Chapter 1. What main trends currently affect the fragility landscape?

1.1.1. Fragility is about more than economic growth

As more nuanced forms of fragility have become better understood, it is now recognised that a lack of economic growth is but one of many dynamics that can contribute to fragility and, by extension, that economic growth alone is not a panacea for fragility. This is reflected in the emergence of an additional category of middle-income fragile states or (MIFFs) (The Economist, 2011[4]). The data generated for this report reinforce the relevance of this category: 30 out of 58 contexts in the 2018 fragility framework are classified as middle income. A recent comparison also demonstrates that a context may be fragile but still able to produce economic growth and qualify as a fast reformer (Whaites, 2017, p. 7[5]). Nonetheless, belief in growth as the cure-all for fragility is difficult to dispel completely. The universality of fragility is increasingly accepted, but it is still tacitly assumed that a country or context eventually will develop its way out of fragility.

This assumption continues to sustain an overarching fixation on poverty reduction and may not take into account dynamics that perpetuate pockets of poverty in middle-income countries and yet often have little to do with actual resources. Middle-income status also changes how donors approach their engagement and where they can reasonably expect to have positive impact (Sumner, 2013[6]). Moreover, recent discussions have highlighted the limitations of assuming that a virtuous cycle of economic growth, social change and institutional development will automatically take place (Whaites, 2017[5]). Indeed, serious new risks manifest when impressive economic growth and its attendant expectations fail to bring commensurate progress on income distribution, job creation, and increased voice and accountability.

1.1.2. Fragility is about more than institutions

The inverse relationship between strong institutions and fragility is another example of initial assumptions that, in hindsight, were overstated. It grew out of the popularity of statebuilding and the crucial importance attached to strengthening institutions, which also correlated with the broadening of the OECD definition of a “fragile state” as one that “has weak capacity to carry out basic governance functions, and lacks the ability to develop mutually constructive relations with society” (OECD, 2011, p. 11[7]). This state-centric lens of fragility focused on authority, legitimacy and capacity, but from a formalised notion of governance. While support to strengthen weak state institutions is critical, any approach to fragility must look at governance more comprehensively – beyond governments – to account for the many ways people experience authority, legitimacy and capacity. Focus on vertical state-society relations, with the state usually embodied in the national/central government, tended to underplay the importance of horizontal society-society relations and local/municipal dynamics, which have an impact on state-level dynamics. Moreover, the earlier emphasis on institution building also skewed perspective too narrowly onto the central, formal state and obscured the impact of people and societies in shaping the foundations on which institutions are built. Ignoring this wider picture often masked why pockets of fragility persist.

Just as prosperity is no guarantee against fragility, a state cannot provide enough services to deliver its way out of fragility either – a realisation that has led to a more nuanced idea of the state’s role. This shift in mind-set recognises that state legitimacy does not derive solely from the existence or demonstration of authority and capacity, for instance through service provision, but from a complex process informed by a multiplicity of issues (Secure Livelihoods Research Consortium, 2017, pp. 7-8[8]). This does not mean that strong state institutions which are able to exercise authority, legitimacy and capacity are undesirable. But it does problematise the simplistic assumption that fragility automatically will be addressed once state institutions are strengthened. In fact, the existence of relatively strong central government institutions within otherwise fragile contexts can provide opportunities for issues of authority and capacity to be instrumentalised for the purpose of diverting resources and attention away from deeper political and societal drivers that can have a larger impact on legitimacy and fragility (Leclercq, 2016[9]). This situation arises especially when certain branches of an authoritarian government are able to cannily use the language of development and local ownership to further consolidate their control (Gisselquist, 2017, p. 1276[10]).

Contexts that may be strong from an institutional standpoint, but inherently fragile because of the way they maintain that strength, can accurately be termed brittle. For example, Kaplan (2014[11]) warned that “interpreting overt stability as a reflection of fundamental strength or resiliency has often set the international community up for a surprise”. A focus on institutions also implicitly appraises regime type. But the phenomenon of so-called “violent democracies” – those with relatively credible electoral processes and middle-income status but also high rates of violence – has tempered expectations about what it is possible for democracy to deliver if other issues are ignored or suppressed.2 This does not mean democracy is undesirable. But it does indicate that fragile contexts do not vote their way out of fragility.

1.1.3. Fragility is about more than conflict

Traditionally, fragility and conflict have been most strongly correlated. Conflict is and certainly always should be a central concern of the international community. However, it does not tell the whole story of fragility. By the same token, the 9 extremely fragile contexts that are experiencing conflict, out of the top 15 extremely fragile in the 2018 framework, should not monopolise international attention or distract development actors from less conspicuous fragile settings that also deserve considerable, albeit different, engagement. The human suffering within active conflict clearly requires special attention. But contexts that are stuck in a fragility trap can lead to potentially more prolonged suffering and thus also need attention.

Places that are stuck in such a trap are variously described as stalled, stagnant or chronically fragile. They are not all the same, however, and their differences matter. It is important to consider the duration of their fragility and differentiate those that are truly chronically fragile from those experiencing temporary fragility (Gisselquist, 2017, p. 1273[10]) – or a transition that could be considered a “fragile moment” that might even last years (Kaplan, 2014, p. 49[11]). Transitions are especially important to identify, as they require fundamentally different international support and present critical windows of opportunity to help peacefully foster change. For some extremely fragile contexts, especially those that have previously experienced conflict, even reaching a point of “resilient stagnation” might be desirable because it ideally provides opportunities to encourage those who hold wealth and power to see the benefits of pressing for more dynamic and widespread development (Putzel, 2010, p. 2[3]). It is thus important to tailor not only interventions in each context but also expectations of what is beneficial and feasible.

1.1.4. All levels of fragility – not just extreme fragility – matter

Fragile contexts can be seen broadly as those places that have been left behind, to use the vocabulary of the 2030 Agenda for Sustainable Development. The subcategory of chronically fragile contexts then can be considered places that have been left behind even among the wider category of fragile contexts. They experience different forms of feedback loops and conflict alone cannot explain why some contexts are trapped in fragility and others are able to exit or move in and out of fragility (Carment and Samy, 2017, p. 25[12]). These also are places that experience lacklustre progress, sluggish growth, only nominal institutional reforms and sometimes sustained, low-intensity conflict. They may demonstrate enough momentum to keep from slipping into deeper instability or back into full-scale conflict, but not enough to propel them towards greater resilience, despite often considerable investment from the international community.

Over the past decade, 27 countries have figured consistently in OECD fragility reports3 (see Box 1.1). Of these, 19 have not experienced major conflict during this period but remain in situations of fragility. These contexts may not grab as many headlines as other fragile contexts, particularly those in active crisis. Regardless of their low profile, or because of it, they nevertheless demand sustained commitment from the development community. Unfortunately, these are also the contexts that can be cited to justify the assumption that fragility is too complicated or cannot be helped and that for a multitude of reasons certain places are destined to stay fragile forever. Yet, conversely, these are also the contexts that with the right mix of resources and attention could build on progress, however modest and static, that already has been achieved and on the substantial investments already made.

And it is likely that more places may fall into this subcategory of chronically fragile contexts in the future. While many of today’s active conflicts hopefully will come to an end, some are likely to persist at lower intensity and contribute to sustained fragility. It can be expected as well that the benefits of global economic growth will still bypass many fragile contexts. Patience and long-term commitment to support true generational change may take as many as 20 to 40 years, according to development bank estimates, and such places will continue to struggle with vulnerabilities throughout the transition (Ncube and Jones, 2013[13]). Expecting more profound and faster change in the most challenging environments, and with less time and fewer tailored resources, is the ultimate fool’s errand of the development sector. Yet it is one that is undertaken with surprising frequency.

1.1.5. Addressing fragility is about moving from complexity and categorisation to concrete action

The complexity of fragility should spur, not deter, new thinking about how best to continue to evolve policy understanding and programmatic approaches. The broad fragility umbrella covers a mix of situations that are fragile for different reasons and more or less resilient for different reasons. Fragility will look very different in a small island developing state than, for instance, in a subnational pocket within a large, natural resource-rich country. It will look different again in a country that has been in conflict for ten years or, for that matter, in one that exhibits impressive economic growth but poor progress in ensuring that growth is inclusive. External actors also can affect all types of fragility. This is highlighted in Trend Twelve: Fragility is a complex and dual-system problem. When these actors become part of a system, they have an influence on it and their actions provide the resources and incentives (or disincentives) for change. Assumptions that fragility can be simplified or isolated into one dynamic or another, have long outgrown their utility for crafting policies and programmes to address it.

At the same time, it is a delicate dance to retain the necessary analytical complexity without nuancing fragility all the way into irrelevance for field-level practitioners. The OECD has put forward a multidimensional framework of fragility that considers fragility as a combination of risks and coping capacities whose interaction shapes how fragility manifests. This is an important step towards sophisticating the way fragility is assessed and clarifying how to respond to it. But this is not the final step in fragility categorisation. Nor should the elements included in the framework be the only consideration. If fragility has demonstrated anything concretely, it is that the category must not get locked in by assumptions and preconceived notions. Developing relevant, concrete and bespoke options for responding to fragility means always retaining a certain degree of flexibility and creativity about how it is captured.

Moving forward, the OECD will continue to explore fragility through different clustering techniques for contexts in similar states of fragility. It also will continue to look at trajectories of fragility and analyse time series to better understand, from multiple perspectives, the behaviour of fragility and the interplay of risks and coping capacities over time. Within this mixed methods approach, it also will be important to reflect on findings, especially those that are anomalies, through a deeper qualitative analysis that takes account of historical, cultural and anthropological features of a given context. All of these features contribute to the overall trajectory of fragility.

