What the OECD Does Not Understand About Fragile States Seth Kaplan

February 3, 2013 | More on Conflict and security, Economics and development | 2 comments

The Organisation for Economic Co-operation and Development (OECD) Development Assistance Committee (DAC) and its International Network on Conflict and Fragility (INCAF) do an admirable job bringing together policymakers, collecting and synthesizing information, and helping set the agenda for donors.

But, as exemplified by Emmanuel Letouzé’s (lead author) and Juana de Catheu (co-author)’s recent report Fragile States 2013: Resource Flows and Trends in a Shifting World, its analysis of fragile states is flawed in a couple of important ways.

My major complaints are:

1) The report does not have an accurate conceptualization of why states are fragile

Although the OECD has reached a “broad agreement . . . on what constitutes fragility,” its definition—“a fragile region or state has weak capacity to carry out basic governance functions, and lacks the ability to develop mutually constructive relations with society” (page 15)—says nothing about the underlying sociopolitical dynamics that play a dominant role in causing states to be fragile.

While the report correctly argues for a shift from a “thin” to a “thick” conceptualization of state fragility, it considers a “thick” conceptualization to encompass “the multiple dimensions of state-society relations” (page 13) and a “substantive understanding centred on the quality of state-society relations and with greater attention to potential stress factors, including economic vulnerability, demographic dynamics, climate change and technological innovation” (page 15). The report is so focused on the nature of the relationship between the state and society that it mentions it 15 times. It refers to social contracts—considered key to fixing these relationships—13 times.

In contrast, it does not mention identity in the whole report, and mentions the word sectarian once. This ignores the most fundamental question determining whether states are fragile or not: do groups of people with different identities, ideologies and interests feel enough solidarity and fraternity with their fellow citizens to work together inclusively? The report thus pays no heed to the fact that “society” as understood in more cohesive nation-states—such as the countries that make up the OECD—does not exist in these deeply divided places. Just look at Syria today. How will this country build any relationship between its future government and society if its various groups are unable to work together with a certain degree of trust? Accepting these political identity divisions as part of the landscape is crucial to determining how to enhance stability and improve governance.

In addition, it assumes that the nature of the state can be determined solely by whether it can “carry out basic governance functions,” ignoring the degree of institutionalization—an important factor in determining legitimacy and resilience (even when the state underperforms). Governments such as India’s may have “weak capacity to carry out basic governance functions,” but they are anything but fragile because of how history has conditioned their peoples to accept common identities, borders, and institutions. In contrast, countries such as the Soviet Union (before 1991) and Syria (before 2011) have relatively robust capacities to carry out basic functions, but do not have the cohesion and legitimacy to survive major changes to how they operate.

The OECD definition depends too much on Western ideological conceptions of how states ought to work, and ignores inconvenient (to Western mindsets) factors such as identity and history—despite ample evidence from across the world that these latter factors are crucial.

These misunderstandings lead the OECD to a wrong assessment of what countries are fragile. Its list of 47 states includes none from the Arab world even though these are now widely recognized as among the most fragile (the list was drawn up based on pre Arab Spring data). On the other hand, it considers Iran and North Korea as fragile even though neither of these have sociopolitical fractures (Iran is diverse but that does not mean Iranians do not have for the most part a common identity). Bad policy and leadership does not necessarily make a country structurally fragile, at least in the short-term. It may lead to poor development outcomes, but then so do the policy choices of some well-developed countries (including members of the OECD). If the governments of these two countries changed, it is likely they both would become dynamic rather soon instead of slipping into the abyss that sectarian divided states such as Syria (a country barely mentioned in the report) finds itself. They are much less likely to implode during any transition than other countries that appear broadly similar precisely because they are not structurally fragile.

The International Dialogue on Peacebuilding and Statebuilding’s New Deal shows a much better understanding of what causes fragility, probably because it better reflects the opinions of fragile states and not just the notions of rich donors. Its focus on “legitimate politics (inclusive political settlements and conflict resolution); security; justice; economic foundations (employment and livelihoods); and revenues and services” (page 35) makes much more sense given the challenges fragile states face.

For more on these issues, see: Social Covenants: The Missing Ingredient in State Building Efforts

2) The report fails to differentiate between fragile states and “fragile moments”

Because they have no sound theoretical basis for what denotes fragility, organizations such as the OECD and the Fund for Peace, which annually publishes the Failed States Index, grope about when trying to formulate lists and indicators. As a result, they often act as if every developing country that has weak governance indicators (by the World Bank’s standards, which fail to capture important information about states) and experiences turbulence is fragile.

But this masks great differences in underlying dynamics, leading to the inclusion of countries that are structurally sound but in need of substantial change (with, for instance, their social contracts and leadership) as well as those that are structurally fragile, and unlikely to fix their problems using the standard set of remedies.

