of the state and/or proximity to a major regime change; and the existence of previous civil conflicts. These conflict-focused studies looked at a wide range of factors, including ethnicity, linguistic diversity, religion, income, trade, inequality, geography, terrain, population size, governance, education and history of conflict.

Risk Factors

11

STUDY / RISK FACTOR Leve

l of G

DP

per c

apita

New

ness

of s

tate

\ Pr

oxim

ity to

regi

me

chan

ge

Prev

ious

civ

il w

ar

Fearon-Laitin (Onset of civil war) .

Collier-Hoeffler (Onset of civil war) .

Reynal-Querol (Prevalence of ethnic wars) .

Hegre et al (Prevalence of civil war) . .

Elbadawi-Sambanis (Prevalence of civil war) .

NS = Not Significant

= Increasing risk

= Decreasing risk

STUDY / RISK FACTOR Leve

l of G

DP

per c

apita

New

ness

of s

tate

\ Pr

oxim

ity to

regi

me

chan

ge

Prev

ious

civ

il w

ar

Fearon-Laitin (Onset of civil war) .

Collier-Hoeffler (Onset of civil war) .

Reynal-Querol (Prevalence of ethnic wars) .

Hegre et al (Prevalence of civil war) . .

Elbadawi-Sambanis (Prevalence of civil war) .

NS = Not Significant

= Increasing risk

= Decreasing risk However, this study fully accepts that many of the factors that cause and sustain conflict cannot be captured accurately in quantitative models, including:

- the fact that there is rarely, if ever, only one conflict going on. Rather, there are multiple layers of conflict, with some actors taking advantage of the outbreak of violence to settle other, often unrelated, disputes;

- belligerents should not be thought of as homogeneous in their aims or strategies. Intra-belligerent dynamics may have a major impact on the inter-belligerent conflict. Examples include the split between Uganda and Rwanda towards the DRC, and splits within rebel factions in the DRC that changed the course of the conflict.

These factors will fundamentally influence the dynamics of a conflict. Indeed, these dynamics may well come to dominate, with the initial causes of the conflict becoming less relevant to efforts to end a conflict or even prevent a resurgence. 1.5 The Instability Framework As will be seen in the following chapter, a multitude of factors have been identified and categorised as affecting the risks of instability. In order to help make sense of all these factors, this project developed the Instability Framework.

Risk Factors

12

Figure 1.1 The Basic Instability Framework

Shocks

Feed-back: Violent Conflict, Political instability, Loss of Territorial Control, Economic Crisis

Risk factors for instability

External Internal

Country capacity and resilience

External stabilising

factors

This paper examines a number of the risk factors for instability (see left arrow above), and some of the feedback mechanisms linking country capability to instability, be they direct or indirect. A better understanding of each of these factors also improves the understanding of the dynamics of country capacity and resilience and their common processes and elements.

Risk Factors

13

2. Historical economic development of different regions

2.1 Differences in development Understanding why some states are prosperous, peaceful and enjoy good and stable governance, and others are not, requires us to turn to history: • by and large economic success today reflects previous economic success: there is

a very high correlation (0.87) of high per capita income in 1960 with high per capita income in 1999;3

• two of the most important factors associated with violent conflict are a low level of GDP per capita and a history of conflict; 4

• the form of government most associated with good governance - democracy - is much more stable at higher levels of GDP per capita.

This section therefore examines the past to explain how countries arrived at their present situation. It examines the relative roles of geography, cultural and religious factors, trade, institutions, policies in explaining differences in GDP per capita. 2.2 Geographical inheritance As a rule of thumb countries in temperate zones are richer and more peaceful than those in the tropics. Is it countries’ geographic inheritance that determines their fate? There are several mechanisms by which geography can influence economic outcomes: first, geography can directly influence economic outcomes: • the tropical location of certain countries can directly influence the quality of land

and labour inputs into an economy through: - low agricultural productivity due to low fertility soils and a high prevalence of

crop pests and animal parasites; - a disease-rich environment (particularly malaria) which directly affects human

health and productivity;5 • a resource curse can develop with the presence of certain minerals and certain

plantation crops. A resource curse can affect economic outcomes both directly and by shaping and sustaining poor institutions.6

Second, geography can hinder trade and economic integration: for example, a landlocked location can restrict access to external markets and hinder export-led growth and technology transfers.7 In addition, trade may also affect a country’s political and economic institutions. For example, the absence of trading elites may have been important in shaping political institutions not conducive to economic development. 8

Risk Factors

14

Finally, a country’s institutions may have been shaped by its original geographical endowment.9 It is possible that countries with “point-sourced” natural resources – geographically concentrated resources such as minerals or certain plantation crops (sugar, cocoa, coffee) have, both today and in the past, thrown up narrow elites with an interest in maintaining their economic and political power.

Explaining terms Geography. Four aspects of geography are important: climate (e.g. temperature, precipitation), topography (e.g. mountainous, landlocked), geology (soils, minerals) and bio-geography (e.g. crops, animals, disease vectors).10 Geographic factors are commonly investigated in the research literature by looking at the presence of valuable minerals, mortality rates of European settlers and the viability of particular forms of agriculture. Policies. Policies include the government’s stance on key macro-economic variables such as taxation, public expenditure, exchange rates and trade. Cultural, ethnic or religious factors. Religion and culture has been investigated in the sociological literature following Weber to see whether particular aspects of a religion or culture (e.g. attitudes towards usury, entrenchment of social divisions) enable or hinder economic activity. Ethno-linguistic and religious fractionalisation variables measure the probability that a person meeting someone else randomly will be of the same group, while dominance variables measure whether a country is dominated by one particular ethnic or religious group. Culture has proved harder to measure systematically so sometimes researchers have used “natural experiments” such as the Cold War divisions of East and West Germany and North and South Korea to investigate the relative roles of religion, culture and ethnicity against other factors. Polity: This refers to the form of government of a social organization. Institutions. Examples of institutions include: the rule of law; property rights; and political arrangements such as democracy, presidential systems, or the presence of civil liberties. Institutions govern how people interact, for example: whether and how they decide what belongs to who; how they choose their leaders; how they settle disputes. As can be seen from the table below, institutions can be both formal and informal, as well as being economic, social, legal and political.11

Risk Factors

15

Level Examples Frequency ofchange

Effect

1. Institutionsrelating to thesocial structure

Mainly informalinstitutions such astraditions, socialnorms, customs.

Very long horizon– 100-1000 yearsbut also maychange in times ofcrisis / shock

Defines theway a societyconductsitself.

2. Basicinstitutionalenvironment:those related tothe “rules of thegame”

Mainly formal rulesdefining propertyrights, politicalsystems, financialinstitutions, and thejudiciary system

Long horizon 10-100 years

Defines theoverallinstitutionalenvironment

3. Institutions ofgovernance: thoserelated to the playof the game

Business contracts,corporategovernance

Mid-term horizon1-10 years

Leads to thebuilding oforganisations

In the literature the role of institutions has been investigated using a wide variety of proxies including: risk of expropriation, constraints on the executive, and perceptions of corruption. As can be seen, level 1 ‘institutions’ are not dissimilar to cultural and religious aspects of a society. There are also relationships between institutions and policies: it is helpful to think of policies as a “flow” concept and institutions as a “stock” concept. Institutions (particularly level 2 institutions) are the cumulative outcome of past policy actions.12

This hypothesis may apply both to countries with and without colonial histories: • between 1500-1900 territories rich in natural resources, but with poor disease

environments and large indigenous populations,13 attracted European settlers who set up short-term extractive economies, based on geographically-concentrated plantations or mines (such as oligarchic domination of agriculture, finance and business in many Latin American countries; slavery in Southern USA and the Caribbean, corvée (forced) labour in the Belgian Congo). This pattern of economic power created narrow rent-seeking European elites. These elites set up political and economic institutions that preserved their income and power to the detriment of the wider population. The influence of these institutions can persist into modern times; 14

• other areas such as North America and Australasia - had better disease environments and here settlers created more broadly based agricultural economies with more dispersed economic and political power (albeit at the cost of the indigenous populations). Settlers here set up institutions (such as stronger rule of law and more inclusive and constitutional political institutions) which they themselves could live under;

• extreme inequalities have emerged in countries (both with and without extensive experience of colonisation) with point-sourced resources. The narrow elites who are mobile during crises have less of a stake in the long-term development of the country. The lack of a substantial middle class reduces demands for investment in education and infrastructure, which, in turn, reduces social mobility and growth.15

Risk Factors

16

2.3 Trade patterns and institutions

‘Unlucky’geography

Lowincome level

Poorinstitutions

Low trade

Figure 2.1. The relationship between geography, trade, institutions and income Academic opinion is divided as to which transmission route (direct effect, institutions, trade) is most important for growth. Geographically related factors have both direct effects (for example, of malaria16) and indirect ones (affecting the development of institutions and trade).17 Schematically, this can be represented as in figure 2.1. Indeed, a vicious circle can develop. For example, disease reduces health, which lowers agricultural productivity, wealth, and impedes institutional development, which further restricts income, meaning that diseases will not be dealt with, and people get poorer, etc. 2.4 Ethnicity, religion, culture, and macroeconomic

policies Religion, ethnicity and cultural differences do not explain why economic growth in the 1960s and 1970s took off in Taiwan, Singapore and Hong Kong, but not in China until the 1980s. The economic literature does not find that religion, culture or ethnicity have a strong statistical association with long run economic performance.18

Similarly, while sound macro-economic policies are undoubtedly important – particularly in enabling economic growth – the role of macro-economic policies in the long run, level of income is much less important once researchers encompass the role of institutions.19 This is consistent with the facts that: • the set of policies associated with the “Washington consensus” have led to very

different experiences in Hong Kong (successful growth), post-Soviet Russia (collapse) and Latin America (mixed performance);

• sustained growth successes such as Taiwan, South Korea, China, or Mauritius, have all been achieved following different policy regimes (none of which conformed to the Washington consensus).

Risk Factors

17

• it seems probable that while good policies (or the removal of bad policies) are needed to allow or encourage the ignition of growth, sound institutions are necessary to sustain growth over longer periods of time.20

2.5 Breaking out of the vicious circle Despite the importance of long run trends in determining current patterns of peace and prosperity, countries can turn themselves around in relatively short time periods. There are large differences in patterns of political instability, violence and economic growth among ostensibly similar countries. For example, Sierra Leone and Botswana have become major diamond exporters. In Botswana the influx of such wealth is associated with stable, democratic government and a per capita income of $3,000.21 In Sierra Leone diamonds have been associated with dictatorship, civil war and one of the lowest per capita incomes in the world ($140).22 In South East Asia, Vietnam has experienced rapid growth since the end of the war in 1975 –leading to rapidly improving development outcomes, while Cambodia has experienced genocide, civil war and political instability.

Why did resource wealth contribute to such different outcomes in Sierra Leone and Botswana? Why has Vietnam outperformed Cambodia? The following chapters seek to explain the answers to these questions by setting out the evidence on the factors that cause and sustain political instability, poor economic performance and violent conflict.