There is constant pressure, in this as in efforts to understand any concept, to simplify. But the emerging lesson about understanding and categorising fragility is that it should be allowed to remain as complex as necessary so that it accurately reflects the reality. At the same time, policy makers and practitioners will need this analysis to be further translated into practical tools and clear recommendations that constitute a mechanism for action and response. There are now several policy frameworks that can be culled to advance better responses to fragility, among them the Fragile States Principles, the New Deal for Engagement in Fragile Contexts, the 2030 Agenda for Sustainable Development, and the Sustaining Peace and refocused prevention agendas. The forward trend for fragility thus should be to capitalise on this moment of accrued wisdom and momentum from the new agendas to address the world’s most intractable challenges – and to address them with concrete models for action that honour complexity and context specificity as well as simplicity.

1.2.1. What is driving the internationalisation of conflict?

There is no single explanation for the growing trend of internationalisation of civil wars, but three causes stand out:

Bad neighbourhoods. Research on earlier internationalised civil wars emphasised their link to what were called regional conflict formations or bad neighbourhoods. Intrastate conflicts rarely emerge in geographical isolation. They commonly arise in regions affected by economic weakness, ethnic grievances, cross-border criminal networks, and struggles between local and/or external powers for predominance. In the presence of such conditions, states may inspire or intervene in neighbours’ conflicts to secure resources, expand their influence or secure themselves from genuine threats to their own security from overspills of violence. In recent years, the Middle East, the Sahel and Central Africa have been most affected.

The threat of transnational terrorism. Since the September 11, 2001 attacks, many powers have intervened in foreign conflicts to stem or contain what they regard as transnational terrorist threats, particularly jihadist groups. A de facto revision of the concept of national self-defence expanded it to include pre-emptive strikes against existing and emerging terrorist bases (Jenner and Popovic, 2017, pp. 4-5[17]). But this is complicated by the fact that many non-state armed groups are deeply involved in local conflicts. Jihadists exploit state fragility and civil wars to expand their influence, building local alliances with less ideological forces, and the boundary between counter-terrorism and actively engaging in a civil war is often unclear or non-existent (International Crisis Group, 2016[18]).

Geopolitical power plays. In an era in which economic and military strength is becoming more diffuse, great power rivalries and regional struggles for primacy are becoming more open. These are taking place in multiple spheres ranging from maritime activities to cyberspace. Competition for influence over fragile contexts – sometimes driven by practical considerations like contests over resources or by less tangible factors such as prestige – are likely to remain part of wider international power plays. Where violent conflict intersects with fragility, this competition risks further complicating efforts to establish sustainable peace.

In worst-case scenarios, as has been seen in the Middle East, all these factors combine to fuel extremely violent clusters of civil wars. Tools such as mediation and peacekeeping have proven difficult or impossible to apply in such complex conflicts, as the United Nations and other external actors have discovered. Additional factors, such as tensions over large movements of refugees and migrants and fears of nuclear proliferation, may also push powers to intervene in other people’s wars even more in the future.

1.2.2. Mechanisms for preventing and defusing internationalised civil wars

Given the limits of current international mechanisms for dealing with internationalised civil wars, what options are available for controlling this phenomenon? The simplest answer is that governments should show restraint and either avoid engaging in other countries’ wars or do so within extremely strict limits.

However, in reality, decision makers in many capitals are likely to encounter overwhelming pressures to intervene in future civil wars. Simply counselling restraint will not suffice. More practicable mechanisms may include:

Regional/international approaches to conflict prevention. Where there is a significant risk of internationalisation of a civil war, outside actors who want peace must factor this risk into their preventive actions. Where conflict looms, the International Crisis Group (2016[19]) has argued, there is a need for multinational “framework diplomacy” and “policymakers should make early and concerted efforts to bring international players to the table to assess their interests, hear their analyses and develop common positions on how to act”. Even if it is impossible to bridge these divisions and external actors support opposing sides in a civil war, policy makers should keep up contacts over limiting and containing the violence. At some point, it may be possible to forge an international or regional agreement on ending a conflict to guide domestic peace efforts, although cases such as Syria show how hard this can be.

Creating incentives for non-intervention. It may be possible to back complex framework diplomacy with coercive or positive steps to incentivise external actors to limit their role in a civil war. One such step might be to ensure that resources illegally extracted by intervening powers from a country in civil war cannot be sold on global markets. Another option could be targeted sanctions and military deterrence against interveners. Additionally, structured incentives, such as regional aid and trade packages or security guarantees, could be offered to actors in regional conflict formations as incentives to co-operation. There are no guarantees of success. Experience has shown that coercive mechanisms such as sanctions may backfire, although the more positive incentives are unlikely to have full effect unless balanced with some realistic penalties.

Reinforcing the international architecture for managing civil wars. A strong case can be made for equipping the United Nations (UN) and regional institutions with tools to anticipate and mediate geopolitical competition over fragile and conflict-affected contexts. Potential steps include investing in the UN Secretariat’s underdeveloped diplomatic capabilities to work with the People’s Republic of China (“China”), the Russian Federation (“Russia”) and other non-Western powers on civil wars; continuing to build up the African Union and other African organisations to handle high-stakes civil wars and bad neighbourhoods on the continent; and a gradual process to establish a regional security architecture in the Middle East to lower tensions among its most powerful players.

In an era of mounting geopolitical tensions, such gradual steps may seem unequal to the risks of further major violence. But efforts to manage internationalised civil wars have the potential to save lives, help states avoid collapse, and reduce frictions among global and regional powers. A prevention agenda focusing on these priorities would help secure not only countries experiencing fragility but the entire world.

1.3.1. The drivers of urban fragility are multidimensional

The incidence of city fragility has implications not just for the immediate residents but also for national, regional and even international stability. Fragile cities are not found only in fragile countries and cities that experience a rapid deterioration in their security, social and economic conditions can generate knock-on effects in ostensibly stable nations. Owing to the ways in which the supply chains and financial flows of cities are increasingly interconnected beyond national borders, disruptions experienced in one city can spread across national and international networks.

As is the case for other contexts, measuring a city’s fragility cannot be readily boiled down to a single factor such as the homicide rate, a water shortage or pollution. Fragility is the manifestation of a convergence of multiple stresses, as illustrated in the OECD fragility framework (OECD, 2016[1]). When these risks accumulate, they undermine the legitimacy of a city’s social contract that binds urban authorities and citizens (Muggah, 2015[23]). A disequilibrium of expectations then arises between municipal leaders and local residents, which in extreme cases can result in a collapse of services. All cities are fragile to a greater or lesser degree, and fragility is not exclusive to poorer urban agglomerations alone. Nor is fragility a permanent condition, as Colombia´s rebounding cities show.

Even so, many of the world's cities today are exhibiting worrying levels of fragility. According to a 2017 news report, the Bank of America Merrill Lynch, citing a USD 75 trillion infrastructure gap, has estimated that 80% of the world’s cities are fragile (Rodrigues, 2017[24]). But fragility relates to more than just the availability and quality of infrastructure, as important as they are. A consortium of researchers, for instance, has identified 11 risk factors for study. These include the speed of population growth, levels of unemployment, income inequality, access to basic services (electricity), homicide rates, terrorism, conflict events and exposure to natural hazards (including cyclones, droughts and floods). This research extends to over 2 100 cities with populations of at least 250 000 inhabitants and examines trends over a 15-year period (Muggah, 2017[25]).

Among the findings of this research is that rapid, unregulated urbanisation appears to be a key driver of fragility. When cities grow exceedingly fast, there tends to be a higher likelihood of city fragility. One reason may be that cities that experience sprawl, such as Karachi or Kinshasa, also are more predisposed to social disorganisation and reduced social efficacy, which in turn are correlated with social tension. Cities registering slower population growth rates tend to be more stable. While the correlation between city size and crime rates is not particularly strong, city growth has a statistically significant effect on crime. In some Colombian cities, for example, a 1% increase in population growth rates is associated with a 1.5% increase in victimisation (Gümüş, 2003[26]).

Other drivers of urban fragility include inequality, unemployment and concentrated poverty (Muggah, 2017[25]). From the United States city of Baltimore to El Salvador’s capital, San Salvador, criminal violence is more prolific in unequal cities compared to those featuring a more equal distribution of income and access to basic services. Social marginalisation together with real and relative deprivation in relation to income, property, service provision and social status are all connected to diminished social capital. There is also some evidence that the clustering and concentration of poverty – for instance, some neighbourhoods in United States cities where more than 40% of the population is poor – are associated with under-performing schools, poor housing and health conditions, incarceration, and higher rates of crime (Valdez, Kaplan and Curtis, 2007[27]).

1.3.2. City fragility can be aggravated and exacerbated

Policing and judicial excesses and deficits can also aggravate urban fragility (Muggah, 2017[25]). When city residents lose confidence in their police officers and the criminal justice system, they frequently resort to private, and in some cases vigilante, solutions. Where the reach of law enforcement is limited, there also tend to be higher levels of mistrust among neighbourhoods and neighbours. As has been shown in cities as diverse as Dili (Jütersonke et al., 2010[28]) and Detroit (Metzger and Booza, 2005[29]), the subjective experience of insecurity and fear has objective effects, among them whether to stay or migrate. Moreover, weak security provision has knock-on effects on other services such as electricity, water and sanitation.