The former group is likely experiencing a “fragile moment” (which, admittedly, may last a few years) that they will be able to escape relatively easily after a transition period. They need more accountable governments and better economic policies, but are likely to have the cohesion and institutions able to produce them after some short-term turmoil. The latter group, in contrast, are likely to be fragile for a very long time, and are in need of long-lasting measures that strengthen cohesion, take into account ethnic or religious divisions, and increase institutionalization. Elections and economic reforms may exacerbate their conditions if not accompanied and implemented in ways that take into account these deeper issues.

Tunisia, for example, is more akin to an Eastern European country transitioning to democracy than a fragile state because of its level of cohesion and institutionalization. It will likely be the most successful product of the Arab Spring. It does not belong on fragile state lists (just like Eastern European countries such as Poland would not have if such lists existed in 1989); it is experiencing a fragile moment that will pass.

Countries such as Syria and Libya (and Iraq, Lebanon, Yemen, and so forth) are much less cohesive and much less institutionalized. Their peoples identify with a subgroup (religious, ethnic, tribal, or ideological) more than with the central state, and have had little experience working together to govern a country. The governments did not exist before World War I, and have a very limited ability to act capably and independently of the underlying social fabric. Even though they appeared on no fragile states list before the Arab Spring, they were fragile then and will continue to be so for many years to come.

For more on these issues, see: Differentiating Between Fragile States and Transition Countries

3) The report makes too big a deal of the fact that more fragile states are now middle-income countries (MIC) than in the past

Income is a bad determinant of country condition, and can easily mislead when it comes to fragility. This is especially true when strongly influenced by natural resources, which can boost average incomes without improving how a state works (or how most people live). Natural resources may, on the contrary, make states less likely to work well by increasing the chance of elite capture and conflict.

Yet, the report—and analysts such as Laurence Chandy and Geoffrey Gertz—argues that middle-income fragile states deserve to be classified separately than their poorer counterparts.

This may be true in a few select cases, such as Bosnia, because “aid is a small part of the development equation in middle-income economies, more attention needs to be focused on how aid can leverage structural change and catalyse non-aid flows to promote development” (page 11).

But it fails to take into account how natural resources (in places such as Nigeria, Angola, and Sudan, all of which depend on oil for over 88 percent of their exports according to the table on page 19) and debt distort country categorization and how small changes in incomes (from as little as $1005 to $1006, the minimum threshold to reach middle-income status) may push countries (such as Pakistan and Yemen) into the middle-income bracket without substantially changing underlying dynamics. Indeed, it is often wrong to use the World Bank classification system to categorize developing countries precisely because of the great number of false readings it produces. (As Jonathan Glennie has noted, between 1978 and 2003, 25 MICs fell into the low-income country (LIC) category; most of the states that have graduated from LIC to MIC status in the last decade were actually MIC at some point in the past.)

The report acknowledges the deficiencies involved with using income as an indicator in a number of places, such as when it admits that “measuring success through the narrow lens of short-term results, using ‘thin’ indicators such as GDP growth and food availability without regard to the impact of development programmes on social and political dynamics, can result in spectacular disaster” (page 47) and points out that “graduation to middle-income status is based on income per capita and does not fully represent the level of development in a country” (page 104).

Increases in income without concomitant increases in institutionalization and inclusiveness explain why the report finds that “in spite of this shift in income level, poverty remains concentrated in fragile states” (pages 11, 29, 30, 35) and that “fragile states are almost always well below their trade and investment potential. This is most obvious in the case of fragile states endowed with abundant natural resources” (page 104).

What would happen to these countries’ categorization if petroleum returned to its 1999 price of $19 per barrel?

Yet, the report makes wide use of the phrase “nearly half of all fragile states are now classified as middle-income countries” and depends on this categorization for a number of important conclusions (such as when it states that “the share of the global poor living in MIFS has increased 17-fold over the past five years alone” on page 32).

The OECD (and the World Bank and other development players) ought to use a categorization system that is more useful for analysis.

For more on these issues, see: Do World Bank Country Classifications Hurt the Poor?

4) The report misleadingly states that “not one of these countries [fragile states] has achieved a single Millennium Development Goal” (pages 1, 30, 35)

As Paul Harvey of the Secure Livelihoods Research Consortium has pointed out, this type of characterization is “misleading.” Based on a similar but subtly different quote in the World Bank’s 2011 World Development Report on conflict, security and development (which actually uses a different set of countries and definition than that used in this report), it ignores the fact that some of the states discussed “are on track / very likely to meet at least five of the eight MDGs (See Gates et al 2010 and MDG Monitor),” including Afghanistan, Bosnia, and Ethiopia.