Risk Factors

18

3. Factors related to the instability of political institutions

3.1 Instability of political institutions Fundamental challenges can arise that cause political institutions to become unstable and therefore subject to upheaval, disruptive changes, or disintegration. There are a number of factors that affect the resilience of political institutions, each of which will be dealt with in turn: political structure; economic development; natural resources; historical factors; social clivages and violence; and external factors. 3.2 Political structure Full democracies are the most stable type of regime.23 Countries undergoing transitions (e.g. former Communist states) and many other types of regimes face high risks of instability (autocracies, various types of hybrid systems).24 This paper does not focus on types of regimes or look at transition countries as a separate category, but looks at political institutions, both formal structures and rules-of-the-game, in a given polity, particularly: • the method of selecting ruling elites (e.g., elections, seizure by force, hereditary

succession); • the constraints placed on ruling elites (e.g. checks and balances provided

legislatures, judiciaries); • the degree of involvement of the public in the political process (e.g. through

political parties or factions); which is closely related to • the degree of inclusion of the polity (e.g. the degree to which minority groups are

represented); and • the neutrality and professionalism of the bureaucracy. Inconsistencies in these ‘rules of the game’ can generate instability. In democracies, these institutions are mutually-reinforcing. Through elections and party politics, many people are involved in selecting the ruling elite whose powers are restricted by the rule of law, autonomous bureaucracies and other organs of the state. It is this mutual reinforcement that gives democracies their stability: they are in stable equilibria and can withstand both exogenous shocks and the pressures caused by internal jostling for power. In autocracies participation is restricted to a small elite who also select the executive and remove constraints to their power – the bureaucracy is used as an instrument of power (through patronage and resource allocation). The internal inconsistency of these institutions makes semi-democracies less stable and resilient.25 3.2.1 Political institutions The effects of selection by ruling elites, constraints on the executive, and political participation, have been investigated by researchers.26 Naturally, is it is hard to separate the three factors, identify how they reinforce each other, and distinguish their

Risk Factors

19

effects. Some studies suggest that constraints on the executive and participation have a very strong and positive effect on a polity’s stability. If the executive is subordinated or controlled by other levels of government and if political competition is institutionalized and functional, this research finds that the relative odds of political instability are dramatically lower. Where there are few constraints on the executive and little participation, even in context of high growth, the relative odds of instability are large.27 In the USA, those who originally drafted the constitution accorded a prominent constitutional position to checks and balances.. The Supreme Court has the ability to veto executive decisions, funds cannot be raised without Congress, and the Senate and Congress are designed so as to require broad coalitions to pass legislation. Similarly, they were concerned to encourage political participation, calling for frequent elections (every two years for congress) and rigorously guarding the rights of assembly and free speech. 28 Various studies attempt to show the relative importance of political institutions and other factors in shaping the risks of state failure (political instability, revolutions and ethnic wars). They use an odds ratio to measure the risk of instability. A ratio close to one (1) indicates that the factor is not strongly associated with state failure, all other things being equal. Studies suggest that where there is a partially democratic system, with neither control on the executive power nor much political participation, the odds ratio is over 17 – roughly ten times larger than the odds ratios associated with socio-economic factors such as infant mortality or trade openness29 - this holds in Africa.30. In both Africa and the rest of the world strong partial democracies (those with either a degree of control over the executive or a high degree of participation) are much less prone to state failure. “If full democracy is to be achieved by means of gradual transition from autocracy, then it seems clear that the path of choice would be through the state of strong partial democracy.”31 Free elections for a President or Prime Minister are not enough – in fact, elections on their own can be dangerous – strong constraints on the ruling elite and/or regularized political participation is needed to create stability.

Lack of openness of executive recruitment

Unconstrained executive power

Fractionalised or limited political participation

Political instability

Risk Factors

20

3.2.2 Bureaucracies Politically neutral bureaucracies can act as a check on executive power through their role in the formulation of policy options and the execution of political decision. Professionalism is also key. Confidence in the state and legitimacy are supported by a civil service seen as neutral, with meritocratic recruitment and promotion, and a continuity of service. This permanence and continuity may be of particular importance for stability if governments change quickly or suddenly.32 By contrast, in “neo-patrimonial” states, factions capture the state. Civil service sinecures may be one of the spoils to be distributed. Such positions are also one of the organs for distributing patronage to clientelistic networks. Non-existent or inadequate pay also forces civil servants to turn to corruption. Such a system will mean the “winner-take-all” principle operates not only at the political level with changes in government, but also in the provision of public services and recruitment for public office. This can increase instability. 3.3 Economic development The durability of polities depends not only on their internal coherence, but also on the socio-economic conditions under which they operate. There are different scholarly views on the nature and the direction of the relationship between political instability and socio-economic development. Some authors submit that political instability disrupts productive activities and increases uncertainty, thereby undermining the incentives for the accumulation of physical capital with detrimental consequences for the rate of economic growth. Other economists argue that economic growth leads to more political instability because growth entails substantial structural changes that undo political coalitions and induce painful readjustments in the balance of power among different interest groups.33 Yet, others suggest that domestic political instability can force otherwise recalcitrant governments to undertake long-delayed reforms.34 The relationship between economic development and political stability depends on the institutional set-up. Przeworski, et al35 provide a panel analysis of data for 1950-1990 linking the likelihood of transition between democracy and autocracy (and vice versa) to the prevailing level of economic development. Using a binary coding of polities as either democracies or autocracies, their findings suggest that: • Autocracies are most unstable at intermediate levels of economic development,

but relatively stable at low and high-income levels. • Democracies are relatively brittle in poor economies, but become increasingly less

likely to collapse at progressively higher levels of economic development. Strikingly, no democracy with a per capita income above c.US$6,000 (in 1985 prices) has ever fallen.

This analysis provides support for Lipset’s original thesis that the more well-to-do a nation is, the better it will be able to sustain democracy and democratic institutions.36 Przeworski et al also report a relationship between levels of education and the stability

Risk Factors

21

of democratic polities. Their results suggest more educated populations are better at sustaining democracies, independent of the level of economic development. One potential explanation for the role of economic development on the stability of democracies is the way in which wealth alters the incentives facing elite groups. In particular, an abundance of wealth may lower the intensity of conflict over its distribution, relieving the pressures generated by those unsatisfied with the social out-turn under democracy. An alternative mechanism relates to the so-called “modernisation” argument that links development to the complexity of economic and social interrelations37. In sophisticated and wealthy economies, individuals are likely to acquire bundles of political affiliations that transcend a simple bipolar categorisation. As a result, support for particular elites is tempered by affiliations to other groups that appeal to other dimensions of interests. So the presence of cross-cutting political affiliations in sophisticated economies acts to lower the potential for any particular elite to successfully and abruptly alter the prevailing polity, by eroding the support base that any one group can command. Despite the plausibility of this argument, attempts at statistical validation have been mixed. Burkhart and Lewis-Beck38 undertook analysis of panel data on 131 countries, and report results suggesting that past economic performance is a strong predictor of current democracy. Other studies conclude against a strong modernisation hypothesis, given the existence of states that have retained autocratic polities despite sustained economic growth and relatively high levels of income per head. Even when rich autocracies do eventually become democratic, it is unclear whether the underlying cause can be justifiably attributed to modernisation. Thus, the impact of economic growth on political stability depends on other factors and, although economic growth can be a trigger for regime transition, there are important counterexamples. 3.4 Natural resources Analysis of natural resource endowments and political institutions has focused attention upon the effect of so-called “point sourced” resources on the stability of regimes, particularly at times of transitions. Results suggest that natural resource riches can contribute to instability.39 More autocratic states may reduce the threat of instability in the short-term through three mechanisms: • A rentier effect, through which states are able to raise revenue through the capture

of economic rents rather than general taxation. The state will then be able to provide a given level of public services at a perceived lower cost to its people. This in turn tends to relieve pressures for public participation in political affairs, as individuals accept minimal accountability as the price for low private taxes.

• A repression effect, through which the state captures economic rents with the explicit aim of funding the military and domestic security services. This alters the pattern of incentives for incumbent and challenging elites, making it easier for incumbents to resort to violent force to suppress their opponents. Knowing this to be the case, potential challengers for political power are deterred from acting.

Risk Factors

22

• A (reverse) modernisation effect, through which the extraction of natural resources distorts the accumulation of productive capital. The concentration of economic activity on labour-intensive extractive industries inhibits wider economic development into secondary and tertiary markets, and the complex mesh of skills and interactions that this entails. By virtue of resource-biased economic development, there are fewer complexities in the political needs of individuals, and consequently reduced imperatives for democratisation.

Resource-rich weak democracies and autocracies with some competition are more prone to insecurity. In regimes where political participation and control of executive are weak, perverse incentives may be created to skew the distribution of wealth through patronage networks. Additionally, incumbent governments have the advantages of knowing the availability of resource rents, and the discretionary power to distribute them. 3.5 Historical Factors A country’s political and social history has important implications for its political stability. The pattern of past polities, and historical transitions from one polity type to another, both matter. Polities survive by acquiring legitimacy. Building a record of effectiveness can create legitimacy, thereby consolidating the stability of a polity.40 Moreover, the age of institutions can have a legitimising effect; longevity has often been linked with stability and political performance. History can also alter the expectations of individuals concerning the behaviour of elites and their own rights and responsibilities as citizens. A young autocracy in which the population had previously lived under democracy is likely to be significantly less stable than a state in which autocracy has prevailed for many decades. The source of this instability lies in the discontent the public may feel as their participation in political affairs is curtailed from that which they have become accustomed. Transitions from military rule, be it to more or less democratic forms of civilian government, raise two specific sets of challenges: the first is that a resurge of societal violence, and the failure of a civilian government to keep adversarial politics within an orderly framework, will increase the likelihood that the military will not acquiesce to civilian rule; and the second is the difficulty of managing military expectations and clarifying its role within a polity, particularly vis-à-vis the selection and functioning of the executive.41 Similarly, political systems moving from historical autocracy to democracy may feel unsettled by the mobilisation of contending elites. 42 This can translate into higher degrees of political instability as new actors try to use the system to have their demands met. In fact, several authors have found evidence that political instability peaks around the second election after the opening up of political participation. 43 This may lead groups either to use violence or otherwise circumvent the formal political processes, if they feel they have lost out.

Risk Factors

23

Jackmann argues that the optimism underpinning decolonisation after World War II neglected the institutional youth of the “new” states and assumed that self-government and democratic government were identical. This failed to take into account the mismatch between pre-existing polities and the states superimposed on them. Looking at the relationship between state legitimacy and development in Africa, Englebert also argues that the colonial genesis of many states has tended to create structures that conflict with pre-existing political institutions, underlying norms of political behaviour, and customary sources of political authority.44 Elites accommodate or domesticate the competing alternative loyalties and allegiances, thus resorting to clientelistic networks to prevent state fragmentation. These new states, mostly (though not exclusively) in Africa, are not the endogenous creations of local history. They are not embedded in the domestic power relations and therefore lack legitimacy and capacity, and are prone to various forms of political instability. 3.6 Societal groups, integration, and violence Fearon and Laitin45 found that political instability in the 3 years previous to the year of observation doubles the risk of civil war. Elbadawi and Sambanis46 concluded that recent political instability increases the risk of civil war in many models, but that the finding is sensitive to the choice of lag structure for the political system variable. Changes in weak political institutions of a country are likely to be accompanied by a heightened risk of civil strife or instability.47 Goldstone finds that democracies, if weak in regard to either executive constraints or the regulation of political competition, can fail to withstand the forces that lead to internal wars, particularly of an ethnic nature.48 First, changes in a democratic direction are likely to be accompanied with reduced repression, which allows communal groups to increase their opportunities for mobilisation. At the same time, it takes time to establish the new institutions and to make them sufficiently efficient to accommodate views and mediate conflicts. Moreover, groups that increase their political influence will raise their expectations for real improvements in their living conditions, but these changes take a long time to realise even with the best intentions. Finally, changes, by definition, alter the power distribution, which means some people gain and others lose. Hence there is a need to look at the incentives and institutions that accommodate and isolate both sides.

Mismatch between state and society

Expectations

History of effective political institutions

Political instability

Risk Factors

24

The notion of discrimination helps explain the relationship between conflict and political instability. Exclusionary politics motivates discriminated groups to resist, which may in turn motivate elites to further repress. According to the State Failure Task Force, two indicators are key in this regard: the ethnicity of the elite in a heterogeneous society; and the existence of public policies that discriminate certain groups. Other authors have examined the role of horizontal inequality (differences between social groups) as a major source of conflict and political instability.49 Reynal-Querol develops a theoretical model that captures the links between political systems and rebellion. She argues that the more inclusive the system, the smaller the probability of instability, since not all governments represent their citizens in the same way, even if they have high level of political rights and civil liberties.50 Additionally, when social divisions run deep and the institutions of conflict management are weak, exogenous shocks, such as deteriorations in the terms of trade, have high political costs, because they magnify distributional conflicts.51 3.7 External factors States do not operate in isolation and external factors, particularly the presence of violent conflict in neighbouring states, have an influence in the onset and sustainability of political instability in a country. Researchers using the Polity IV dataset found that violence in two or more neighbouring countries was always associated with a substantially higher risk of failure in a given country. The political neighbourhood may be important. The model run by Gates, and based on the Polity IV data, indicates that polities surrounded by states diametrically different (in terms of the Polity regime typology) are likely to expire twice as quickly as identical polities geographically located among comparable polities.52 According to this study, the political neighbourhood is particularly important for the stability of polities with open and competitive executive recruitment. Sudden changes in a country’s relationship with the external environment may act as triggers of instability if they fundamentally challenge the institutional consistency of the political system.

Large-scale internal conflict

Political discrimination

Political instability

New or challenged political institutions

Low trade openness

Political instability

Violence in neighbouring countries

Risk Factors

25

Low openness to trade may encourage political instability. The State Failure Task Force Report III finds low trade openness to be a highly sensitive indicator of the costs imposed on a country by the inability or the unwillingness of elites and authorities to cooperate in upholding a fair economic regulatory framework. As well as inhibiting trade, this may encourage political instability through rent-seeking and distribution conflicts. Summary As has been shown, there are many factors that affect the stability of political institutions. Political stability is one of the key components of overall country stability. Extreme political instability can lead to violence, the subject of the next section.