City fragility is exacerbated by exposure to sudden environmental shocks and slow-onset disasters (Muggah, 2016[30]). Flooding, storm surges, high winds, rising sea levels and extreme weather events are increasing in scale and intensity and affect urban residents and city infrastructure. The extent of these risks is chilling. A recent review of more than 1 300 cities determined that 56% of them are exposed to severe natural disasters (Verisk Maplecroft, 2015[31]). This is all the more worrying considering that over two-thirds of the world’s cities are coastal and that 1.5 billion people live in low-lying coastal areas.

When all these factors are combined and compared across cities, global patterns can be more readily detected. A review of 2 100 cities found that roughly 14% of all the world’s cities could be classified as “very fragile” in 2015, the last year for which comparative data were available (Igarapé Institute/UN University/World Economic Forum, 2015[32]). By way of contrast, 66% of the cities were classified as having “medium fragility” and 16% registered “low fragility” scores. African and Asian cities were often the most fragile.

The review classified cities in Europe and Oceania with the lowest fragility scores on average. The least fragile cities are largely concentrated in Australia, Canada, Japan, Norway and the United States. For example, the United States cities of Sarasota, Florida; Syracuse, New York; and Ann Arbor, Michigan, figure at the top of the list of least fragile cities, followed by Bournemouth (United Kingdom), Sakai (Japan), Canberra (Australia) and Oslo (Norway). Even so, many North American and European cities have also recorded deteriorating fragility scores over the past 15 years.

Urban fragility tends to be concentrated in a handful of countries (Figure 1.2). Three of the four most fragile cities – Mogadishu, Kismayo and Marka – are in Somalia. Six of the top 25 most fragile cities are in Iraq; five are in Yemen; and four are in Afghanistan. The drivers of fragility in these 25 cities include high rates of conflict-related violence and terrorism, above-average unemployment, and low access to basic services (Muggah, 2017[25]).

Figure 1.2. Top 25 most fragile cities, 2015

Source: Adapted from (Igarapé Institute/UN University/World Economic Forum, 2015[32]), “Fragile cities”, (data visualisation), https://igarape.org.br/en/apps/fragile-cities-data-visualization/.

 StatLink https://doi.org/10.1787/888933786800

1.3.3. Smarter and more resilient cities

The good news is that decision makers are waking up to the risks of city fragility. There is growing investment, not just in developing smarter cities (Muggah, 2014[33]) but also in cities that are more resilient (Berkowitz and Muggah, 2017[34]) to anthropogenic and climate threats. As shown by some Colombian cities, this requires integrated and inclusive solutions that build the capacities of a city’s individuals, neighbourhoods, businesses and institutional systems to survive, adapt and grow. It is not just about preparing for future shocks, but also investing in routine and daily challenges such as social and economic inequality, poor service delivery, and uneven population growth that sap a city’s ability to respond to disasters when they inevitably occur.

The lesson from Colombia is that urban fragility can be designed out. This requires building city plans and solutions that are designed inter-systemically, with co-ordination across sectors. Take the case of Medellín, which had a homicide rate of 38 per 100 000 in the 1990s – the highest ever recorded. Rather than focusing on a single risk such as threats to law and order, public authorities adopted an inter-sector approach that is known as urban acupuncture. A city’s ability to overcome fragility requires re-imagining the urban landscape and accounting for multidimensional risk factors rather than one single risk factor.

1.4.1. Different interpretations of value can distort the framing of aid

This phenomenon has contributed to simplistic notions of effectiveness. The emphasis on finding “what works” in fragile contexts implies that the only methods to be tried should be those that are already known to succeed. Yet the outcome of development engagements in fragile contexts is by nature uncertain, although lessons learned over the past decade have indicated these engagements should use long-term, adaptive and holistic approaches tailored to the specificities of each context and in which local people and institutions are in the lead (Vernon, 2017[36]; OECD, 2007[37]).

These approaches, however, tend to sit uncomfortably with over-simplified notions of causality and effectiveness. A truism of aid is that the most easily measurable interventions are often the least transformational, while the most transformational are often the least measurable (Natsios, 2010[38]). This is especially true in fragile situations, where structural change is essential. Development actors may know the right goals to aim for, but that does not mean it is easy to achieve them. There are no algorithms or simple linear pathways for reducing fragility.

Indeed, no one can be certain of how to achieve success in fragile contexts. History suggests that progress will happen differently in different contexts. Good aid programmes are those that encourage and accompany this evolution, applying resources and knowledge intelligently, responsively, carefully and collaboratively to maximise positive, if uncertain, outcomes. Achieving these will take many years and many donor country electoral cycles. Such uncertain, context-specific and time-specific programming lends itself poorly to the idea of sticking to what is already known to work.

The focus on “what works” feeds into mechanisms designed to demonstrate value for money. Given the scarcity of funds, it is essential to show voters and beneficiaries that money is being well spent. But value-for-money analysis is often done a priori as part of decision making before the impact (i.e. value) of a programme is known. This puts programmes that are appropriate for fragile contexts at a disadvantage compared with simpler and more predictable approaches elsewhere. Judging programmes solely through the lens of value for money or proven success can contribute to poor decisions, leading to funds being allocated to places where operations are cheaper and easier and to shorter term, simpler and less transformational programmes rather than to programmes that address the complex causes of fragility.

1.4.2. Fear can distort the framing of aid

Fear has become more audible in the discourse around aid to fragile contexts. In these times, politicians may benefit by framing aid as a way to keep citizens at home safe from uncontrolled migration or terrorism. This helps to reassure voters that their governments are keeping them safe and also helps to justify aid in the public mind.

But this framing, when translated into practice, can also distort development programming. An example is the use of aid to keep potential migrants where they are. Much aid to Syrian refugees is tied to preventing them from leaving Jordan, Lebanon and Turkey, which may not be the best solution for them as some may never be able to return home, whatever the outcome of the war (Supporting Syria and the Region Conference, 2016[39]). Nor is this approach necessarily the best policy for countries hosting these refugees, where the rapid influx of large numbers of refugees has resulted in the countries’ citizens being as concerned as their counterparts in wealthier nations about what this will mean for their livelihoods and future.

Ambitions that aid will somehow solve the migration issue in the short term are misplaced. Research shows that development progress tends at first to stimulate migration rather than deter it, and remittances from migrants are often critical to the resilience of people back home (Clemens, 2014[40]). Funds that seek to reduce migration can therefore risk undermining both development progress and coping strategies. Moreover, both experience and common sense suggest that surge funding of this sort is unlikely to succeed in fragile places with underdeveloped institutions and limited absorption capacity.

Fear of terrorism is also influencing development programmes and more of them are likely to be expected to address the risk of violent extremism in some way. This is worrying on several levels. Much has yet to be learned – first, about drivers of violent extremism and second, about effective responses to it. Yet, many projects appear to have been rushed into operation. Lessons learned thus far suggest that reducing extremism calls for nuanced, long-term programming approaches designed to build social cohesion and based on a detailed understanding of local conditions. These also need to operate in partnerships of trust with communities and especially with the people most likely to be at risk and who are not always easy to identify (Royal United Services Institute, 2017[41]). The right programmes can take many years to develop and succeed. Their impact, though, can be undermined by the heavy-handed approaches of security actors, which have been found to be a major source of grievance (UNDP, 2017, p. 65[42]). Some of these actors have been trained and supported by the same donors who are providing development aid.

Fear and narrow definitions of value are just some of the ways in which development assistance is being pushed into uncomfortable territory where it is becoming harder to find the balance between short-term perceived interests and altruism. The distortion of aid by extraneous factors will always occur (Klouda, n.d.[43]), especially as the proportion of international aid programmed in fragile contexts continues to increase (OECD, 2016[1]). The foregoing analysis suggests that for international interventions to be effective in fragile and conflict-affected contexts, they must be clearly and consistently framed above all to reduce fragility and foster resilience. Making this a central pillar means that other factors can be clearly judged by whether they will contribute to or undermine this goal.

This also would allow value-for-money analysis to factor in the need for programmes to be intentionally adaptable, long-term, holistic and tailored to local specificities within a non-linear view of change. It would also allow greater acknowledgement of both the higher cost of operating and the higher risk of failure in such contexts. Because aid in fragile contexts will remain particularly prone to ethical dilemmas, more transparency about the challenges of working in these environments would help to move the debate about development and aid onto a more stable footing. Finally, it would be timely to provide more nuanced guidance to officials within donor governments, who are often struggling with the tension between values and interests, to help them navigate this difficult terrain. It is important to find a balance between these matters, because if aid fails in the fragile contexts where need is greatest, then it could undermine the case for providing aid at all.

1.5.1. Fragile contexts present new challenges for aid

The increasing concentration of aid in fragile contexts lays bare the difficulties facing donors. First, these are the hardest settings in which to achieve results and the most likely to produce failures and scandals. In addition, the institutional design of aid agencies often prevents them from adapting their way of working to fragile contexts. For instance, security concerns mean that staff of the International Monetary Fund cannot even visit some fragile settings. Other donor staff who are able to work in fragile environments face daunting restrictions on their movements and contacts; many are confined to heavily fortified compounds with little access to partner governments and still less to non-state players who might be able to inform decisions. Staff turnover in such environments is often high. Finally, funding cycles are often dominated by short-term humanitarian responses, making it difficult to design longer-term strategies needed to address the fragility and its deeper causes.