For more on these issues, see: Are Fragile States Really Failing to Meet the MDGs?

SLRC has a nice chart monitoring progress towards the MDGs for the developing countries that they track.

5) The report shows a lack of understanding of what social cohesion is and how inequality affects fragility

Instead of considering how intergroup dynamics create fragility—as people such as Frances Stewart, Jeffrey Herbst, and Joel Barkan have—the OECD believes that poverty and inequality are influential, seeing that “situations of fragility clearly have common elements—including poverty, inequality and vulnerability” (page 11). As a result, it argues that “by weakening social cohesion, poverty is also a cause of fragility” (page 34).

This conceptualization of social cohesion and the factors that determine how well a society holds together through difficult times is rather “thin,” only taking into account economic indicators. A “thicker” conceptualization would see social cohesion as a “complex cultural, psychological and social phenomenon,” as Duncan Green put it on his blog earlier last year, understanding that how strong the “social glue” and feelings of togetherness that tie people to each other (what I call “horizontal social cohesion”) matters much more to resilience.

Although reducing social inequity and inequality are important, they most matter in fragile states when they overlap with ethnic, religious, clan, or class differences, creating ”fault lines” that lead to conflict. Elsewhere, large differences in income may hurt the poor but can be sustained for long periods of time without producing fragility (see Brazil, Botswana, Chile, Hong Kong, and so on). Robust growth that helps everyone may, in fact, be accompanied by sharp increases in inequality that pose no threat at all.

The countries most successful at promoting development and reducing poverty have been horizontally social cohesive, not vertically so. Places like China, Vietnam, Indonesia, and the rest of East Asia have improved the lives of over a billion poor people over the past two generations not because they had “good governance” (as defined by the World Bank), progressive taxes, or followed a certain policy playbook, but because they had the social glue that lubricated business and encouraged leaders to focus on inclusive development.

For more on these issues, see: Horizontal Versus Vertical Social Cohesion: Why the Differences Matter

6) The report suggests that aid to fragile states is improving despite the lack of evidence for this

The OECD report states that “some recent evaluations point to progress in the quality of aid provided to fragile states” (page 63), but provides very little evidence to support this claim. In fact, it only mentions three reports, and one of these (Chandy, 2011) is based on the other two, which only evaluate themselves, and are not based on any independent verification. Most donors are not evaluated at all.

It also concludes (on the same page!) that “every fragile state has had at least one aid shock; and all have experienced considerable volatility”—something very hard to reconcile with this claim.

Aid needs to be evaluated on a comprehensive basis for what it does to countries and not on a project-by-project basis to determine its overall effects on state fragility. Donors could run hundreds of successful individual projects yet do little to decrease fragility. They could, in fact, increase fragility—by, for instance, supporting only one area or ethnic group at the expense of others—even though each of their projects worked “successfully.”

For more on these issues, see: Brookings: How Not to Evaluate the Effectiveness of Aid

*        *        *

Despite these many criticisms, I did also find some very useful data and analysis in this report:

1) The financial information is unsurpassable. There is no better source for finding out information on the financial flows to the 47 countries it covers. The charts and graphics are great.

2) Chapter 3 provides a very useful forecast of possible future trouble spots:

Based on aid and growth prospects, countries that are of particular concern in the coming years include those that:

• are considered chronically under-aided: Madagascar, Togo, Niger, Bangladesh, Guinea, Burkina Faso and Nepal notwithstanding the fact that all, except Niger, can expect slight increases in aid between 2012-15;
• combine projections of falling aid and slow growth, such as Sudan, Chad and Kosovo. Among these, Kosovo faces a particularly sharp fall in aid; and
• demonstrate high aid-dependency and projections of falling aid, such as Afghanistan. (page 91)

3) The analysis of the “eight global factors of fragility” (page 18) provided a useful framework to analyze these issues even though I would call them “exacerbaters,” not factors. These aggravate underlying structural drivers of fragility but do not cause them. In some countries, however, they may tilt the environment more towards conflict because of how they affect the incentives of major actors.

4) The emphasis on leveraging aid to promote investment, remittances to promote resilience, and domestic sources of revenue to enhance accountability is also welcome. As the report correctly indicates, “in fragile states, there remains significant scope for leveraging ODA and remittances to increase private sector inflows” (page 12). Wherever possible, such countries ought to lessen “their dependence on aid by reforming their tax administration and policies” (page 43). “A large body of literature outlines the importance of remittances as a source of external finance, and their use and impact on development issues such as health and education, nutrition and poverty, productive investment, socio-economic equality, women’s empowerment, and even climate change mitigation” (page 73).

5) The macro analysis of foreign aid and financial flows is very welcome, and provides much insight. As mentioned above, the information on aid shocks is particularly troubling.