Risk Factors

26

4. Factors which cause and sustain violent conflict 4.1 The overall incidence of conflict Almost all violent conflicts are now intra-state. On average two or three civil wars start in each year – this rate of ‘ignition’ has been pretty constant over the last fifty years. Currently, in an average year about two conflicts end. This ‘termination rate’ has varied quite dramatically over time as the average duration of a civil war has waxed from the 1950s and waned since the early 1990s, thereafter declining.53 The interaction of these trends has meant that the overall incidence of conflict has declined since the 1990s – driven down largely by the decrease in the duration of ongoing wars. Figure 4.1: Proportion of countries in civil conflict 1950-200154

4.2 Factors associated with civil wars Researchers comparing civil wars since the 1950s using statistical methods have found that the factors most consistently and strongly associated with the start of civil wars are: • poverty – a low level of GDP per capita; • a recent history of conflict (which in turn proxies a range of different factors); • fragile political institutions. These findings are found in almost every cross-country statistical analysis of civil war (see sections 1.4 and 4.5).55 In addition, some studies suggest the importance of other factors in the ignition or duration of conflicts:56 • the presence of natural resources; • bad neighbourhoods or “regional conflict formations”57 • the size of the country’s diaspora;

Risk Factors

27

• the extent to which the country is dominated by one ethnic group • a country’s trade openness. • inequality between groups, as both a cause and sustainer of conflict; • mediation and preventative diplomacy, in preventing and shortening conflict; and • the proliferation of arms 4.3 Poverty is strongly associated with civil war One of the strongest findings, which is found in almost all research on civil war, is that low levels of GDP per capita are associated with a high risk of violent conflict - both that conflict will erupt and that it will be prolonged once it starts. Figure 4.2 below shows how, other factors (such as natural resources, diasporas, previous civil wars, type of political institutions) constant, a country at or around US$250 GDP per capita has, on average, a 15% risk of seeing a civil war in the following five years. Once a country reaches a GDP per capita of around $5,000 the risk of civil war, on average, reduces to below 1%. Figure 4.2 Risk of civil war and GDP per capita, 1960-199958

0%2%4%6%8%

10%12%14%16%18%

$0 $1,000 $2,000 $3,000 $4,000 $5,000US$ GDP per capita

Pro

babi

lity

of o

bser

ving

a n

ew

conf

lict i

n ne

xt 5

yea

rs

This poverty-conflict link can also be shown by plotting the number of conflict-affected countries per income group (figure 4.3). Not only does this show that less developed countries are more prone to conflict, but that this correlation has intensified since the mid-1990s. The implication from all this is that, in the long run, economic development is one of the best conflict prevention strategies.

Risk Factors

28

Figure 4.3 Conflict-Affected Countries by Income Group, 1992-200159

Are low levels of GDP causally associated with conflict? There could be unmeasured factors (e.g. social capital) that affect both poverty and conflict, which would mean there is not a causal relationship. If there is a causal relationship, as seems likely, the influences probably run both ways: conflict reduces levels of GDP per capita, and low GDP may lead to conflicts60. Low levels of GDP probably cause or sustain conflict through the following two mechanisms: • State weakness. Low levels of GDP per capita are associated with low state

capacity; in particular, weak capacity to settle societal conflicts via political and judicial means rather than violence. There are several sub-mechanisms:

- the governments of poorer countries typically have lower state revenues with which to pacify potential opponents;61

- the political institutions of poorer countries may be under-funded and less effective. There may be, on paper, fair and effective ‘rules of the game’ but, for example, effective rule of law may be undermined by underpaid judges, lawyers and policemen;

- low state capacity could imply a lack of military might. A government unable to carry out effective counter-insurgency campaigns would pose less of a deterrent to potential rebels, meaning that more such conflicts would be expected in poorer countries, and linked to this;

- richer countries are better able to meet the needs of the population, which reduces the potential causes of grievances;

- wealthier countries tend to have better infrastructure (particularly in more remote areas) and a greater government reach into distant regions, both of which are likely to make insurgency more difficult. Greater wealth may mean fewer popular grievances. This argument is supported by findings that countries with mountainous regions are more prone to conflict62.

Risk Factors

29

• Economic opportunity costs for potential rebels. Low levels of GDP reflect a lack of economic opportunities for potential rebels – who have few stakes in society (employment, investments) to deter them from using violence.

Figure 4.4 The Poverty-Conflict mechanism

Low opportunity

cost for young men

Easier/cheaper to

recruit/retain

‘soldiers’

StateWeakness

StateWeakness

Weak statemediation

mechanism/institutions

Elite Behaviour

‘Hollowed-out’state

IncreasedRisk ofViolentConflict

IncreasedRisk ofViolentConflict

Crush Oppositionmilitarily

Buy off Opposition

Low State Revenueswith which to

‘Poverty’low percapita

incomes

‘Poverty’low percapita

incomes

Is proxy

for

Military is insufficient

deterrent to rebels

Weak Government

outreachinto Rural

Areas

It should be stressed that, other than the opportunity cost mechanism, studies do not posit a direct causal relationship running from poverty to conflict. That is, that widespread systematic conflict is not a result of the poor rising up and rebelling against their lot (though more contained conflict among poor people over scarce resources, such as land and water, does occur).63 Instead, most models of conflict point towards a top-down process, with elites drawing on/manipulating poor peoples’ very real grievances (which may be related to poverty, especially relative poverty) in order to further the elite’s narrower interests. This is explored in more detail in the section on ethnicity below.

4.4 Previous history of violent conflict Roughly half of all outbreaks of violent conflict occur in countries that have recently had a violent conflict. Indeed, countries emerging from conflict run about a 30% risk that they will return to conflict within 5 years.64 Almost all the major statistical studies of conflict find conflict legacy factors - proxied in the large-N models by a recent history of conflict - to be important in explaining the risk of future conflict. There are probably several mechanisms at work which can explain this finding. Previous conflicts affect the risk of future conflict by weakening institutions and increasing poverty, and by driving out manufacturing, thereby increasing a country’s dependence both on natural resources and informal or illegal activities. They can also

Risk Factors

30

worsen inter-group inequality and heighten groups’ sense of identity, and can exacerbate grievances owing to wartime atrocities or the level of civilian deaths.

4.5 Fragile Political Institutions A common finding is that robust, established, and internally-consistent political institutions reduce the risk of violent conflict. This manifests itself in a variety of ways in the research literature: • some research finds that the newness of a state or proximity to independence is a

risk factor associated with civil war;65 • some research finds that the frequency of regime change is associated with violent

conflict;66 The mechanisms by which fragile political institutions heighten the risk of civil war are explored in more detail in section 2.67

4.6 Natural resources A number of statistical studies of conflict highlight the importance of a country’s dependence on a narrow range of natural resources.68 Countries that earn either a small or a large share of their national income from natural resources were found to be at less risk of conflict than countries where natural resources contributed between one-quarter and one-third of their income. Some claim this correlation is spurious, because the variable used is a poor.69 Other econometric studies, while refuting the significance of natural resources per se, have found that conflict is more prevalent among oil producers.70

It is clear from a number of country case studies that natural resources can, though need not necessarily, play a role in violent conflict.71 Obvious cases where resources have played a role include DRC, Sierra Leone, Angola, Sudan (of late, though not necessarily at the outset of the conflict between the Government and SPLA) and Colombia. However, in many conflicts, natural resources do not play a role, e.g. Nepal, Senegal, and Lebanon72. Financing and financial gain are still important. Soldiers and rebels resort to looting to help fund themselves.73

Clearly, resources can play a role. More recent country-focused research has tended to focus on the circumstances under which natural resources can increase the risk or duration of conflict. This work has highlighted the importance of point-sourced commodities over diffuse ones.

Point-sourced resources are those whose revenues are narrowly funnelled, and therefore more easily captured. Not all mineral resources need be point-sourced. Diamonds which require deep shaft mines are point-sourced, but alluvial diamonds tend not to be. Similarly, not all point-sourced resources are minerals. Depending on how they are marketed, the revenues from some agricultural products (coffee, cocoa) can also be funnelled, as can aid, as well as stocks that can be captured or controlled by an elite (land or arms). One study of 13 conflicts found that oil, non-fuel minerals and illicit drugs to be the main natural resources fuelling conflict.74

These recent studies have also thrown light on the mechanisms.

Risk Factors

31

Onset of Conflict:

• As described in sections 4.3 (and again in section 5.2.2.b), the existence of resources inspires negative behaviour by certain elite groups, which cannot be constrained by weak institutions. This further weakens the state and leads to conflict through the mechanisms described under poverty in section 4.3;

• Resources (particularly oil) concentrated in one part of a country can incite that region to secede (e.g. Biafra in Nigeria, Aceh in Indonesia);

• Rebels can fund start-up costs by selling the rights (booty futures) to exploit resources in areas they hope to capture (e.g Kabila during his take-over of what was then Zaire);

• A government may resort to pre-emptive repression to protect resources, exacerbating grievances and creating early casualties (Sudan and Indonesia in Aceh);

• The prospect of earning a share of another country’s resources may feature in a foreign country’s decision to intervene in support of a rebel group (e.g. Liberian activity in Sierra Leone).

Despite the number of onset mechanisms, there is evidence from case studies that, except for oil, natural resources are less important in the onset than in determining the duration of conflict.75

Duration of conflict

• Sale of war booties or revenues from those already sold can prolong the conflict (Angola, Sierra Leone and the 1998-2000 conflict in the DRC);

• Looting of resources can finance rebel operations (including arms purchases). If the rebel is the stronger party, this can act to shorten the war (and prolong it if it is the weaker party);

• If exploitation appears financially profitable, resources may prolong a conflict by reducing the incentives for peace.

• Finally, the mechanisms may operate in the reverse direction. Resources could shorten conflicts or reduce their intensity, either because access to lootable resources could undermine coherence within rebel organisations (as some leaders may prefer to focus more on looting than fighting thereby cutting battle deaths), or if belligerents chose to collaborate locally in exploiting the resources;

Different types of natural resources affect conflict in different ways and through different mechanisms. Broadly the evidence is that: • Oil is associated with the ignition of new conflicts – in particular separatist

conflicts; • Gems, opium, coca and cannabis help to increase the duration of ongoing

conflicts; • The role of timber has not yet been systematically tested; and • There is a lack of statistical correlations between conflict and other primary

commodities, such as agricultural produce and non-fuel minerals. The dynamics of conflicts associated with natural resources can also have an important impact on the risk of future conflict. Conflict changes the nature of the

Risk Factors

32

economy, as formal sector activities (especially manufacturing) are driven out or close down. Modern intra-state conflicts are associated with the growth of informal and often illegal activities. Rebels may be better networked into the global economy via links with organised crime than formal businesses based in the capital city. This growing informalisation and links with international crime creates a new economic elite with few interests in establishing positive institutions. In this regard, past conflicts can pre-dispose countries to further conflicts. Figure 4.6 Weak institutions as a driver of instability

Age of institutions

Limited pol. participation

Political discrimination

Economic decline

Limited electoral processes

Natural resources

Large-scale internal conflict

Weak

institutions

Political Instability

No executive constraints

4.7 Ethnic Dominance

There is mixed support for the argument that ethnicity causes or sustains conflict.76 A number of statistical studies have found, other risk factors being equal, that having two large and roughly equal ethnic groups (studies have typically used language-based or religious proxies for the degree of ethnicity) makes a country more at risk of conflict than when there is a greater degree of ethnic diversity. An alternative way of putting this is that extremely heterogeneous or homogeneous societies tend to be both at lower risk of conflict breaking out and, once violence has started, tend to have shorter conflicts.

Given that the nature of ethnic variation is very country-specific, there are clearly limits to how well this can be modelled using cross-country econometric methods. However, the broad findings reported above tend to be borne out by case studies.77 Ethnic diversity is thought to constrain widespread conflict in Nigeria while group dominance heightened tensions in Northern Ireland, Cote d’Ivoire and Macedonia.

Risk Factors

33

A number of theories of how this occurs have been proposed.78 These try to explain why more polarised societies are at greater risk of conflict. A unifying theme is that making groups act violently is difficult. Groups with a stronger sense of identity are most likely to act collectively.

• In homogeneous societies (where most people are from the same ethnic or religious group) there is a greater chance that opposing leaders will cause a movement to divide;

• In very diverse societies, a rebel group is more likely to have to draw people from different ethnic groups if it is to grow to a size capable of threatening the government. Maintaining unity will be harder in a movement comprising different ethnic groups. Again, the government may divide the movements, playing off different groups and interests against each other, in order to build its own coalition.