The challenges to traditional aid approaches run even deeper, however. Bilateral and multilateral aid agencies have traditionally seen sovereign governments as their natural partners and/or arenas of action. But in fragile contexts, states are often weak or predatory. Many other actors fill the partial vacuum of politics and administration, among them traditional leaders, faith organisations, social movements and armed groups. The actions of individuals and organisations are constrained by these different facets of what is considered public authority, and in ways that are poorly understood by researchers and still less by aid agencies. Furthermore, the instruments of aid – funding cycles, logical framework approaches, project management, and monitoring and evaluation – assume a level of stability and predictability that is often absent in such settings.

Overall, the level of aid dependence of governments in fragile contexts has fallen back slightly over the past several years.4 The relationship of aid to the delivery of essential services in these places raises a number of complex questions including whether aid creates incentives or disincentives for government ownership of these services. In best-case scenarios, the importance of aid can be exaggerated. In Mozambique, for example, health and education are overwhelmingly government-funded, with only water, sanitation and hygiene (WASH) largely aid-funded (Figure 1.4.). But in worst-case scenarios, donors’ willingness to fund basic services can absolve the partner government from doing so. In South Sudan, for example, it is reported that donors fund 80% of health care and that the government funds just 1.1% (Foltyn, 2017[48]).

Figure 1.4. Donors and government spending by sectors in Mozambique, 2016

Source: Adapted from (Government Spending Watch, 2018[49]), “Government Spending Watch” (database), www.governmentspendingwatch.org/spending-data.

 StatLink https://doi.org/10.1787/888933786838

One recent response to the difficulties faced by development actors in fragile contexts is to turn to the international private sector for help. However, international companies face many of the same problems in operating in these difficult settings. High levels of risk and unpredictability do not encourage long-term investment and the possibility of corruption and abuse poses serious reputational risks to brand-conscious companies.

Companies could seek to offset risk through public-private partnership agreements with donors and/or governments in fragile contexts. However, some in the development community and private sector have raised concerns about the higher cost of capital, the lack of savings and benefits, complex and costly procurement procedures, and inflexibility of such agreements in these settings (Green, 20 November 2015[50]).

International companies are most likely to take on the risks of operating in fragile contexts when returns are commensurately high. Extractive industries drill and mine in many such places. But extractives as a sector are capital intensive, create relatively few local jobs, and have a chequered record on human rights and environmental protection.

Local small and medium-sized enterprises (SMEs) are the one part of the private sector that undoubtedly has a great deal to contribute to livelihoods and well-being in fragile contexts. However, SMEs often find it hard to navigate the complicated systems to access funding.

1.5.2. Options for improving the scale and impact of financial flows to fragile contexts

In trying to find effective ways to reduce poverty and vulnerability in fragile contexts, development actors increasingly accept that they must learn to “dance” with the system (Meadows, n.d.[51]), which is a matter of grappling with the messy realities of power and politics and navigating the unpredictable tides of events, opportunities and threats. This often means abandoning or significantly adapting approaches to statebuilding and best practice developed in more stable settings.

Aid professionals have responded to these challenges by setting up networks to find ways of providing aid and support that function better in fragile contexts. One is the Doing Development Differently (DDD) network, whose 2014 manifesto set out its proposed way of working ( Doing Development Differently, 2014[52]).

DDD approaches give weight to understanding the specificities of the local context in order to be “politically smart [and] locally led” (Booth and Unsworth, 2014[53]) and to “work with the grain” (Levy, 2014[54]) of existing institutions. Responses to complex, messy problems need to be iterative, as donors and implementers adapt to changing circumstances and to lessons learned as their work progresses. These approaches are increasingly described as adaptive management.

Several new research programmes also are exploring the role of aid in fragile contexts and the efficacy of these new approaches. A recent analysis of theories of change among donors who seek to promote social and political accountability in fragile settings found an interesting bifurcation in thinking:

In addition to the ideas outlined above, several additional options are worth exploring to try to improve the developmental contribution of financial flows in fragile contexts:

Diasporas and remittances As Figure 1.3 shows and as Chapter 6 discusses further, remittances to fragility-affected countries already eclipse official development aid and foreign direct investment. They also are expected to continue to rise faster and more steadily than either of these other sources. Diasporas that send the remittances also have good knowledge of local contexts and how to support development. Some donors are exploring whether instruments such as diaspora bonds can improve the developmental impact of such flows (Famoroti, 12 April 2017[56]).

Domestic resource mobilisation. Revenue raised from taxation and royalties on natural resources is growing in importance with respect to aid flows, but in many fragile contexts remains at low levels as a percentage of gross domestic product (GDP). Domestic resource mobilisation offers a way to further reduce aid dependence and strengthen the social contract among citizens, the state and the private sector. Until now, however, aid agencies have failed to recognise its potential. Aid figures reported to the OECD suggest only 0.2% of aid to places affected by fragility or conflict in 2015 and 2016 – a trifling USD 116 million in 2015 and USD 110 million in 2016 – was dedicated to technical assistance for domestic resource mobilisation.5 Chapter 7 explores this subject in greater detail.

Localisation. Pushing power and decision making as close as possible to local levels makes good sense in fragile contexts to deal with the enormous variations in conditions over space and time. So far, the localisation agenda has been more apparent in statements than action, however.

1.6.1. States and societies are not uniform

Society is not a cohesive entity forming a single political community in fragile contexts. Particularly in fragile contexts, society tends to be fractured into various groups that often are based on ethnic, religious, clan or other identity. These groups typically have little mutual trust; different perceptions of history and ongoing dynamics (including what others might view as fact); and different concepts of the legitimacy of rules and public authority including the state. Fragility rises or declines with the ability of such groups to work together, especially in the pursuit of public goods. The level of co-operation among groups affects how security apparatuses, administrative bodies and legal systems perform. The more cohesive the society, the more likely these institutional entities will work as advertised, inclusively and without bias.

Relationships among identity groups are much affected by history. Historical legacies may influence the actions of leaders, levels of trust and perceptions regarding a wide range of issues. Any lingering resentment or trauma may make co-operation harder to achieve, as was evident in conflict-prone contexts such as the Balkans and Africa’s Great Lakes region. The chances that divisive leaders may emerge are greater when historical grievances are unresolved. Such legacies are hard to change, especially in the short term.

In fragile settings, horizontal society-society dynamics have an important impact on how vertical state-society relationships evolve. They also influence whether a social contract can be fashioned and the nature of any contract that might ultimately be achieved. In such contexts, developing a social covenant or some other form of pact that brings together various ethnic, religious, clan and ideological groups may be essential to progress on other fronts. All else being equal, once a society settles on shared, fundamental principles and values – for instance, who is or can become a citizen, what makes a government legitimate, or how to accommodate myriad ethnic, religious and regional identities – it is more likely to be able to forge a sustainable social contract. This is particularly the case when institutions are unable to equitably enforce rules and commitments. A positive example is what Tunisia has done in the years since its uprising and political transition in 2011. In contrast, societies that have stark differences in opinion over fundamental principles and values, such as Libya, Syria and Yemen, are likely to struggle to forge a consensus across groups on the type of social contract and government they need.

States, the other component of the social contract, typically do not work uniformly across territories in fragile contexts. The effectiveness of institutions and their ability to serve different parts of society equitably vary across location and entity. Different parts of a government work differently for different people, sometimes based on wealth or identity and sometimes on relationship and corruption. Such differences are apparent between urban and rural areas and between the centre and outlying areas. What is coherent and functional in one place may be incoherent and dysfunctional in another. Furthermore, factors such as limited or non-existent infrastructure, a lack of adequately trained managers, and insufficient or unstable revenue sources hamper the ability of institutions to project authority across all of a state’s territory.

1.6.2. Social contracts require effective leaders and impartial institutions

In some cases, leaders intentionally keep state institutions weak or biased. They may direct a disproportionate amount of resources to regions where their supporters live, for instance, depriving other areas in the process. They may deliberately keep the judiciary, electoral commission, finance ministry and other pivotal agencies weak in order to stay in power or harvest government resources for their own use. In these cases, weak institutions are as much a product of political calculus as of a lack of human, financial and technical resources. Because of how the incentives in these contexts shape behaviour, fragile environments struggle to break out of the vicious cycle whereby weak institutions encourage fickle or biased leadership that works to undermine the institutions that could challenge its authority.

Ultimately, impersonal, capable institutions are needed to achieve fairness. These are institutions that encourage leaders to work on behalf of all parts of a society and make all groups in that society feel they have a set of known and enforceable rights and remedies that cannot be overturned arbitrarily or ignored. In this sense, a country can never have a robust social contract that is genuinely inclusive until its institutions work equitably for everyone, the powerful included.

1.6.3. Exclusion and marginalisation must be reversed

Exclusion and marginalisation are to some extent natural products of the dynamics in fragile contexts. Those with power naturally prioritise their own supporters over everyone else and institutions naturally favour those with power or wealth over everyone else. Reversing this dynamic calls for building greater cohesion across groups and strengthening institutions so that society and the state work more inclusively. These can boost the chance for a robust social contract.

While elections and economic and social reform may be essential to social contract formation, they are rarely sufficient. Too often, they also are overemphasised at the expense of policies that address the deeper, more fundamental dynamics that influence a context’s social cohesion and institutions. Indeed, elections and reforms often produce much less change than expected and can be as much a hindrance as a help to push progress towards an inclusive and legitimate social contract. In Sri Lanka, for example, the Sinhala majority has repeatedly prevented the kind of compromise that would satisfy the minority Tamils. In Kenya, Nigeria and Ukraine, electoral competition for political power has repeatedly increased social divisions rather than healed them. In Guatemala, democracy has repeatedly failed to empower disadvantaged groups.

Social institutions including civil society organisations (CSOs) can play an important role, negative or positive, in the forging of a social contract. They can strengthen, or weaken, social cohesion. They can enhance the accountability of state institutions or work to enhance their partiality. The shrinkage of civil society space in recent years, which typically results from the perception of elites that CSOs are a threat, reduces their scope to have any influence at all.

The state-society relationship, and perception of the legitimacy of public authority and the national rules system (i.e. the state), can vary tremendously by group. Groups that benefit – or perceive that they benefit – from public resources and the rules of the game can have significantly better relations with the state and better perceptions of legitimacy than those who do not.

Social contract concepts are likely to continue to inform actors working on fragility, especially international actors. They will see better results the more they take into account the unique dynamics that drive local contexts in a way that addresses these issues. Policies and projects should prioritise building social cohesion, reducing horizontal inequalities, strengthening the relationships among groups, and ensuring institutions work equitably across different parts of a society and territory. If they do not, international actors risk achieving short-term gains such as higher growth and improved macroeconomic indicators at the expense of long-term prospects, as happened in many Arab countries in the years before the 2011 uprisings. Development and socio-economic gains must be inclusive and broad-based to be sustainable.

1.7.1. Men and boys as vulnerable

Conventional wisdom says that women and children are among the most vulnerable in situations of conflict and fragility along with elderly men, people living with disabilities, and people of diverse sexual orientations and gender identities. While these groups indeed face multiple vulnerabilities, this conventional classification does them a disservice when it disregards the sometimes vastly different needs and vulnerabilities of different women, men and others. The needs of a child of an urban, upper-class family will differ from the needs of an elderly person with a disability living in a rural area. Furthermore, taken together, the categories of people mentioned above comprise, depending on the context, between two-thirds to three-quarters of a society’s population. This hardly allows for a prioritisation of the most vulnerable.

These blanket approaches to identifying the most vulnerable in a society also often assume that young adult, able-bodied men, particularly those without families, have few or no vulnerabilities. New research on Afghan and Syrian refugee men, however, paints a different picture and shows the many vulnerabilities they face. These include exploitation in the labour market, police harassment, lack of access to services because these are geared towards other groups, and sexual violence, abuse and exploitation (International Rescue Committee, 2016[69]; Khattab and Myrttinen, 2017[70]; Turner, 2016[71]; Myrttinen, forthcoming[72]). These vulnerabilities are worsened by men’s internalised expectations – to be tough, stoic and independent, to not seek help, and to fulfil the role of breadwinner for the extended family by sending remittances. Service providers often do not understand these vulnerabilities and lack the skills and resources to deal with them.

1.7.2. Participation of women and girls in armed violence

The vast majority of armed actors are men and, to a lesser extent, boys. But women and occasionally girls also are also direct and indirect participants. This is evidenced by women’s participation in substantial numbers in the Maoist People’s Liberation Army in Nepal, the Kurdish People’s Protection Units (YPG) in Syria and the Revolutionary Armed Forces of Colombia (FARC). They make up an estimated 30-40% of these three forces. However, it is not only in organisations that have at least rhetorically embraced gender equality that women are active as combatants, supporters or recruiters. In recent years, women have also voluntarily joined groups that seek to roll back women’s rights, social space and mobility such as violent far-right and neo-Nazi groups in Europe and North America and Islamist groups like the so-called Islamic State and JAS (Boko Haram) in Nigeria (Fangen, 1997[73]; Milton-Edwards and Attia, 2017[74]; Lakhani and Ahmadi, 30 November 2016[75]). Substantial numbers of women and girls also are members of violent armed gangs, such as the Central American maras.

Certainly, not all women and girls, nor for that matter all men and boys, join such groups voluntarily. However, dominant narratives highlighting only abduction, coercion or brainwashing of women members obscure the fact that women, as much as men, have grievances and political convictions that may compel them to take up arms and that they also do so for other economic or personal reasons. These reasons may include going against dominant gender norms, escaping abusive or repressive family structures, protecting themselves and their family, and finding a new sense of purpose and belonging.

Assumptions that women never join armed groups voluntarily or that, if they do, they were manipulated by men, amount to a denial of women’s agency. Such assumptions also have direct consequences in both peace processes and peacebuilding. Treating women as innately peaceful or solely as victims of war, even when they are former combatants, is positive discrimination and has often led to the sidelining of women (Vinas, 2015[76]). In peace negotiations, which tend to focus on bringing armed actors to agreement on a settlement, women often are either absent or relegated to secondary negotiating tables to discuss what are termed women’s issues. In reintegration processes, too, women combatants have been routinely sidelined, lumped together with other women victims (as in Nepal) or shunted into vocational training programmes that reflect gendered stereotypes more than the women’s skills or wishes. Similarly, in terms of preventing violent extremism, simplistic approaches that focus mainly on using mothers as their sons’ de-radicalisers, potentially place these women in danger; these approaches also overlook women’s multiple roles in violent extremism and complex reasons for participating (Satterthwaite and Huckerby, 2013[77]).

1.7.3. A more effective way forward to overcome gender stereotypes

Conflict- and fragility-affected situations and post-conflict settings should be approached from a comprehensive, nuanced and context-sensitive gender angle. This allows for a deeper understanding of social dynamics and thus improved and effective policy responses to situations of conflict and fragility. A gender angle means breaking away from conventional stereotypes and examining how different women, men and others face new threats as well as new opportunities. In practice, it means seeing men and boys as potentially vulnerable due to dominant social norms. It also means starting to see women, girls, and those with diverse sexual orientation and gender identity not “as objects for charity [but rather as] rights-bearers” who possess social and political agency including, in some cases, the agency to take up arms (Davis, Fabbri and Alphonse, 2014[78]).

It is also necessary to move away from quantitative approaches to vulnerability – such as asking who may be most vulnerable, especially if these are based on pre-determined assumptions rather than research – and to qualitative understandings where there is an examination of how and why different people are differently vulnerable in a given situation.

1.8.1. A failure to understand the implications of hope and aspirations

While thwarted ambitions were part of the discussion around the Arab Spring, the development community still has not internalised the implications for how it can better understand and support the realisation of those ambitions through development programming. This is despite widespread evidence of the critical role that hope and ambition play in shaping people’s perceptions of their well-being. Assessment tools often overlook perceptions, although these can considerably impact how people experience satisfaction or dissatisfaction and therefore can inform subsequent consequences (UN/World Bank, 2018[79]). The migration literature echoes these aspects particularly strongly, finding that feelings of “inescapable stagnation” (Carling and Talleraas, 2016, p. 7[80]) and a “lack of hope in local futures” (Carling and Talleraas, 2016, p. 36[80]) lead people to take risks in order to find a better future elsewhere. Despite the vast differences in contextual dynamics, there are striking similarities between the goals and aspirations of refugees from places as diverse as Cameroon, Jordan, Malaysia and Turkey. (Barbelet and Wake, 2017, p. 11[81]). Displaced people across six different contexts echoed this same lack of optimism about meeting their aspirations, with recent field research finding that the overwhelming majority of people who are receiving humanitarian aid do not feel they are on an empowered trajectory towards self-reliance (OECD, 2017[82]).

Education is perhaps second only to safety as a central input into decision making about whether to stay or to go, and even when adult refugees have lost hope for their own lives, they transfer their aspirations onto their children (Barbelet and Wake, 2017, pp. 11-12[81]). Education also has the valuable potential to be able to reduce the appeal of groups that prey on despair and resentment, with some research warning that extinguishing the hope that results from learning creates conditions that perpetuate inter-generational poverty, fuels instability and undermines prospects for recovery (Watkins and Zyck, 2014, p. 2[83]).

Hope, or the lack thereof, plays a similar role in areas afflicted by high levels of violence. Youth are especially susceptible to reacting when they feel abandoned by their government or their society. Central American maras, or gangs, provide a feeling of belonging and sense of purpose to youth who have so little to look forward to that they feel, as one Salvadoran anthropologist put it, “born dead” (International Crisis Group, 2017, p. 12[84]) A lack of community-level mechanisms that are sufficiently empowered and supported to deal with the scale of desperation leads to a situation where youth, feeling doubly disserved by both state and traditional authorities, have “less and less to lose” (Sears, 2017, p. 17[85]). This sentiment is far from unique to developing countries. A community leader from Los Angeles has summed up the violence he sees as being about “a lethal absence of hope” and added, “Nobody has ever met a hopeful kid who joined a gang” (Cowan, 2017[86]).

Frustrated aspirations contribute to perceptions of exclusion and marginalisation, and these are now increasingly recognised as conflict risks (UN/World Bank, 2017, pp. 16-17[87]). Unfortunately, people who understand how to market mobilisation as an antidote to grievances have long recognised these sentiments as opportunities. Mobilisation, and the belonging that follows it, becomes an attractive beacon in an otherwise bleak and hopeless landscape. Researchers conducting interviews across several African countries recently asked youth which emotion best captured their decision to voluntarily join an extremist group; the most common response was “hope and excitement” (UNDP, 2017, p. 74[42]). Again, this sentiment is not unique to developing countries alone. Research on foreign fighters who have travelled from Western countries to fight with militant groups in Iraq and Syria is still in its nascent stages, but preliminary findings indicate that “the absence of a future and feelings of exclusion” contribute to their search for a cause to embrace as well (van Ginkel et al., 2016, p. 54[88]).