• In more polarised societies a large minority may fear political domination and permanent economic, political and cultural exclusion. A dominant group or groups that discriminate against other groups make it more likely that those discriminated against will act. When there is elite manipulation of identity, this may not be sufficient to motivate a group to act violently, in the absence of a real fear of violence.79

A number of these explanations involve an inter-action between ethnic groups and inequality or discrimination, to which we now turn.

4.8 Inequality

Econometric studies do not provide consistent evidence of the significance of inequality between the richest and poorest in society (vertical inequality), though contradictions between models may be explained by variations in how the models are specified.80 Some work suggests that, while vertical inequality may not be significant in causing conflict, it can prolong it.81

Evidence from case studies is that inter-group inequalities can increase the risk that a country will experience violent conflict. Leaders who have sought to use ethnicity or other identity characteristics to mobilise a group against another have been found more likely to succeed when there are substantial inequalities among groups that cause resentment.82 These inter-group inequalities have been termed ‘horizontal inequalities’.

Despite the obvious difficulties of data problems associated with any general work on horizontal inequality, there have been a couple of attempts to test this statistically. There is evidence that, neither a country being divided religiously nor having a government pursuing policies of state religiosity, was in itself enough to heighten the risk of conflict. Rather, it is the combination of these two that best predicts religious intra-state conflict.83 Horizontal (including spatial) inequality has been found to help explain the variation in the intensity of conflict in Nepal.84 Caste features strongly among these Nepali horizontal inequalities, including access to civil service jobs. Statistical evidence thus tends to support the idea that systematic exclusion can be a cause of conflict.

Risk Factors

34

This involves the interaction of sharp horizontal inequality with other factors, notably group perceptions and identity. The grievances associated with the former help overcome the collective action problems associated with mobilising people around the latter. Clearly, such differences in the allocation or economic and political power between groups may persist for some time without conflict breaking out. Horizontal inequality is thus likely to be a structural factor predisposing a country to conflict, the trigger being one or more proximate factors.

Figure 4.7 Horizontal Inequalities links to conflict

Strong ‘Group’ Identity

TriggeringFactor

OutbreakViolentConflict

OutbreakViolentConflict

Real or Perceived

GroupGrievance

MarkedInter-GroupInequality

SystematicDiscriminationagainst Group

+HorizontalInequality

4.9 Diasporas

Some econometric studies show a correlation between large diasporas and the risk that a country will experience violent conflict.85 The evidence from systematic country case studies is ambiguous. If the definition of ‘diaspora’ is broadened to include people from the same ethnic group living in neighbouring countries, then evidence from case studies is strengthened to support a link.86

Two key mechanisms have been proposed:

• Splinter groups of diasporas may provide funding for rebel groups, making conflict financially feasible;

• Having ethnic kin living in a neighbouring country may increase the risk of conflict if these migrant communities offer support to their kin (e.g. Albanians’ support for Kosovo) or if it nurtures nationalisms, encouraging secession;

However, the mechanism may be in reverse, with conflict-affected countries having large diasporas, because conflict (or fear of it) forces people to leave. Similarly, diasporas may be as likely to support the government as the rebels.87 Finally, overall, diasporas have a positive effect: in 2001 diasporas gave $72 billion to developing countries in remittances, in comparison to the $52 billion given in international development assistance.88

4.10 Bad Neighbourhoods There is evidence to support that violence in neighbouring states can be a destabilising force, leading to conflict. “The conflicts in Central Asia, Western, Central, and the Horn of Africa all share characteristics of ‘Regional Conflict Formations’”.89

Risk Factors

35

Violence is another transnational relation (related to transnational flows like arms, refugees, drugs, resources, etc. and networks like armed groups, NGO’s, criminal networks, splinter diaspora groups, terrorists, etc.) that presents new challenges for the state. The incapacity of a state to deal with it can lead to the spread of conflict across an entire region. Such regional conflicts “are comprised of interrelated civil wars, operating at different levels (local, provincial, national, and regional levels) without respect for frontiers.”90 Cross-border identity groups, hostile neighbours, and informal economies make the spread of other violence more likely when conflict breaks out.

4.11 Preventative diplomacy and peacekeeping Explaining the global decline in the duration and intensity of violent conflicts since the early 1990s is difficult. However, one factor is almost certainly the dramatic increase in efforts by the UN, World Bank, regional security organisations, states, and NGOs to prevent conflicts from breaking out and to dampen or settle ongoing conflicts.

• UN peacekeeping operations (in post-conflict situations) have increased from an average of 4 per year from 1948 to 1988 to around 15 per year from the 1990s onwards.91 Three new missions were launched in 2004;92

• this increase has been accompanied by an increased willingness by the UN Security Council to authorise the use of force in such peacekeeping operations. Resolutions under Chapter VII have increased rapidly since the early 1990s;

• these peacekeeping operations have become more complex, and almost all new operations are accompanied by Disarmament, Demobilization and Re-integration programmes. The number of DDR programs have increased from zero in the 1980s to between 6 and 12 per annum from the early 1990s;93

• there have been between 1 and 6 formal UN preventative diplomacy missions (where the UN Secretary General sends personal envoys to crisis zones before a conflict breaks into violence (with the consent of the member state)) each year since 1990;94

• more common are UN initiatives to make peace, once conflict has already erupted into violence. UN peacemaking missions have increased from 3 to 5 per year in the early 1990s to around 15 per annum since 2000.95

Efforts at peace diplomacy by other regional organisations, key states and NGOs have also increased. For example, OSCE conflict prevention and peace-building missions have also increased – from around 5 per year in the early 1990s to between 15 and 20 per year since 1999.96 Figure 4.9 shows the rise in all mediation efforts since World War Two. Researchers estimate that, on average, mediation achieves some positive effect in almost 40% of cases, but it is hard to have an impact on the incidence and duration of violent conflict.97

Risk Factors

36

Figure 4.9 Mediation and other conflict-management efforts 1945-1995

0200400600800

100012001400160018002000

1945-55 1956-65 1966-75 1976-85 1986-95

Multilateral Conferences

Referrals to InternationalOrganizationsArbitration

Negotiation

Mediation

Summary Most violent conflicts are now intra-state rather than between states. They can be seen as an extreme form of political instability: where all effort to mediate a change of regime has collapsed. As has been seen, economic development is strongly related to both political instability and violent conflict. Economic decline is a major part of national instability. The next section examines this in more detail.

Risk Factors

37

5. Factors specific to the creation and sustaining of poor economic performance Poor economic performance is taken to mean stagnation in per capita income or, less commonly, a prolonged absolute decline in real per capita GDP. Some studies on growth have instead used the more specific ‘output per worker’. In the following, the two measures are judged close enough for purposes of broad generalisation98.

Few countries have sustained either continuous growth in per capita incomes (with exceptions in East Asia) or continuous decline in aggregate national income over periods of 10 years or more. Most countries have experienced shorter periods of both growth and stagnation. Although absolute decline in a country’s total national income is rare, many countries’ economic growth has failed to keep pace with population growth, with the result that per capita incomes have declined. Table 5.1 presents 42 countries which have experienced growth failures.

Table 5.1 Economic decline and stagnation Countries whose per capita incomes have declined to the extent that, in 1998, they were no higher in real terms than…… …in 1960 or before …during the 1960s …during the 1970s …during the 1980s Central African Republic Côte d’Ivoire Burundi Kenya Chad Mauritania Cameroon Republic of Congo Congo, Dem. Rep. Togo Gabon Ecuador Ghana Bolivia Malawi Paraguay Liberia Jamaica Mali Trinidad & Tobago Madagascar Zimbabwe Jordan Niger El Salvador Nigeria Guatemala Rwanda Guyana Senegal Honduras Sierra Leone Peru Somalia Algeria Zambia Iran Haiti Saudi Arabia Nicaragua Philippines Venezuela Source: Murshed (2004), table 1.

This section is not an investigation into the causal and sustaining factors of poverty dynamics. Some key measures of poverty, notably infant mortality, are only weakly correlated with a country’s per capita income over time99.

Before considering what causes and sustains economic stagnation and decline it is necessary briefly to consider the evidence for what causes and sustains growth.

5.1 What causes and sustains growth? 5.1.1 Inputs, Institutions & Policies There is broad agreement that growth requires both an appropriate level and mix of factor inputs (at its simplest, labour, physical and natural capital, and, critically,

Risk Factors

38

technology) and appropriate institutions. Studies into the effectiveness of aid have highlighted the importance of good policies for growth.

Econometric studies into growth have highlighted the importance of inputs.100 There is a robust correlation between investment rates and growth. Government investment in infrastructure (such as roads) and people (education) is also important for growth. However, education, while necessary, is not sufficient. It is important that people are productively employed if education is to contribute to increasing output per worker. Investment in equipment may have additional benefits as a means of transferring technology and ideas, which are key to improving productivity.

However, differences in inputs alone cannot explain the variation in growth experiences across countries. Clearly, other factors are at play. Section 2 noted that explanations for the variation in countries’ long-term prosperity include differences in geography (and with it, agricultural potential, the prevalence of diseases, and the ease of access to markets), culture, ethnicity and religion. Econometric studies of growth since about 1960 suggest other factors are important in the medium-term. Economic freedom and property rights are associated with faster growth. Regulated financial services (banks and stock markets) are now agreed to have a positive effect on growth. Openness to trade also appears good, though there is uncertainty about the conditions necessary to realise its benefits. Some studies suggest the benefits may be greatest to those countries already specialising in manufacturing exports. Shorter-term influences include those of economic policies and external shocks.

A theme that emerges from both the econometric and case study evidence is the importance of institutions. Variation in the quality of institutions is thought to go a long way to helping explain why, despite considerable resources, countries such as Nigeria, Zambia, and the DRC have experienced economic decline, while countries with a similar level of income in 1960 - and on the face of it, poorer prospects - have prospered.

A number of recent econometric studies suggest that, while geography and trade openness undoubtedly play a role, their main impact is through the way they help shape a country’s institutions, rather than because of their direct influence on production or other sources of income. Similarly (and, as noted in previous sections), natural resource abundance can either work for or against long-term growth, depending on the quality of institutions. There is thus a growing body of evidence that one of the key factors, if not the key factor, explaining the variation in growth across countries, is the quality of their institutions.

Policies also matter, at least in the short-term. Studies into the impact of aid on growth have tended to highlight the importance of good policies for aid effectiveness.101 However, while good policies may be necessary, evidence suggests that institutions may be more important for growth over the medium-term.

Evidence in support of the dominance of institutions comes from the varied experience of countries that have pursued the core policy objectives of the Washington Consensus.102 The growth experience of those countries in Africa and Latin America that also adhered to the institutional reforms recommended by the IMF and World Bank has tended to be worse than those countries that implemented the policies through home-grown institutional arrangements (notably, but not exclusively, in East Asia). Institutions help explain this103:

Risk Factors

39

• East Asia relied on local (including informal) institutional innovations and capacities, rather than imported models (which tend to be less successful because institutional innovations do not travel well);

• The East Asian policies were better, because they combined incentives with discipline, whereas in Latin America companies tended either to face incentives or discipline (but rarely the two simultaneously).

The clear lesson from Latin America is that policies, in the form of removing macro-economic distortions (over-valued exchange rates, etc), are not enough to deliver growth. It is equally important to establish institutions that protect property and contractual rights104.

There is evidence from a set of case studies that virtuous or vicious cycles of growth or decline can be triggered by even moderate changes in country-specific policies and institutional arrangements, often interacting with external factors.105 This is highly significant because evidence also suggests that, while institutions are key, growth can be initiated without deep and extensive institutional reform. Rather, credible government assertions that the rules of the game have changed for good may be enough. Sustaining the ensuing growth does, however, require institutional reform.106

5.1.2 The Institutions/Policy Nexus One can see policies as a flow and institutions as a stock, implying that policies can change institutions, but slowly.107 So, while policies to remove economic distortions are important for growth, they are not enough without the right institutions.

However, it would be wrong to think of institutions as passively reflecting the impact of past policies: the quality of institutions also has an important effect on the quality of economic policies.

Moreover, much of this evidence is drawn from statistical patterns. These correlations need not imply causation. To be sure of the primacy of institutions in explaining growth (and so decline), it is necessary also to assess the quality of the interpretations given for these relationships, and, in particular, the causal mechanisms proposed linking institutions with economic outcomes.

Three mechanisms appear credible: • the enforcement-capability paradox; • from institutions via macro-economic instability to slow or no growth; • eroded institutions leaving an economy vulnerable to shocks.

At the core of each of these is the interaction between elites and institutions.

The enforcement-capability paradox holds that sustained growth requires states to be powerful enough to enforce the rules of the game, but that states powerful enough to do so also have a large potential for self-serving discretionary action108. While the solution is to ensure that there are adequate checks and balances on the power of executive, the problem is that such controls tend to follow rather than precede economic growth.109 Many developing countries’ political systems have not resolved this tension, resulting in endemic institutional instability.