Violent extremism is packaged to provide purpose, belonging, direction and a chance to express a clear personal identity. These represent gaps that cannot be filled by jobs alone even if there were enough jobs, and in most fragile contexts, there are not. While unemployment is a major problem, often underlying it is “hopelessness and a belief that there is no fairness” and it is this, not the lack of jobs alone, that burnishes the appeal of violent livelihoods (Mercy Corps, 2015, p. 23[89]). Affiliation with extremist groups can also provide hope of economic advancement, with loans and payments providing the means to fund more standard business ambitions where there are few other options (Mercy Corps, 2016[90]).

1.8.2. Greater focus on the human element in development

Despite this cross-cutting evidence, why have considerations of hope not featured more prominently in discussions of development impact? The obvious explanation is that hope is subjective and difficult to measure, and thus it would seem out of place and quite likely even unprofessional in a project document or a proposal. As the development community has moved towards greater rigour with tracking results – a commendable professionalisation of the sector – it has left little appropriate space for inclusion or prioritisation of softer and less quantifiable programming goals.

But beyond these practical concerns, there is also a more systemic explanation. One of the key recommendations from States of Fragility 2016 was to put people at the centre of better programming. This related to the report’s acknowledgment of the current preference for structural responses to fragility’s challenges, which are typically more visible and easier to manage and evaluate and may yield more immediate results. But downplaying the human element in development and the individual agency of people leads to a deficit of the delivery of dividends that actually matter to people. Doing better will require continuing to shift the mind-set with which development is approached to one that seriously takes into account well-being as a core component. A focus on well-being is not only about the satisfaction of objective needs and wants, but also about the quality of life that people experience. Far from being an altruistic approach, this also can tangibly affect progress. The role of shame, pride and self-esteem in poverty alleviation continues to be explored including the possibility that shame may perpetuate poverty by undermining individual agency.

1.8.3. Patience for development to deliver is wearing thin

In many fragile contexts, patience to wait and see what development will eventually deliver is wearing thin. Hope was raised that people’s lives would considerably improve and now that hope is fading. Places where expectations were inflated, intentionally or unintentionally, are potentially at higher risk of resentment boiling over when expectations are dashed than are places where expectations never existed in the first place (Mercy Corps, 2015, p. 36[89]). This risk is especially exacerbated where competition for people’s futures is rife and development is no longer the only game in town that is ostensibly offering an aspirational horizon.

Development never was and never will be a panacea. As raised elsewhere in this report, it is fallacious to think that a country can develop its way out of fragility. Nevertheless, development should itself aspire to meet people’s aspirations and, at the very least, understand that to not do this is to potentially negatively affect the very dynamics that development sees itself as an answer to. With the decades of accrued development wisdom, it is now time to evolve to a “new humanism” (Gass, 2017, p. 5[91]) and a greater recognition of the primacy of human agency in affecting peaceful outcomes, ones where development is not an end in itself but a means to deliver the hope of a better future.

1.9.1. All forms of violence are not considered equal

Yet while there is a growing recognition of the relationships among various forms of violence, and specifically how violent extremism largely emerges from other types of conflict, it has not fully penetrated the international community’s approach. Foreign policies, donor commitments and media attention continue to emphasise violent extremism over other forms of violence. This can be seen in the resources poured into defeating the so-called Islamic State in Raqqah and Mosul, which took precedence over tackling the core factors that allowed the group to emerge through ending the Syrian conflict and addressing the exclusion felt by many Sunnis in Iraq. Politicians and media often focus on terrorist attacks and the subsequent calls to address extremism in Berlin, Kabul, London, Mogadishu, Maiduguri, Paris and New York. A result is that attention tends to fix onto these sensational incidents rather than the more persistent violence that affects more people globally, impedes the ability to address fragility and achieve the SDGs, and foments the conditions that allow extremist groups to emerge.

On the first point, as States of Fragility 2016 clearly laid out, violent conflict is a small proportion of the type of violence that affects people worldwide (Figure 1.5). Mc Evoy and Hideg (2017[92]), writing for the Small Arms Survey on global levels of violence, report that 18% of the 560 000 violent deaths in 2016 were due to armed conflict. Of the 99 000 deaths related to armed conflict, only 26% were due to terrorism. However, because of the prominence of terrorist attacks in the media, some people assume these attacks are more common than they really are and as a result, people tend to be more afraid of terrorism than other threats (Kahneman, 2013[93]). This puts policy makers and politicians in a bind as they must choose between being responsive to actual risk or to perceived risk (McGraw, Todorov and Kunreuther, 2011[94]). This attention contributes to the overemphasis on programming to prevent and counter violent extremism compared to other forms of violence reduction programming or even more development-oriented programming that addresses deeper grievances and drivers of multiple forms of violence.

Figure 1.5. Global overview of violent deaths, 2016

Source: Adapted from (Mc Evoy and Hideg, 2017[92]), Global Violent Deaths 2017: Time to Decide, p. 18, www.smallarmssurvey.org/fileadmin/docs/U-Reports/SAS-Report-GVD2017.pdf.

1.9.2. Violence affects all dimensions of fragility

On the second point, the interplay between violence and fragility that keeps societies stuck, unable to address the fragile conditions that contribute to violence or to reduce violence enough to address fragility, is clear. Based on World Bank data, 72.8% of the 705.55 million people living in extreme poverty (defined as USD 1.90 per day) in 2015 were clustered in the 58 countries and settings considered fragile in the OECD fragility framework. These fragile contexts also are often immersed in protracted conflict. Unless fragility in these places is addressed, this percentage is projected to rise to 80% by 2030, the year for achievement of the SDGs.6

Violence frequently is regarded as only affecting the security dimension of fragility. But this interplay between violence and fragility is apparent in each of the other four dimensions of fragility – societal, political, environmental and economic – as well. It creates a negative spiral where fragility and violence worsen over time. In the societal dimension, for instance, grievances among groups can erode trust and create conditions that elites can use to mobilise supporters for violence. This violence destroys whatever trust remains among social groups. Moreover, violence can cause displacement, which further disintegrates relationships and breaks social bonds. Population movements often worsen segregation between groups, contributing to polarisation (Enos, 2017[95]), exacerbating fragility in societies and increasing the ability of recruiters for armed groups to justify the use of violence.

Similarly, along the political dimension, violence creates insecurity, making it harder for governments to provide services whether due to conflict or to other forms of violence such as gangs who control urban neighbourhoods. This adds stress to already fragile state-society relationships where marginalised populations feel the government does not serve them, and it allows other groups – sometimes violent ones – to gain support by promising services. In the economic and environmental dimensions, violence creates the insecure conditions that impede access to markets or productive land, deters investment that would create economic growth, and causes overuse of resources. In 2017, conflict cost the global economy USD 14.3 trillion (Institute for Economics and Peace, 2017[96]). Even low-level but persistent violence has significant costs. In the Middle Belt region of Nigeria, violence between farmers and pastoralists costs the Nigerian economy up to USD 13 billion a year (McDougal et al., 2015[97]).

The interplay between violence and fragility further creates the conditions that allow violent extremist groups to emerge. Most violent extremist groups are able to gain support when the government is largely absent, when there are strong grievances either between groups and/or with the government, and when people feel marginalised and excluded (International Crisis Group, 2016[18]). Most often, there is pre-existing conflict, as was seen in Afghanistan, Iraq, Libya, Somalia, Syria and Yemen. Consequently, to address violent extremism other forms of violence and fragility must be addressed.

1.9.3. Fragility and violence must be addressed together

What are the best ways to address violence and fragility simultaneously? The danger of tackling these issues sequentially is not knowing where to start. Addressing fragility is difficult in the presence of persistent violence. A recent review by Zürcher (2017[98]) of development programmes and their ability to reduce violence finds that these programmes were effective only in more secure areas. However, addressing violence first and ignoring fragility until violence subsides carries its own risks that the conditions that led to the violence remain in place and risk igniting renewed violence (World Bank, 2011[99]). The security approaches used to address gang violence and organised crime in urban areas of Latin America such as Rio de Janeiro, Guatemala City, Bogotá and San Salvador, as well as in protracted conflicts like those in Afghanistan, Iraq and Nigeria, are rarely paired with follow-on programming to address underlying fragility. As a result, violence often returns after much-heralded initial reductions.

One possible way to get a better handle on the violence-fragility link is to layer short-term and longer-term programming. Short-term programming can quickly reduce violence or address immediate needs to stabilise people’s lives. However, that approach is often not sustainable, as experience across the world has shown. Therefore, at the same time or immediately after violence subsides, significant resources should be dedicated to addressing the longer-term issues – particularly those related to inclusion (social, political and economic) and state-society relations – rather than persist in the habitual practice of focusing primarily on infrastructure in post-conflict reconstruction.

In Afghanistan, on a small scale, Mercy Corps combined unconditional cash transfers, which gave young people some short-term economic relief, with vocational training, which gave the youth skills to have more profitable livelihoods. This approach led to decreased support for armed opposition groups like the Taliban, more so than either cash or vocational training alone (Mercy Corps, 2018[100]). Research on this project, as well as a recent study of a cognitive behavioural therapy and cash programmes in Liberia (Blattman, Jamison and Sheridan, 2017[101]), demonstrate the complexity of what motivates people over the short and long term to participate in violence and shows how both need to be addressed. One could imagine using a similar approach with a military surge to increase security in an area in the short term and quickly moving afterwards to implement governance and economic development programmes to address underlying fragility in the long term. Box 1.3 explores related issues of security sector reform in the context of spreading violent conflict.

To return to a point discussed above, while many knew when the Iraqi city of Mosul would ultimately fall, the international community has not quickly moved into the recovery and development phase after gaining control and improving security in this area. Its delay risks the emergence of a new militant group or revival of previously active ones. This is not to say it is easy to address the violence-fragility link. However, the link will not be broken without changing how the international community invests resources and until reducing all types of violence, not just those forms that are the most politically expedient, is a priority. Ultimately, preventing the emergence of violent extremist groups will only be possible by making stronger inroads in addressing fragility and filling the vacuums of governance and rule of law and the isolation that these groups seek to exploit. What is clear is that by myopically focusing on violent extremism, the international community will continue to ignore the conditions that give rise to it, as well as the wider fragility dynamics that provide and protect the spaces permissible to violence of many forms.

1.10.1. Global financial incentives have changed

Global financial incentives have changed as the illicit economy has grown. In short, there are new and multiple places to invest illicit funds. This has important implications for fragility. In the past, the profits from what might be termed organised crime were invested locally. Think, for instance, of the late-19^th century robber barons in the United States. Now, however, criminal money is syphoned out of fragility-affected and developing countries, producing two sets of self-reinforcing incentives. First, important and emerging financial centres around the world have an interest in masking the extent to which they host criminal proceeds. Second, few incentives exist for the local investment of criminal money in many developing and fragile economies. In such a scenario, there is little interest on the part of criminal bosses in seeking long-term stability and security, but every interest in maintaining the fragile status quo.

The result is a growing gulf between what can be seen as places where crime openly occurs, be it in the poppy fields of Afghanistan or around the sourcing of environmental commodities from Asia or Africa, and the places where profits are hidden. Afghan farmers, African poachers and Central American enforcers earn little; the money extracted higher up the economic chain never passes through the most fragile base of the illicit pyramid. This business model funds external middlemen and criminal empires that have few interests in – and in fact strong disincentives for – long-term stability in their zones of extraction. The system has becoming strongly self-reinforcing: criminal value is taken out, promoting harm and poor governance and sustaining fragility, while its benefits are exploited elsewhere. The illicit economy is unequal and exploitative with little prospect of lifting up a second generation of the poor through sustainable investments in things such as education. In the long term, there are only losers on the ground.

1.10.2. Illicit economies lead to their own forms of governance

The second area of concern for fragility is that illicit economies spawn what might be called crime governance. This is in contrast to the strong legal governance of the places where illicit money is invested. At its most basic level, crime governance includes the development of extortion economies in which money is paid to those who promise protection or, failing payment, becomes the source of instability. For example, across the poor and fragile contexts of Central America, and linked to local drug economies, this protection economy plays out on a daily basis. But extortion “taxes”, as they can be called, are present now in many cities in the developing world. Such relationships are unexplored because their victims remain silent. Yet crime governance can be seen where criminal actors call the shots in important, although marginalised, parts of urban complexes around the developing world. Rio de Janeiro, Cape Town and Manila are just some examples. Such zones of fragility and violent economy exist also in the developed world. Legitimate forms of service delivery are driven out and people live in pockets of criminal governance where the gangs hold sway.

Crime governance also has wider implications for political fragility when criminal figures are elected to local or national legislatures and state and criminal interests overlap within government itself. This blurring of the boundaries destabilises regions, promotes criminal interests and ensures that criminal actors are protected by politics. The Balkans, for example, is now widely seen as deeply afflicted. But this region is only one case among several. These relationships are core sources of fragility at the local, national and regional levels. Where criminal markets are politicised, responses then become much more complex. The old exculpatory Cold War refrain – he may be a warlord but he’s our warlord – is being repurposed for the age of the criminal ecosystem.

1.10.3. Organised crime erodes sustainable development

Third, illicit markets and organised crime erode sustainable development across a range of sectors. In contexts that are fragile and where people’s livelihoods are tenuous, the shocks that organised crime delivers to the economy and the environment erode individual and family resilience. The cost is incalculably high – whether from the trafficking of environmental commodities such as timber or animal products that cause soil erosion or undermine livelihoods in tourism; illegal mining of state resources; overfishing by criminal operations; health systems that are infiltrated by counterfeit pharmaceuticals; or the control of local drug markets decimating school systems. Organised crime impacts all the Sustainable Development Goals and, in just over 10% of cases, achieving the goals are directly linked to reducing organised crime (Global Initiative against Transnational Organized Crime, 2015[104]).

Environmental crime markets, beyond their general effect on scarce resources, are of particular concern in fragile contexts where they reduce local resources in a way that is not recoverable and they empower violent and corrupt actors. One feature of these markets is a form of commodity hopping, where criminal groups seek out the next commodity for exploitation when existing markets are depleted. Those engaged in rhino horn trafficking, for example, are as acutely aware as any conservationist that this is a finite resource. Commodity hopping creates spirals of resource depletion, greatly increasing environmental fragility while feeding external tastes and so deepening the unequal divisions between places of extraction and investment/consumption.

In an age of such interconnected global criminal markets, then, the biggest mistake is to assume that fragile zones are disconnected from the global illicit economy. With few exceptions, places that can be classed as fragile demonstrate a web of connections – as sources, transit zones, markets, etc. – for illegal activities and the extraction of profits. In every case, these unequal and exploitative relationships shape local political economies, and almost always with a focus on sustaining the system that maintains fragility, by extracting profits, protecting those who are violent locally, and protecting those externally who feed off and protect criminal activities and interests. Therefore, considerations of illicit economies must be as inextricably linked to attempts to address fragility as they are to its perpetuation.

1.11.1. Climate change affects livelihood security and migration

The link between climate change and displacement is an underestimated global trend. In 2016, 24.2 million people were displaced as a consequence of sudden onset natural hazards (Internal Displacement Monitoring Centre, 2017, p. 31[108]). Climate change is projected to increase the frequency and magnitude of extreme weather events (Field et al., 2012[109]). In tandem with poor coping capacity, these contribute to disasters9 that in turn will contribute to increased displacement and changing migration trends.

Migration in response to changing environments has occurred throughout history. Seasonal migration is a familiar and key livelihood strategy across South Asia, the Middle East, Africa and other regions. However, migration trends are changing. Due in part to the impacts of a warming climate, the duration of circular migration patterns is increasing, as is the need to migrate permanently and to new places. Transboundary migration shows a marked rise in recent years, with nine out of ten international refugees moving to a low-income or middle-income country (WFP, 2017[110]). Evidence suggests that countries with the highest level of food insecurity, which is often linked to climate change and armed conflict, have the highest outward migration of refugees (WFP, 2017, p. 77[110]). Environmental degradation and change – mediated in different ways by governance and political and societal issues – are important structural factors contributing to livelihood insecurity and migration.

The impacts of climate change on Africa, where 80% of people depend on agriculture for their livelihood, have been severe. Drylands are especially affected by climate change. These include the Lake Chad basin region, northern Kenya, southern Ethiopia and South Sudan. In South Sudan, a temperature increase of 2-3 degrees Celsius has exacerbated drought and food and livelihood insecurity (UNDP, 2017[111]; Omondi and Vhurumuku, 2014[112]). A combination of drought and poorly planned development projects has spurred a rapid increase in the movement of people and displacement within and across South Sudan’s borders, which have heightened grievances against the state and led to famine and communal conflicts (Human Rights Watch, 2017[113]).

Migration, along with the prospect of alternative jobs and an additional labour force, can present opportunities for migrants as well as for host societies and countries. However, research on the links between climate change, migration and fragility is currently not sufficient to inform policy responses that can promote migration that is co-ordinated and pragmatic and that also does not drive people towards higher-risk areas such as already fragile cities, coastal megacities, river deltas, and places where livelihoods are already at risk and projected to worsen (Greenpeace, 2017[114]).

1.11.2. Cities are unprepared for extreme weather events and disasters

The world is poorly prepared for a rapid increase in climate change-related natural disasters affecting cities. It is estimated that assets worth USD 158 trillion – double the total annual output of the global economy – will be in jeopardy by 2050 without preventive action (GFDRR, 2016[115]). Although cities offer many advantages for many individuals, urbanisation often exacerbates and highlights inequalities through the proximity of rich and poor. This, in turn, can be a factor leading to instability and conflict. Climate impacts, which are especially devastating in cities given the concentration of people and assets, can deepen inequalities, further fuelling grievances.

The relationship between disasters and fragility is frequently mutually reinforcing. Disasters place additional stresses on already overburdened governments. They also displace large volumes of people, reduce economic opportunities and shift access to resources. A lack of structural and individual coping mechanisms, such as social safety nets, insurance mechanisms and social networks to support migration, can fuel grievances. This is especially the case where national or international disaster relief is inadequate or inequitably distributed.

Poorly planned and executed humanitarian responses to disasters also can worsen tensions and increase fragility. An example is when these responses do not take account of sensitive conflict dynamics such as power relations among different identity groups who frequently live in very close proximity in cities. In fragile cities, moreover, weak governance can undercut efforts to build resilience, making the impact of a disaster more severe. Conversely, disaster risk reduction and disaster management initiatives offer opportunities, if done properly, to address climate-related fragility risks and build peace.