Risk Factors

40

Macro-economic instability in part flows from the paradox just stated. Statistical evidence shows that the degree of macro-economic variability helps explain variation in countries’ growth experiences from the 1960s onwards110. The core problem here is that, all too often, macro-economic stability is sacrificed in the name of short-term political gain, as elites compete for the resources they need to attain or retain power.

This is not to say that elite competition or conflict is the only cause of poor economic performance. Indeed, in some countries this competition may not be the main factor. However, there is a growing body of evidence that it tends to be up there with the main factors. To understand why, it is useful to consider Ghana.111 From independence until the early 1980s, Ghana’s rulers used a variety of techniques for redistributing resources towards their supporters. Use was made of an over-valued exchange rate and the Cocoa Marketing Board to redistribute income, from the cocoa farmers who dominated the economy and exports, to the regime’s urban and ethnic supporters. This reliance on economically highly inefficient practices was necessitated in part by the central government’s inability to control or raise taxes in the countryside. It was also politically rational, because it allowed the income most appropriate to be targetted more selectively. The economic impact of this redistributive policy was extreme instability and a sharp absolute decline in real per capita GDP, to the extent that, by 1983, per capita incomes were below those on independence in 1957112. Besides being damaging economically, relying on redistribution to maintain power was of limited success politically. The creation of an increasingly corrupt one-party state with its bureaucracy partly subjugated to dispensing patronage did not bring political stability. Instead, this increased the political stakes, and the potential returns from being in power induced intense political instability with frequent coups as groups fought to attain and retain power.

There are many examples of neo-patrimonialism. Many leaders have succeeded in hollowing out their states through increasingly desperate efforts to extract rent from increasingly bankrupt countries113. In extreme cases of ‘spoils –politics’, this has eventually led to violent conflict.

Vulnerability to exogenous shocks is increased in countries where the state (implying key institutions) has been ‘hollowed out’. Resulting weaknesses in the civil service mean that it is poorly prepared to react to exogenous shocks. As a consequence, shocks have a greater impact on economic growth114.

What is clear from the above is that, ultimately, the manifestations of economic decline listed above are all either forms or consequences of political instability.

However, this is not to lay the blame for economic decline and violent conflict solely on a country’s elites. Rather, external influences or actors may play a major role in this competition. The following parts of this section explore the positive and negative roles of external influences on economic stagnation and decline.

Risk Factors

41

5.2 Factors explaining stagnation & decline There are two groups of explanations for stagnation and decline:

• Structural constraints, which are often deep-seated and fixed in the medium- to short-term, make it harder to achieve growth and increase the vulnerability of the economy;

• Contingent factors may cause growth to falter or even output to contract. 5.2.1 Structural constraints These in turn can be split into a number of categories, though these tend to be inter-linked:

a. adverse starting conditions, such as being landlocked, having a large proportion of the population living away from the coast, being prone to diseases such as malaria or those affecting livestock;

b. Section 5.1.1 suggests insufficient inputs are a constraint on growth, where governments may be under-investing in key public goods (education, infrastructure, agricultural research), private sector efforts may not be sufficient (low investment, or investment in form that does not bring in new technology), or the stock of inputs may be eroding (AIDS deaths or a brain drain may be slashing the number of educated or otherwise skilled workers115). Reasons for this under-investment could include:

• low government fiscal or extractive capacity which may severely limit a government’s ability to invest in public goods while maintaining the macro-economic stability also necessary for growth. Institutional issues clearly play a part here. Tax collection will be difficult in an under-developed and essentially rural-based but sparsely populated country116. In part, the impact of this low revenue raising ability can be offset by flexible development assistance;

• some regimes’ lack of credibility mean that even good policies may not entice the private investment necessary for growth;

• However, it may not be inadequate investment itself that that is the constraint (the levels of which have not necessarily been particularly low in many stagnant economies), but rather low productivity that is constraining growth117;

• inadequate access to (protected) northern markets may constrain some countries, given that trade appears important for growth. The Least Developed Countries have largely failed to take full advantage of preferential trade access agreements (such as the EU’s Lomé Convention Agreement). This has contributed to the Least Developed Countries’ share of fast growing world trade declining from 1.9% in 1970 to 0.5% in 1999; during the same period, sub-Saharan Africa’s share decline from 3.7% to 1.4%118. This is partly because the majority of the poorest countries remain highly dependent on the production and export of traditional commodities (such as agriculture, textiles and clothing), whereas much of

Risk Factors

42

the increase in trade has been in manufactures and services (manufactures had increased to 85% of merchandise trade in 1998).

This is problematic because, although protection has declined substantially over the last 30 years, it remains significant for the sorts of products the poorest countries are trying to export (notably agriculture and labour-intensive industrial products, in which the poorest countries have a comparative advantage). The IMF and World Bank119 estimate that in the case of developing countries, the static gains alone from eliminating all barriers to merchandise trade would be in the order of US$80-180 billion a year, i.e. more than twice the annual flow of aid to these countries120.

The significance of these barriers goes well beyond the monetary loss. They may even inadvertently be reinforcing countries’ typical dependence on a narrow range of economic products that are easily captured by elites. By reducing access to markets for products or commodities involving large numbers of people, northern trade barriers make it harder for countries to diversify out of point-sourced commodities (described in section 4.6). While these commodities can provide the resources needed for growth, they can undermine it if key institutions are weak (as is examined further in section 5.2.2.b).

While there is evidence that these structural factors play a part in explaining slow or no growth, only the low fiscal and extractive capacity has been put forward as helping explain absolute decline - even then, it was proposed121 as part of a broader explanation involving elite behaviour. Some of the other factors similarly beg questions (such as, why have many governments typically chosen to under-invest in education?122), which lend themselves to answers using the elite-institutions framework set out in section 5.1.2.

Figure 5.1 attempts to map out the likely relationships between the structural factors and economic stagnation. It indicates that the relationship between these structural factors, and stagnation or slow/stalled growth is not always direct. Such factors help explain both countries’ propensity to stagnation,123 and the duration of economic decline.

Risk Factors

43

Figure 5.1 The links between structural factors and economic stagnation and decline

Geography

Institutions

Productivity

EconomicStagnationEconomicStagnation

Economic Decline

Economic Decline

ExogenousPartly

Endogenous Endogenous

C o n t i n g e n t F a c t o r s

FactorEndowments

Trade

Source: Builds on Figure1.3 in Rodrik (2003) 5.2.2 Contingent factors Bad macro-economic policies or the hollowing out of the state, often resulting from the adverse elite-institutions nexus (described in section 5.12), present some of the most convincing explanations of the causes of absolute economic decline. There are, however, a number of other (independent) factors that can help explain it.

a. Political instability. Although this will often be linked to the elite-institutions nexus, section 3 outlined other causes, suggesting that political instability may have independent effects on growth124.

The evidence for political instability adversely affecting growth is mixed. Though some econometric studies support the link, others either find the relationship to be very weak, or even with the wrong sign125. There is more evidence that the relationship holds in Africa, if not elsewhere126. Indeed, one major series of 30 systematic country case studies focused on explaining growth in Africa has found that “politics over-rides economics in policy making in most countries”127.

The principal mechanism by which political instability is thought to reduce economic growth is via investment. Investment can be deterred because:

• political instability disrupts productive activity and increases uncertainty;

• the political instability is a symptom of the extent to which the state favours diversion/redistribution over production (the mechanism in section 2). That is, de facto (if not formal de jure) property rights are unclear and the risk of appropriation is too high.

Risk Factors

44

Figure 5.2 The Mechanisms linking Political Instability to Economic Decline

Political InstabilityPolitical Instability

Increased Uncertainty

Risk of Expropriation of PropertyJudged high

Dropin

Investment

Economic Stagnation or Decline

Economic Stagnation or DeclineIntensified

competitionover

Resources

b. Resource Curse - Natural Resources and economic decline. The presence of certain types of natural resources can lead to economic decline or stagnation. It is not natural resources per se which are associated with economic decline but the presence of point-sourced natural resource endowments. Oil, diamonds and certain plantation crops such as coffee and cocoa are considered to be point-sourced, whereas fisheries and other types of agriculture have diffuse patterns of ownership and production.128

Table 5.1 listed countries which experienced significantly declining per capita incomes since 1960 - all but six countries can be described as having point-sourced natural resource endowments. Moreover, only six point-sourced countries (Botswana, Chile, Dominican Republic, Indonesia, Egypt and Tunisia) experienced growth rates averaging more than 2.5% per annum between 1960 and 1999.129

A number of mechanisms have been proposed to explain how the presence of point-sourced natural resources affects the economic performance of a country130. These can be categorised into four broad groups. Typically, several of these may be taking place at once; certainly there is no single explanation of how point-sourced resources impact on a resource-dependent economy. The four are:

All other things being equal, point-sourced natural resources will have a positive effect upon the growth of GDP per capita. That is, used wisely, these resources can be a blessing. Unfortunately, other things tend not to be equal, and this positive effect is commonly outweighed by the following negative effects:

• Point-sourced natural resources crowd out other sectors of the economy through: real exchange rate rises, which make exports more expensive and imports cheaper; consequent reductions in trade openness, if the shrinking manufacturing sector is protected; a reduction in investment because the primary commodity based economy is more volatile; and a reduction in investment in education as the returns to education are lower in primary sector dominated economies.

• The presence of point-sourced natural resources warps the political economy of a country by encouraging rent-seeking competition, facilitating corruption, and triggering conflicts over the control of natural resource revenues – conflict which in some circumstances turns violent;

• Though clearly less common than the earlier three, should conflicts over resources turn violent, protagonists can use point-sourced natural resources to finance the ongoing costs of fighting, thus prolonging the war.

Risk Factors

45

These four mechanisms interact with each other: for example, in the presence of a well functioning political system (such as Botswana or Norway) policy decisions can be taken to mitigate the resource mis-allocation problem. This interaction makes it difficult to judge the relative importance of each mechanism. One attempt, looking at 39 countries with natural resource related growth failures, judged that the resource misallocation mechanisms are very important.131 c. Policy mistakes The argument that policy mistakes have adversely affected growth is an old one. Those calling for structural adjustment in the early 1980s pointed to the impact of the inward-orientated, anti-export bias of much of Africa and Latin America. This has been repeatedly contrasted with the more outward-orientated (but in many ways no less protectionist) policies of much more rapidly growing East Asia.

The political economy work presented in section 5.1.2 would suggest that some of these ‘mistakes’ might well not have been mistakes as such, but rather the subjugation of the economy for political purposes.

More recent arguments point to mistakes on the behalf of those advising developing country governments. Evidence was presented in section 6.1.1 that the Washington Consensus has often failed to have the impact on growth, as predicted by the IMF and World Bank. This was not because the core economic policies being promoted were themselves wrong, but because the IFIs did not adequately understand the importance of appropriate institutional structures.132

A second example of a donor’s ‘policy mistake’ is more accurately an implementation failure. The donors have often inadvertently induced a high degree of macro-economic instability in heavily aid-dependent countries through the disbursement patterns of their programme aid (be it balance of payments or direct budget support). Where donors choose to enforce aid conditionality, it can induce large swings in the volume of quick-disbursing aid, in effect either knocking countries’ expenditure plans off-track or forcing them to rely even more heavily on distortionary methods of government financing. Even more worrying, the same result can come about by problems within donors’ own organisations, which can delay disbursements, thereby reducing the predictability of a key revenue source for many poor, but well-performing governments. Country capacity and resilience are, thus, key to the extent to which exogenous shocks destabilise a country.

d. Exogenous shocks Poorer countries tend to be more vulnerable to exogenous shocks both because of their geographical location (predisposing them to natural hazards) and the structure of their economies (where the frequent dependence on a narrow range of activities or commodities makes them vulnerable to changes in the prices of their exports and key imports).

Many major disasters occur in developing countries133. Moreover, these occur more frequently in low-income countries than in other developing countries (occurring on average once every 2½ years in low-income countries compared with 4½ years in other developing countries). The magnitude of these shocks on developing countries’ output is such that shocks of an equivalent proportion of GNI would present challenges to the management of far more developed countries. For example the 1992

Risk Factors

46

southern African drought is estimated to have cut output in South Africa by 2% and by up to 8% in the more agriculturally-dependent Malawi134.

Low-income countries are also more vulnerable to commodity price shocks. The effect on Uganda of a sharp fall in coffee prices is estimated to have averaged 3.5% of GDP per year over 6 years from 1987 to 1992135.

As highlighted in section 5.1.2, countries’ vulnerability to shocks is clearly not entirely exogenous. Weak institutions and government capacity will exacerbate vulnerability either by an inability or failure to prepare, or weaknesses in the response to these crises136.

Indeed, the effect on growth from an inability to manage the distributive conflicts resulting from the initial shock may be greater than the direct effect of the shock. There is evidence that this was the case with the absolute decline in Venezuela’s per capita income since 1980, where a fall in income from lower oil prices could account for only about half the overall decline in income137.

Nonetheless, as the Venezuela case shows, exogenous shocks can trigger economic decline. However, the Uganda case also shows that it need not, given that growth actually increased during the coffee price slump. A similar conclusion can be drawn from the speed with which much of East Asia bounced back from the severe financial crisis of 1997. This suggests that, while there is clearly a role for exogenous factors affecting growth in the short-term, the longer-term impact of the shock appears to depend on the quality of the country’s institutions.

e. Bad neighbourhood effects. A separate form of exogenous shock can be the impact of bad neighbours on a country’s growth. Examples include: the impact on a landlocked country of a deterioration in its seaward neighbour’s transport infrastructure; and the effect of an outbreak of violent conflict within a neighbouring region or a perceived increased risk of conflict leading to a regional arms race.

There is proof that the impact of intra-state conflicts on neighbouring states’ growth prospects is damaging. One estimate is that an intra-state conflict reduces the annual growth rate of each of that country’s neighbours by 0.9 percentage points138. Given that intra-state wars now last on average 7 years, the cumulative effect on neighbours’ growth foregone can be considerable.

However, this effect is also thought to be short-run. Provided that the conflict does not fundamentally affect the structure of neighbouring countries’ economies (and, presumably, their political stability), their growth is expected to bounce back once the war stops. One study, which found a significant short-term effect on neighbours, did not find a significant impact on their medium-term growth (i.e. over 25 years)139.

However, as has been examined in the section on conflict, wars can also spread across borders, with deleterious economic effects. This is an extreme example of a bad neighbourhood effect.

The means by which conflict lowers neighbouring countries’ growth appears to be country-specific. While migration, human capital or even investment factors may adversely affect individual countries, generally their effect is not sufficiently systematic across all countries to be significant. A separate mechanism may be through neighbouring conflicts lowering growth by spurring countries to increase their military expenditure. One estimate is that over time a neighbour typically

Risk Factors

47

increases its defence spending by an amount equivalent to a one-off cost of 4.3% of initial GDP140. This represents a considerable opportunity cost of spending on growth-inducing public goods (education, infrastructure).

Figure 5.3 Mechanisms Linking Exogenous Shocks to Economic Decline

Depending on a country’s initial starting position, these neighbourhood effects might be enough to stall growth, tip a country into short-term decline or prolong either stagnation or decline.

f. Inequality A recent, but now fairly commonly held view, is that the level of a countries inequality determines its rate and pattern of economic growth. More unequal countries are expected to grow more slowly. Although there is some theoretical support for this view, the empirical evidence is inconclusive.

Asset inequality may be more likely to deter growth than income inequality141.

There are two ways inequality can have an effect:

• countries with high wealth or income inequality will tend to under-utilise their productive growth potential to a greater extent than more equal societies, because the poor will tend to find it harder to access the finance needed to realise their potential;

• higher inequality is associated with greater political instability and social conflict, which, as noted earlier, is associated with lower growth142.

Economic Stagnation or DeclineEconomic Stagnation or Decline

Natural Hazards orCommodity Price Shocks

‘Bad’Neighbour

E x o g e n o u s S h o c k s

Increasedtransport and other

costs

Drop in Investment

Intensifiedcompetition

over Resources

Higher militaryexpenditurein responseto conflict

Lower spending on

growth-enhancing public goods

Lower Productivity and possiblyInvestment

Increased Uncertainty

Economic Stagnation or DeclineEconomic Stagnation or Decline

Natural Hazards orCommodity Price Shocks

‘Bad’Neighbour

E x o g e n o u s S h o c k s

Increasedtransport and other

costs

Drop in Investment

Intensifiedcompetition

over Resources

Higher militaryexpenditurein responseto conflict

Lower spending on

growth-enhancing public goods

Lower Productivity and possiblyInvestment

Increased Uncertainty

Risk Factors

48

Figure 5.4 Mechanisms Hypothesised Linking Inequality to Economic Decline

High Income

Inequality

High Income

Inequality

Countryunder-utilises

ProductivePotential

Drop in Investment

Economic Stagnation or Decline

Economic Stagnation or Decline

Political Instability

LowerProductivity

?

Research suggests there may be a two-way relationship between inequality and growth; that growth (or stagnation) may itself affect the degree of inequality. Though econometricians are more sceptical of a relationship from growth to inequality, there is some evidence from case studies that the relationship may be too country-specific for any generalisation to be made143.

Summary There is broad agreement among economists that economic growth requires the appropriate level and mix of “factor inputs” labour, physical and natural capital and technology. However, these are only proximate drivers of growth. Good quality political and economic institutions are needed to ensure the level and mix of inputs is appropriate and used productively. Institutions both set the incentives elites face and, ideally, constrain their behaviour so they act in the interest of the majority. This merely emphasises that all of the factors that determine the level of instability are interlinked, as are their effects. Political instability, violence, and economic decline can all reinforce each other, creating a vicious downward spiral. It is essential to understand the mechanisms behind these forces, in order to develop solutions to respond to them and to create a virtuous circle to increase stability.

Risk Factors

49

Appendix 1

Botswana and Sierra Leone Case Studies Both Botswana and Sierra Leone gained their independence from British colonial rule within a few years of each other (1966 and 1961 respectively). Both had similar levels of per capita GDP, and rich diamond deposits. Nonetheless, they have followed very different paths and histories since independence. At independence Sierra Leone was not an obvious candidate for state collapse, and had few ‘inherent’ governance problems such as size, distance, population density or a history of violent conflict. Instead it had a favourable social endowment with significant capacity of administrators, lawyers and other professionals that had been built up over decades. Botswana on the other hand, which gained independence in 1961, had virtually no infrastructure, a handful of university and secondary school graduates, and previous outbreaks of violent conflict, yet it is now one of Africa’s success stories. The interests of the elites at national and local level and those of the broad mass of the population are what have differed fundamentally between the two, as well as the nature of their diamond resources. In Botswana the interests of the elite have more or less converged with ‘the national interest,’ while in Sierra Leone, from very early on they were highly divergent, with central government and local chiefs protecting very specific powers and economic interests. The basic dynamics of these political and economic alignments or misalignments in these two countries developed in the interaction between the patterns of colonial power dynamics and pre-existing governance relationships and structures. In addition, the diamond resources in Botswana were easier to control than those in Sierra Leone.

Risk Factors

50

Botswana

Summary/conclusion • Alignment of elite and broader population’s interest, formed by unique interaction

between: a broad-based, accountable, pre-colonial governance system; colonial neglect; the development of an inclusive national identity; and the incorporation of the ruling party of urban and non-urban elites and interest.

• This created an environment where ‘good’ economic policies could prevail. • There are significant challenges that face the current state and it will be a

substantial test of the system to see if it can respond and adapt itself to address the threat of HIV/AIDS and broaden out the distribution of growth.

Outcomes Since independence, Botswana has been politically, socially and economically stable. It has benefited from an absence of violent conflict and the highest per capita growth rates in the world over the last 35 years. The increases in per capita income, from US$1,070 in 1961 to US$8,800 in 2002, have also been relatively evenly distributed. There have been three democratically elected presidents and the transitions between them have been successfully managed. The smoothness of these transitions is in part due to the fact that they were all from the same party - the Botswana Democratic Party - which has ruled continuously for the last 40 years. However, the distribution of the gains from economic growth is emerging as a serious challenge for the future, as is the devastating impact that HIV/AIDS is having on human development outcomes, particularly life expectancy, which had previously seen dramatic improvements. Resource base Pre-independence Botswana had little in the way of physical resources. It was landlocked, sparsely populated and with a largely pastoralist, subsistence economy. The first large kimberlite diamond deposits were discovered around the time of independence, and now the three large mines are the economic motor of the country. Most of the Botswanan economy is diamond related, with a substantial proportion of the population employed in the sector; diamonds make up around 40% GDP and 90% of export earnings. Kimberlite diamond deposits are easier to control, because a mine can be coordoned off and guarded. Elites and institutions Botswana seems to have had a quite a unique legacy of institutions that combined to largely engage and harness elite interest and behaviour to those of the broader population. Accountability, inclusiveness and national identity Twsana governance structures included kgotlas, technically an advisory body, which in practice provided a forum for consultation, criticising the King, and generally holding rulers to account. In addition, while cattle were individually owned and the source of status and wealth, there were underlying factors that produced a general alignment of the economic and political interest of communities and their aristocracy. These factors include the communal ownership of land and mafisa, a clientalistic

Risk Factors

51

pastoralist production system that included a high level of the social and economic integration between large cattle owners everyone else, through systems of lending and usufruct rights. While today only around 50% of the population is actually from the Tswana tribes, ethnicity has not played a dominant role in Botswana’s development. The dilution of the role of ethnicity was reinforced and extended through institutional arrangements, including non-Tswana wardenships, and the threat provided by a series of common enemies (particularly the expansionism of the Zulu kingdom and then the Afrikaaners, in the earlier part of the 19th century). The lack of colonial rule The British held the territory mainly to have access to the interior of the continent, and thereby deny this to other powers. For the early part of the colonial period, the level of actual colonial engagement was extremely limited. Attempts to finally establish some authority over the chiefs only occurred in the 1930s, which succeeded only in further uniting the chiefs, who brought a joint legal case challenging the decision. While the case was lost, World War II hindered implementation of the policy. The formation of the Botswana Democratic Party (BDP) The early stirrings of party politics in Botswana began in 1960 with the Bechuanaland Peoples Party which mainly consisted of a narrow group of urban elite and workers. In response, and led by chief Khama, the BDP was formed, integrating an emerging elite of teachers and civil servants with traditional chiefs. As the chief of largest Tswana tribe, and with his European university education, it was Khama himself who bridged the gap between urban and rural interests and traditional loyalties. While the BDP has had a monopoly on power since then, it does appear to have been responsive in its policies on a number of occasions to the threat of losing power. Policies Botswana has largely and very successfully implemented economic policy more or less along the lines of the Washington Consensus. The national development planning processes is central to managing the economy and public financial management systems have a good track record. While the obvious conclusion is that the policies were more or less right, at least for this situation, the more substantive question seems to concern the way in which it was possible to conduct “good economics”. The answer seems to lie in the relationships between elite and national interests discussed above. For example, the diamond industry is 50% state-owned. This generated a lot of government revenue and has not acted as source of major distortions and corruption. The wealth that this has generated also gave the state the capacity to hire in the required technical assistance from very early on. The success of the pragmatic decision not to insist on ‘Africanisation’ provides an interesting contrast to the technical assistance supplied and controlled by the aid industry which is often neither trusted nor used effectively by the recipients, and has, as such, been much less successful.

Risk Factors

52

Sierra Leone

Summary/conclusions • Corruption was a key issue in undermining the authority and capacity of the state,

but also became a symptom of disaffection and state failure. • The underlying causes of the conflict centre more on the exclusion and alienation

of the population, the breakdown of the social contract at a national level, and of disillusionment, in particular, of rural young men.

• This both caused, and was the consequence of, an increasingly extremely weak state; remaining institutions became a superficial political, social or economic infrastructure.

• Diamonds did not cause the problems that eventually resulted in the outbreak of violent conflict. However, they did motivate particular movements of forces, and help to sustain and affect its course.

• Regional interference and sponsorship of parties to the conflict - in particular of Liberia’s NPFL - contributed significantly to the start, and conduct of the conflict.

• One of the most fundamental aspects of state weakness was the lack of a coherent central government, beyond what was essentially a rent-seeking elite.

• The use of militias (e.g. Kamajors) and tribal forces was central to the nature of the conflict. The use of private security forces has been a notable feature of Sierra Leone’s conflict. Mining concessions were used as incentives for both domestic and international parties in the conflict.

• Lack of consistent and committed engagement by the international community at an early stage of the conflict at an international level, contributed to the length of the conflict. While the regional organisation, ECOWAS, was committed to finding a solution to the conflict, it had few appropriate resources with which to do so.