1.11.3. Volatile food prices and provision

The UN’s first-ever global assessment on food security and nutrition finds that world hunger is on the rise, driven by a combination of factors that include climate change and conflict (FAO/IFAD/UNICEF/WFP/WHO, 2017[116]). In a reversal of its steady, decade-long decline, hunger now afflicts 11% of the global population. Between 2015 and 2016, the estimated number of chronically undernourished people rose to 815 million from 777 million. Some of the highest proportions of food-insecure and malnourished children in the world are now concentrated in conflict zones. Many of these areas also are adversely affected by climate change, such as northeast Nigeria, Somalia and Yemen. But even in regions that are more peaceful, droughts or floods linked to climate change and to the El Niño weather phenomenon10 can interact with pockets of instability to fuel conflicts and challenge governance, which contributes to communal and border conflicts over natural resource access (UN OCHA, 2017[117]).

The scientific evidence reliably shows that increasing temperatures, the stronger El Niño effect in 2015 and 2016, and reduction in groundwater resources are major drivers of food insecurity. Food insecurity in turn increases fragility (Box 1.4). There also is clear evidence that climate change is reducing crop quality and yield (Food Security Information Network, 2017[118]; Myers et al., 2017[119]), which “ultimately depend on a dynamic balance of appropriate biophysical resources, including soil quality, water availability, sunlight, CO2 and temperature suitability” (Myers et al., 2017[119]). Lower crop yield and quality decrease food supply and increase prices. In Kenya, in a recent example, drought led to sharp food price increases including a 31% increase in the price of the main staple, maize flour (Okiror, 2017[120]). The drought became a major issue in the Kenyan presidential elections of August 2017, demonstrating how food security issues at the local level can scale up and be politicised at the national level (Okiror, 2017[120]).

Food insecurity, particularly relating to food provision, and price volatility can be strong drivers of discontent (Hendrix and Brinkman, 2013[121]). There are important gendered components of this risk. For example, reducing food insecurity, especially in countries affected by conflict or the legacy of conflict, would have some impact towards reducing incentives for men to join armed groups (FAO, 2017[122]).

The impact of compound climate risks on food security is rightly, albeit slowly, emerging as a major global concern. Adequate policy and practical responses, however, still are lacking. This past year saw acknowledgement of the relationship between food security issues and security issues such as protests and recruitment to insurgent groups, and examples of a more conflict-preventive, risk-based approach are emerging among certain donors in fragile contexts (USAID, 2016[123]).

1.11.4. Climate trends are interrelated

As climate extremes and conflicts are increasing, so too are geopolitical and socio-economic extremes. The world is currently and simultaneously facing high levels of public uncertainty over shifts in the world order, the highest numbers of displaced people in decades, and a peak in global hunger. Migration, extreme weather events and food security are but three examples of interrelated, climate-sensitive drivers of fragility, and they illustrate the complexity of addressing fragility. Strategies that do not take into account the systemic and linked nature of these climate-fragility risks will fail, and may exacerbate the risks they set out to address. Linked risks need a linked response. Reducing vulnerability while creating opportunity – the cornerstone of managing fragility – depends on informed, flexible strategies that build both social and institutional resilience to face linked threats that climate change may pose in fragile situations.

1.12.1. Complexity in the aid system

Even under the best of circumstances, when policy makers accept an authoritative recommendation to implement a well-designed aid package, it is not a straight line from recommendation to implementation. How that package ends up getting delivered depends not just on intent and best practices but also on the requirements and constraints imposed on it by budgeting, management, security, planning, contracting and other bureaucratic functions. It depends as well on the knowledge, norms, incentives, processes and competing priorities faced by personnel in those offices (Lamb and Mixon, 2013[131]; Lamb, Mixon and Halterman, 2013[132]). The interactions among all of these factors within and between donor agencies are generally too complex for anyone to be in a position to predict, with any degree of confidence, how any particular policy decision will be implemented on the ground.

For example, there might be a recommendation to engage with local communities. But the donor’s security office decides whether donor personnel are permitted to travel to areas to meet these communities and the security officer risks getting fired if personnel are kidnapped or killed. A contracting officer may or may not recognise the importance of flexibility in the methods of aid delivery but could be too risk-averse to approve a waiver of certain requirements if doing so would set a precedent. Indeed, there are dozens of reasons why best practices do not get implemented that have nothing to do with the complexity of fragile environments but everything to do with unseen dynamics, practices and regulations within donor agencies.

1.12.2. The dual-system problem

By the time an aid package makes its way through this complex aid system, it is sometimes unrecognisable from what had originally been recommended. Only then does it enter the complex system whose fragility is its target. No wonder it is so hard to solve problems in fragile and conflict-affected environments. Recommendations and decisions about aid packages have to make their way through two complex systems before they turn into outcomes. First the complex aid system transforms aid decisions into aid packages in unpredictable ways. Then the complex fragility system transforms aid packages into development outcomes in unpredictable ways (Figure 1.8).

It is this dual-system problem (Lamb and Gregg, 2016[133]; Lamb and Gregg, forthcoming[134]) that complicates the ability of donors to deliver assistance appropriately. The development community has to acknowledge the dual-system problem and find ways to overcome it, if it wants to improve life for people touched by sectors that are fragile.

1.12.3. Systems thinking starts at home

Many analysts and practitioners continue to employ linear theories of change (e.g. from inputs and activities to outputs, outcomes and impacts) and linear research methods (e.g. correlations, regressions and indices). However, it is fairly common today at least to acknowledge that fragile sectors are in fact complex systems. More appropriate methods, among them systems mapping and political economy analysis, are starting to be employed (OECD, 2017[135]).

Much less acknowledged is that donor agencies, foreign and domestic policy institutions, and the international aid system as a whole are also, technically speaking, complex systems. In fact, authors of programme evaluations, best practices and lessons learned studies frequently produce recommendations that their own systems are fundamentally incapable of delivering, because they are unaware of bureaucratic constraints. It would be better not to focus recommendations only on what policy makers and practitioners should do in the field in the future, but also to focus the underlying research on why policy and aid systems have failed to implement those exact recommendations in the past.

Figure 1.8. The dual-system

Source: Contributed by the author, Robert Lamb.

Human systems are combinations of components. Some are designed, such as constitutions, legislatures, schools and other formal institutions. Some are so-called informal components that are not designed but emerge organically, such as norms, values, social groupings, economic networks and demographics. All of these system components interact in complex combinations and over time the system as a whole behaves in certain ways as a result. In the worst case, cognitive biases, misunderstandings, fear and self-interested behaviours fall into vicious cycles that lead to mass violence, economic crisis or famine. In the best case, opportunities, hope, social progress and behavioural norms fall into virtuous cycles of steady improvement in the quality of life for most people in the system – a stable equilibrium that makes life predictable and pleasant. In between those extremes lies a wide range of dynamics and outcomes often simplified under the label of fragility. In other words, fragility emerges from complexity.

This form of thinking, called systems thinking, forces recognition that the stability of better-developed sectors also emerges from complexity. The relative predictability of those sectors, however, creates an illusion of control. It is assumed that formal (designed) institutions are what make a sector resilient, when in fact it is the constructive interaction between these formal and informal dynamics that produces stable and prosperous sectors in any society. Building more formality into sectors without accounting for informal dynamics is a common approach that has been repeatedly shown to fail (Kaplan and Freeman, 2015[136]). The distinction between formal and informal is not always clear or constructive, in any event. It is better simply to understand each system on its own terms. This is something the OECD multidimensional fragility framework begins to make possible because it attempts to capture not just risks, but also coping capacities.

1.12.4. From systems thinking to systems doing

To overcome the dual-system problem in international development and fragility, systems thinking and methods need to be applied much more seriously to both fragile environments and donor systems. Linear thinking and methods will not work in either, since they generally assume independence between factors. This assumption is almost never appropriate. In complex systems, a great many interrelated components need to be accounted for. Among these are non-linear causality or feedback loops that lead to resistance to change, exponential change or oscillating changes, depending on the system’s structure; accumulations and depletions of resources (e.g. money, resentments, arable land, populations, etc.) that make it seem like nothing in the system is changing until a tipping point is reached and some small change suddenly cascades in surprising ways; and, usually, long delays between causes and their apparent effects that can make it difficult to distinguish between short-term and long-term results (e.g. a short-term success complicates long-term progress).

Methods for studying and planning for complex systems, both qualitatively and quantitatively, do exist (Meyers, 2009[137]). Systems thinking is a minimal requirement, not just in understanding higher order effects and unintended consequences of aid on the ground but also in determining what aid packages are fit for purpose and possible to deliver through donor systems in the first place (Wright, 2008[138]). More formal methods, such as political economy analysis and system dynamics modelling, are helpful in fragile environments. But these methods can and should be applied as well to the policy systems of donor countries in order to identify and rectify sources of resistance to reasonable recommendations and better ways of working (Serrat, 2011[139]; Stroh, 2015[140]; Senge, 1990[141]). Collective strategies can be developed in ways that account for complexity in both systems and make it possible to identify barriers to success from policy decisions all the way through to development outcomes (Lamb, 2018[142]).

The dual-system problem can be solved. But it begins with a recognition that problems in fragile contexts will never be adequately addressed with the assistance of donors who overlook the complexity of their own systems. That complexity has hidden the back office functions, emergent behaviours and non-linear dynamics that have prevented the institutionalisation of effective aid practices for more than half a century. Now that this complexity is known, in both donor systems and fragile environments, making solutions to fragility more effective should not require another half century.