Outcomes Sierra Leone had many of the classic risk factors that lead to instability, cursed as it was with: rich resources; a rent-seeking elite; and declining state capacity. Once the conflict began, it descended rapidly into a more and more vicious conflict. At independence in 1961 average per capita incomes were estimated at US$1,070, but by 2002 they had declined to only US$480, one of the lowest in the world. The Truth and Reconciliation Commission, set up in the aftermath of the conflict to establish a truthful and authoritative historical report, found that, “by the early 1990’s, greed, corruption and bad governance had led to institutional collapse, through the weakening of the Army, the police, the judiciary, and the civil service. The entire economy was undermined by grave mismanagement.”144 Coups were common, and in 1991 a military coup ousted the government, but failed to suppress rebel forces. In the ensuing war, coup was followed by counter-coup and the scale and frequency of atrocities escalated. There was a configuration and reconfiguration of formal and informal military alliances, which became more and more difficult to distinguish on the ground. The blurring of loyalties and activities extended into the rebel groups and government soldiers, leading to the emergence of “sobels” (soldier/rebels). Most combatants, including the official army, did not receive pay, and so adopted a “pay yourself” approach, ensuring that looting and pillaging became a pervasive feature of the conflict.

Risk Factors

53

The 1991-2002 war is said to have resulted in at least 50,000 lives being lost and, at different points, to hundreds of thousands of people being displaced, with all of the adverse effects on the social and economic fabric of communities and societies that this entails. Resource base (e.g. natural resources, aid etc) Currently the alluvial diamond industry makes up a large proportion of the economy. However this dependence is itself evidence, not so much of the overwhelming wealth of diamond resources in Sierra Leone, but of the relatively underdeveloped status of the agricultural and mining sectors. In comparison, though they have dropped off, official aid flows were averaging £200 million a year between 1993-5. Diversification into other commodities such as rutile and kimberlite diamonds has been undermined by the conflict. Popular analyses of Sierra Leone have focused on the causal role of diamonds in the conflict. In particular they highlight the accessibility and portability of alluvial diamonds as the commodity over which the war was fought. This makes them more difficult to control centrally than kimberlite mines, and easy for them to be gathered and sold by rebel groups. While diamonds did in some way play a central role in prolonging the conflict, the underlying importance of ‘grievance issues’ is often underestimated. High unemployment, lack of effective government, including rule of law capacity, widespread corruption and inequality between elites and the masses were more fundamental reasons for the initiation of armed conflict. Currently Sierra Leone is highly aid dependent and is likely to remain so for the foreseeable future until an effective tax base is rebuilt (for example, over 50% of the 2003 budget was from donor grants). Elites and institutions Early History The early colonial history of Sierra Leone was focused on the creation of a colonial base (the Western peninsular around Freetown in 1787) and the designation of the rest of the territory as a protectorate in 1896. The dominant tribe in the colonial arena came to be the Krio, who began as returnee slaves, and became in time, a fully-fledged tribe. British interest in the rest of the country was largely limited to protecting trade routes and thwarting other colonial interests. Freetown became the increasingly powerful base for central government, and the policies pursued benefitted the colony at the expense of the protectorate. The traditional role of chiefs was co-opted. Setting up competing local government structures effectively undermined the traditional power mechanisms for holding elites and rulers to account. Local positions with a wide range of local regulatory powers gave extremely powerful patrons access to cash levies, land allocations, and increased the ability to divert taxation and punish dissenters. After independence, mismanagement and self-serving corruption exploded in government systems. When combined with the exclusion and marginalisation already felt by the regions, this created an environment fit for dissent and recourse to armed conflict.

Risk Factors

54

Economic Policies Unsurprisingly macro-economic policy in Sierra Leone has been characterised by poor fiscal management, serial devaluation, and the partial implementation of Structural Adjustment measure. The central feature and perhaps also the most interesting wider lesson from the Sierra Leone experience is that, because the fundamental power dynamics were focused on short-term elite gain and the exclusion of other groups, this affected most government policies, even those that should have benefitted the broader population. Structural reforms such as privatisation and the removal of exchange rate control (aimed at reducing the ability of the state distortions), in practice, provided major rent-seeking and patronage opportunities for an elite of ‘insiders’. In the context of the fundamental inability of the state to regulate and provide the right framework of incentives, some of the reforms were more akin to the introduction of unregulated, parallel markets than the broadening of a market economy. External interventions Most accounts of external interventions seem to point to a systematic failure to do anything to contribute to halting the conflict, and may even have aggravated matters. Regional interventions often played into the conflict, with malign intentions in the case of Charles Taylor, or with stabilising ones in the case of ECOMOG. Western interventions also, may have been well intended, but many were seen as ill-conceived. For example, sustained pressure from the international community during the negotiation of the Abidjan Peace Agreement (from governments, the UN and IFI’s) resulted in the withdrawal of the private security company Executive Outcomes, which, not only placed a large contractual financial burden on the government, but also enabled the RUF to exploit the withdrawal of an opposing force, and break the ceasefire, resulting in many deaths. The push by Western governments and the UN for ‘democracy before peace’ in 1996 elections some argue was a major miscalculation. The early peace keeping efforts were not particularly effective. The later deployment of up to 17,000 troops at the height of its mandate and the involvement of the UK showed a marked improvement and demonstrated how seriously the UN took its peacekeeping role. Sierra Leone is faced with the usual problems of donor behaviour potentially undermining the reconstruction of the state through the volatility of aid flows. Too many, non-prioritised and often incompatible demands are placed upon very fragile systems with extremely low capacity. 1 Patrick Chabal and Jean-Pascal Daloz (1999) Africa works: Disorder as political instrument. 2 Two key ones are those reported in Sambanis (2003) and the Africa Growth Project, whose results are not yet out but which is reported on in Nkurunzizia & Bates (2003). 3 Easterley, W (2003) “National Policies and Economic Growth: A Reappraisal, Development” Research Institute, New York University 4 Strategy Unit review of econometric studies of conflict. 5 Daron Acemoglu, (with Simon Johnson and James Robinson) Understanding Institutions, Lionel Robbins Lectures, LSE, February 2004. Easterley & Levine (2002). 6 Murshed, M (2003), When Does Natural Resource Abundance Lead to a Resource Curse, IIED-EEP, Working Paper 03-02, http://www.iied.org/. Also, Daron Acemoglu, (with Simon Johnson and James Robinson) Understanding Institutions, Lionel Robbins Lectures, LSE, February 2004.

Risk Factors

55

7 Jeffrey Frankel and David Romer (1999) “Does Trade Cause growth?” American Economic Review, June 1999. Jeffrey Sachs and Andrew Warner (1995) Economic Reform and the Process of Global Integration, Brookings Papers on Economic Activity. Antonio Ciccone and Francisco Alcala (2001) Trade and Productivity, CEPR Working aper 2001. 8 Daron Acemoglu, (with Simon Johnson and James Robinson) Understanding Institutions, Lionel Robbins Lectures, LSE, February 2004. 9 Easterley & Levine (2002); Mc Arthur & Sachs (2001) “Institutions and Geography: Comment on Acemoglu, Johnson and Robinson (2000)”, NBER WP 8114, 2001; Acemoglu (2004) “Understanding Institutions”, Lionel Robbins Lectures 2004; Murshed (2003), 10 Ola Olsson (2003) Geography and institutions: A Review of Plausible and Implausible Linkages, Working paper 106, Economics Department, Göteborg University. 11 Jütting, J (2003) Institutions and Development: A Critical Review, OECD Development Centre, Paris. 12 Dani Rodrik, Arvind Subramanian, Francesco Trebbi (2002). 13 J.Sachs & J.Gallup, "Agriculture, Climate, and Technology: Why are the Tropics Falling Behind?" American Journal of Agricultural Economics. Vol. 82, pp. 731-777, August 2000, J.Sachs, "Globalization and Patterns of Economic Development," Review of World Economics, Vol. 136(4), Kiel Institute of World Economics, 2000, and J.Sachs and J.McArthur, "Institutions and Geography: Comment on Acemoglu, Johnson and Robinson," with John W. McArthur. NBER Working Paper 8114. February 2001. 14 See Michela Wrong (2000) In the Footsteps of Mister Kurtz for a discussion of the persistence of institutions in the Congo from King Leopold’s to Mobutu’s regime. 15 Murshed, M (2003), When Does Natural Resource Abundance Lead to a Resource Curse, IIED-EEP, Working Paper 03-02. 16 McArthur & Sachs (2001) “Institutions and Geography: Comment on Acemoglu, Johnson and Robinson (2000)”, NBER WP 8114. 17 Easterley & Levine (2002). Daron Acemoglu, (with Simon Johnson and James Robinson) Understanding Institutions, Lionel Robbins Lectures, LSE, February 2004. Rodrik, Subramanian & Trebbi (2002). 18 Daron Acemoglu, (with Simon Johnson and James Robinson) “Understanding Institutions,” Lionel Robbins Lectures, LSE, February 2004, or Dani Rodrik, Arvind Subramanian, Francesco Trebbi (2002) Institutions rule: The Primacy of Institutions over Geography and Integration in Economic Development. 19 Easterly (2003). “National Policies and Economic Growth: A Reappraisal, Development” 20 Dani Rodrik (2004) “Getting Institutions Right”, CESifo DICE Report. 21 World Bank, 2002 data Atlas methodology. 22 World Bank, 2002 data Atlas methodology. 23 Murshed, M (2003) “When Does Natural Resource Abundance Lead to a Resource Curse”, 24 Thomas Carothers (2002), “The End of the Transition Paradigm”, Journal of Democracy, Vol. 13:1. Johns Hopkins University Press and NED. 25 Gates, Hegre, Jones and Strand (2003) “Institutional Inconsistency and Political Instability: the Duration of Polities”, Michigan State University. 26 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004), “It’s all about State Structure – New Findings on Revolutionary origins from Global Data”. Gates, Hegre, Jones and Strand (2003). 27 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004) 28 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004) 29 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004). 30 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004). 31 Jack Goldstone; Ted Robert Gurr; Monty Marshall and Jay Ulfelder (2004). 32 Sir Andrew Turnbull, “Civil Service Reform: Delivery & Values,” 20th February 2004, http://www.nottingham.ac.uk/economics/news/Sir_Andrew_Turnbull-Notes.pdf 33 Campos, Nauro and Jeffrey Nugent (2000) Who is afraid of Political Instability? Working Paper Number 326, July 2000. Available at: http://eres.bus.umich.edu/docs/workpap-dav/wp326.pdf 34 Campos, Nuget and Robinson (1999) 35 Przeworski, Alvarez, Cheibub and Limongi (2000) Democracy and Development: Political Institutions and Well-Being in the World, 1950-1990, Cambridge University Press. 36 Seymour Martin Lipset (1959), “Some Social Requisites of Democracy: Economic Development and Political Legitimacy”, The American Political Science Review, Vol. 53.

Risk Factors

56

37 Barro (1999) “Determinants of Democracy”, The Journal of Political Economy, Vol. 107. 38 Burkhart and Lewis-Beck (1994) “Comparative Democracy: The Economic Development Thesis”, The American Political Science Review, Vol. 88. 39 Michael L. Ross (2001) “Does Oil Hinder Democracy?”, World Politics, Vol. 53, also Collier, P. 1998. “On Economic Causes of Civil War.” Oxford Economic Papers 50(4): 563–73, and P.Collier 1999. “On the Economic Consequences of Civil War.” Oxford Economic Papers 51: 168–83. 40 Seymour Martin Lipset (1994) “The Social Requisites of Democracy Revisited: 1993 Presidential Address” – American Sociological Review, Vol.59. 41 Myron Weiner (1987) “Empirical Democratic Theory and the Transition from Authoritarianism to Democracy”, PS, Vol. 20. 42 Caty Clement (2003) “The Nuts and Bolts of State Collapse”, Not published. 43 State Failure Task Force: “Phase III Findings”, 2000. Hegre et al (2001) and Reynal-Querol (2002), among others. 44 Englebert (2000). 45 Fearon, James & David Laitin, 2003. ‘Ethnicity, Insurgency, and Civil War’ American Political Science Review 97 (1): 75-90. 46 Elbadawi, Ibrahim & Nicholas Sambanis, 2002. ‘How Much War Will We See? Explaining the Prevalence of Civil War’, Journal of Conflict Resolution 46(3): 307-334. 47 Hegre (2003) 48 Goldstone (2004), 27. 49 See for instance Frances Stewart and Valpy FitzGerald (2000) “The Costs of War in Poor Countries: Conclusions and Policy Recommendations” in War and Underdevelopment, Oxford University Press. 50 Reynal-Querol (2002), p.2. 51 Dani Rodrik (1998) “Where did all the growth go?”. Available at: http://ksghome.harvard.edu/~.drodrik.academic.ksg/conftext.pdf 52 Gates et al (2003), p.22. 53 Collier et al (2003) Breaking the Conflict Trap, World Bank 54 Collier et al (2003) Breaking the Conflict Trap, World Bank 55 See for example Collier & Hoeffler (2003, 2004a), Fearon & Laitin (2003), Hegre et al (2001), Elbadawi & Sambanis (2002), de Soysa (2002), Reynal-Querol (2002) 56 de Soysa (2003) ) “Paradise is a Bazaar? Greed, Creed and Governance in Civil War, 1989-99” Journal of Peace Research, 39(4): 395-416. State Failure Task Force ‘Phase III Findings’, Goldstone et al (2000). http://www.cidcm.umd.edu/inscr/stfail/, 57 Rubin, Armstrong & Ntegeye, see B.Rubin et al., “Draft Discussion Paper I: Conceptual Overview Of The Origin, Structure, And Dynamics Of Regional Conflict Formations,” G.Ntegeye, et al., “Draft Discussion Paper II: Origin, Structure And Dynamics Of The Regional Conflict Formation In The Great Lakes Region of Central Africa,” and A.Armstrong et al.“Draft Discussion Paper III: Regional Approaches To Conflict Management,” in B. Rubin et al., Conference On Regional Conflict Formation In The Great Lakes Region of Central Africa, 11-13th Novermber 2001, Nairobi, New York, Centre On International Co-operation, NYU, 2001, www.nyu.edu.. 58 Data from Collier Hoeffler model. Used by Macartan Humphreys, Economic and Violent Conflict, Harvard University, 2003. 59 UNCTAD (2003) The Least Developed Countries Report 2004, UNCTAD, Geneva. 60 Collier et al (2003) Beyond the Conflict Trap. However researchers use past income to predict future conflict, this points towards causality also flowing from levels of GDP to conflict. 61 de Soysa (2002). 62 Fearon & Laitin (2003) 63 This explanation may also help explain the fact that econometric studies have not found a measure of inequality between the richest and poorest a significant determinant of the risk of conflict. This inequality, measured using the gini coefficient, is termed ‘vertical’ inequality. See section 4.8. 64 Collier-hoeffler model 65 Fearon-Laitin and Hegre et al 66 Hegre, Havard, Tanja Ellingsen, Scott Gates, Nils Peter Gleditsch (2001) “Toward a Democratic Ciil Peace? Democracy, Political Change, and Civil War, 1816-1992” American Political Science Review 67 US State failure task Force; Marta Reynal-Querol (2002) Political Systems, Stability and Civil Wars. Available at: http://econ.worldbank.org/files/15110_MRQPoliticalSystems.pdf; Hegre et al, also Jennifer Gandhi1 and James Vreeland, "Political Institutions and Civil War: Unpacking Anocracy" August 30, 2004, www.yale.edu/ycias/ocvprogram/Gandhi_and_Vreeland1.pdf , Annalisa Zinn,

Risk Factors

57

Anocracy and the Onset of Civil War: A Reconsideration, January 17, 2004, www.yale.edu/cpworkshop/ZinnPaper.pdf . 68 This work is most closely associated with Collier (Collier & Hoeffler (2004a, b), Collier et al (2003)). Others have failed to find natural resources significant (see work of Fearon (2004), and Fearon & Laitin (2003)), though they tend to argue that this is because of data & specification problems with the Collier & Hoeffler model, and they are supportive of the concept that opportunity is more important than grievance as the motivating factor. 69 See Sambanis (2003) for summary 23/24 and evidence from country case studies. 70 See e.g de Soysa (2003). 71 See Sambanis (2003) and Ballentine & Sherman (2003) Beyond Greed & Grievance, and Ross (2003a) “How do Natural Resources Influence Civil Wars? Evidence from 13 Cases”. 72 Synthesised in Sambanis (2003) 73 Sambanis (2003). Though Ross (2003a) has cast doubt on the looting as a cause of conflict, finding no evidence in the 13 cases he studied “that rebel groups funded start up costs from looting natural resources or extorting money from resource funds” (page 37). 74 Ross (2003a). 75 Ross (2003b) “What do we know about Natural Resources and Civil War?” 76 See summary in Chapter 1. 77 Reported in Sambanis (2003). 78 See Eldabawi & Sambanis (2000 & 2002), Fearon & Laitin (2003) & Sambanis (2003). 79 Sambanis (2003). 80 Cramer, C (2003) “Does Inequality Cause Conflict?” Journal of International Development, 15(4): 397-412 81 Collier & Hoeffler (2003) on duration and (2004a) on onset. 82 Stewart, F (2000) ‘Crisis Prevention: Tackling Horizontal Inequalities’, Queen Elizabeth House Working Paper Series 33, Oxford. Another comparative set of case studies presented in Sambanis (2003) also provides evidence – from Senegal and Lebanon - of violence driven by inter-regional inequality 83 Nordås, Ragnhild (2004) 'State Religiosity and Civil War: Does Religious Heterogeneity and the Role of Religion in States Influence the Risk of Intrastate Armed Conflict?', Department of Sociology and Political Science, NTNU. 84 Murshed, Mansoob & Scott Gates (2004, forthcoming) “Spatial-Horizontal Inequality and the Maoist Insurgency in Nepal” 85 Proxied by the ratio of the nationals of the conflict-affected country living in the USA over the national population living back home. 86 Sambanis (2003). 87 Examples given in Humphreys M (2002) ‘Economics and Violent Conflict’, Harvard University 88 UNDP, Least Developed Countries Report 2004, New York, UN, 2004. 89 G.Zachariah “Regional Framework For State Reconstruction In The Democratic Republic Of The Congo,” Journal of International Affairs, Autumn 2004. 90 G.Zachariah (2004). 91 Phil Orchard (2003) University of British Columbia. Human Security Report 2004. 92 UN DPKO, personal communication. 93 Joseph J. Collins, Robert Perito & Joanna Spear ‘Conflict Prevention and Resolution Forum: October 14, 2003’ http://www.sfcg.org/documents/CPRF/CPRFOctober2003.pdf Bridging the “Post-Conflict Security Gap” Human Security Report 2004. 94 UN DPA 95 UN DPA 96 Human Security Report 2004. 97 Developing data sets in International Conflict Management: Bercovitch’s Correlate of Mediation Project. Jacob Bercovitch and Judith Fretter, University of Canterbury, 2003. 98 One reason for this is that measured GDP/GNI is itself an imperfect measure of actual output, since by definition it does not accurately capture output from the informal sector. It is not known, therefore, whether an increase in measured output reflects: an increase in the output of the formal sector; or a shift into the formal sector of previously under-estimated informal activity. 99 Macrae at al (2004) Aid to Poorly Performing Countries: A Critical Review of Debates and Issues, ODI, London available at http://www.odi.org.uk/publications/poorly_performing_countries/Aid_to_PPCs.pdf

Risk Factors

58

100 This and parts of the next paragraph draw heavily on Temple (1999), (http://www.nuff.ox.ac.uk/Users/Temple/abstracts.htm) who synthesised numerous econometric studies from the mid-1980s onwards which sought to explain post-1960 growth experiences. 101 There has been considerable research into this since the late 1990s. Early work at the World Bank concluded that aid is only effective where there are good policies. While more recent econometric work has tended to qualify this – and note exceptions, such as in countries emerging from conflict – evidence from case studies tends to support this initial broad generalisation. This debate is briefly summarised in Beynon (2003) “Poverty Efficient Aid Allocations – Collier/Dollar Revisited” ESAU Working Paper 2, ODI. 102 Notably adherence to sound money, fiscal solvency, respect for property rights and market-orientated incentives (see Rodrik, 2003b). 103 This and the 2 bullet points are taken from Rodrik (2003b). 104 Keefer, P & S Knack (1998) “Political Stability and Economic Stagnation” in Borner (1998). 105 Rodrik (2003a) In Search of Prosperity (page 9) 106 Rodrik (2003a) In Search of Prosperity presents evidence from India that economic growth in the 1990s can be traced back to relatively minor changes in policy in the 1980s that changed officials’ attitudes and supported a belief that the rules had changed. Similarly, there is evidence that Indonesian growth stalled in part in the late 1990s because institutional reform failed to keep pace with growth. 107 Rodrik, Subramanian & Trebbi (2002) 108 Borner (1998), “Introduction”. 109 See Chang, Ha-Joon (2001) “Institutional Development in Developing Countries in a Historical Perspective – Lessons from Developed Countries in Earlier Times” paper presented to European Association of Evolutionary Political Economy (EAEPE) 2001 Annual Meeting. 110 Temple (1999). 117 The following is taken from Acemoglu et al (2002) 112 Easterly (2001) 113 See Allen, C (1999) “Warefare, Endemic Violence & State Collapse in Africa” Review of Africa Political Economy, 81, pp 367-84 on spoils politics, and de Walle (2004) ”The Economic Correlates of State Failure” in Rotberg (2004) Why States Fail for descriptions of how states are hollowed out. 114 de Walle (2004) 115 It is predicted that some countries in sub-Saharan Africa face economic collapse unless they can bring their HIV/AIDS epidemics under control (WHO (2004) Changing History: World Health Report 2004) 116 See de Walle (2004) 117 Rodrik (1999). 118 IMF and World Bank (2001) Market Access for Developing Countries’ Exports, 27 April. Much of the detail in this paragraph is from that report. 119 IMF and World Bank (2001). 120 Though presumably the largest share of this would go to the more developed developing countries rather than CRI. 121 de Walle (2004) 122 see Akyuz, Y & Charles Gore (2001) “African Economic Development in a Comparative Perspective” Cambridge Journal of Economics, 25(3), pp 265-88 on how African emphasis on redistribution led to increased govt consumption and insufficient investment. 123 de Walle (2004) 124 Indeed, as noted in section 5, depending on the level of income and the degree of democracy, the causality may run from economic growth to political instability. 125 For reviews see Nkurunziza & Bates (2003) “Political Institutions and Economic Growth in Africa” Centre for Study of African Economies WPS/2003-03, and Campos & Nugent (2000) “Who is Afraid of Political Instability?” Working Paper 326. 126 Campos & Nugent (2000) 127 The studies are being undertaken as part of the African Economic Research Consortium’s Growth Project, as described in Nkurunziza & Bates (2003). 128 When does Natural Resource Abundance Lead to a Resource Curse? S Mansoob Murshed (2003). Papyrakis , E & R Gerlagh (2004) “The resource curse hypothesis and its transmission channels” Journal of Comparative Economics, 32, pp 181-193. Bannon, I & Collier P (2003) Natural Resources and Violent Conflict: Options and Actions, World Bank. 129 Murshed (2003), see also Stevens, P (2005 forthcoming) “’Resource Curse’ – Is It Inevitable and if not, Why Not?”

Risk Factors

59

130 For a comprehensive survey, see Stevens P (2003) Resource Impact – Curse or Blessing? A Literature Survey, University of Dundee. 131 Papyrakis, E & R Gerlagh (2004). 132 A cogent example of his argument is Rodrik (2003b). He notes that none of the countries which achieved fast growth – including those who adopted institutional structures different to those promoted by the donors – did so without minimal adherence to key principles of sound economic governance, notably “property rights, market-orientated incentives, sound money, fiscal solvency” (Rodrik, 2003b, page 15). 133 IMF (2003) Fund Assistance for Countries facing Exogenous Shocks available at http://www.imf.org/external/np/pdr/sustain/2003/080803.pdf Between 1990 and 1998 94% of the worlds major disasters and 97% of disaster-related deaths were in developing countries. 134 IMF (2003). 135 IMF (2003). 136 The ‘hollowing out of the state’ related by de Walle’s (2004) makes sense in this context. Purely weather-related explanations of drought are no longer deemed valid, with researchers such as Sen pointing to political and other institutional determinants. 137 Hausman, R (2003) “Venezuela’s Growth Implosion: A Neoclassical Story?” in Rodrik (2003) In Search of Prosperity 138 Collier & Hoeffler (2004b), available at http://www.copenhagenconsensus.com/Files/Filer/CC/Papers/Conflicts_230404.pdf 139 Murdoch & Sandler (2002) “Economic Growth, Civil Wars, and Spatial Spillovers”, Journal of Conflict Resolution, 46, pp55-73 140 This is the present value of higher spending over the 7 years of the average conflict and a decade following cessation of hostilities. Reported in Collier & Hoeffler (2004b). 141 Ferreira, F (1999) “Inequality and Economic Performance: A Brief Overview of the Theories of Growth & Distribution” prepared for World Bank website at http://www.worldbank.org/poverty/inequal/index.htm 142 See e.g. Rodrik (1999) “Where did all the growth go?” 143 See Bourguignon F (2004) “The Poverty Growth Inequality Triangle” paper presented at Indian Council for Research on International Economic Relations, New Delhi, Feb 2004. 144 Sierra Leone Truth and Reconciliation Commission Report, ‘Findings on the Causes of the Conflict’ p.6, http://www.nuigalway.ie/human_rights/Docs/Publications/Sierra%20Leone%20Report/SL.TRC.V2.2.Findings.pdf