Monday, December 7, 2015

Modelling fragility: Stepping towards universality, by Thomas Wheeler

The universal focus and more nuanced conceptualisation of the OECD’s new states of fragility model is to be welcomed, says Thomas Wheeler. But we should remember improved aid flows are only one part of the solution to highly complex problems

At Saferworld we viewed the OECD’s transition from thinking about “Fragile States” to a more nuanced “States of Fragility” approach as a step in the right direction. It is important to remember what these data-based models are intended for, which is to ‘help international actors do a better job in assisting these states in making a sustainable exit from poverty and insecurity.’   

The good thing about the original move towards listing states as fragile was the recognition that development assistance should not ignore a country’s vulnerability to shocks, especially those related to political crises and conflict. Past practice of treating all low-income countries as the same while ignoring politics and conflict was not only ineffective, but also counter-productive.

Closer to a multidimensional model… and policy responses

Nonetheless, a single list of fragile states has its problems. While perhaps useful for donor bureaucracies and for monitoring aid flows, the binary approach groups together countries facing very different types of vulnerabilities, and does not distinguish the wide range of issues which may exist locally within a country and/or across its borders. The risk is that binary models create binary responses, with a template approach to fragility deployed across what are very different contexts. The reliability of the indicators and data that determine whether a country is on or off the fragile states list should be of concern given the significant implications they have for the kind of response they might subsequently trigger.

Finally, we cannot ignore the politics of the fragile states list. That some countries are powerless over whether or not they are on it is seen as but one reflection of their wider powerlessness in the international system. Meanwhile, non-western powers suspect that the list – and the very notion of fragility – may serve to circumvent norms of state sovereignty. Legitimate or not, these perceptions are not inconsequential: they make the whole exercise look like an OECD-owned process, which in turn shrinks the space for global cooperation to address the root causes of the interlinked development, humanitarian and security crises the world is confronting today.
The new ‘states of fragility’ model addresses some of the weaknesses of its predecessor. Assessing countries against five dimensions of fragility – violence, justice, institutions, economic foundations and resilience - moves us away from binary thinking and allows for more nuanced conceptualisation and, hopefully, policy responses. Both Seth Kaplan and Frauke de Weijer have rightly suggested that the indicators used under each of the five dimensions could be improved. Significantly, both have argued that key issues are missing from the dimensions, including transnational stress factors, social cohesion and horizontal inequalities. Nonetheless, with some further discussion and analysis followed by some tweaking to the model, these issues could likely be addressed.   

Monitoring fragility through the SDGs?
What is most welcome about the new model is its universal focus. In being an analytical tool that can be applied to all countries, it brings the OECD’s thinking more into line with that of the 2030 Agenda and the new Sustainable Development Goals (SDGs).  Critically, this means accepting that, while some will need special support, all societies could be more peaceful, just and inclusive. Aligning with the universally-agreed framework could be one way to get around some of the political sensitivities that past OECD approaches have been restricted by. And one could ask why, in the context of agreement by all countries to a new common framework, the OECD would need to revise and maintain a separate conceptualisation and framing of fragility.

While the targets in the SDGs don’t cover every issue that is important for fragility, they do cover the vast majority, including transnational factors and inequalities between social groups. So once the framing of the dimensions of fragility are agreed, the next step towards alignment with the SDGs would be to cluster the most relevant targets underneath them. We have done this with a focus on peaceful, just and inclusive societies, clustering 17 targets around what we see as 9 dimensions of peace.

Given that every SDG target will be tracked with one or two global indicators – along with some additional thematic indicators - they could be the basis for a fragility monitoring system, perhaps turning targets green when a country scores above average, yellow when it hovers below, and red when it lags dangerously behind. More red lights would signal greater risk. Disaggregation of indicators by region and social identity would allow us not only to generate an even more nuanced vision of where vulnerabilities are located sub-nationally, but also to identify horizontal inequalities, which will add significant analytical strength to the model.

There are some fairly significant obstacles in the way. First, the global SDG indicators are in the last stages of development, and whether they measure the right things is still open to question. Second, whether one or two indicators can effectively monitor a single issue is questionable. Third, some decisions on the disaggregation of indicators will be made at national level, potentially restricting their usefulness for differentiating between social groups or regions. Finally, data availability – especially in conflict-affected states – will remain a huge challenge. This means proxy indicators, including composites, may need to be relied upon while these hurdles are overcome through concerted support for the creation of strong data ecosystem.

From measuring fragility to formulating a response
Well thought-through indicators and new data are good ways to make the right priorities visible – but we also need a conversation about how we improve our responses to them, looking beyond aid. Monitoring tools are not short cuts to more effective policy. Putting aside the fact that quantitative assessments are always going to be limited (along with the accuracy of our data), the bureaucratic impulse to measure, categorize and compare needs to be resisted when it is used to dictate what engagement looks like on the ground.

The latest OECD report may already be making this mistake, suggesting that “Comparing actual aid flows with the characteristics of fragile environments could also improve the monitoring of aid, by offering a more nuanced and meaningful assessment of the extent to which aid is targeting the greatest needs in each country.” The report goes on to lament that “just 4% of ODA to fragile states and economies was allocated to the PSGs [Peacebuilding and Statebuilding Goals] for legitimate politics, 1.4% for security and 3% for justice.” However, while exclusionary politics, insecurity and injustice are indeed key drivers of vulnerability, they are not challenges primarily solved by outside financing.
How the international community responds to drivers of fragility will differ from country to country; we must constantly beware generalized prescriptions that identify things outsiders can control – like aid flows – as solutions to what are highly complex problems primarily shaped by domestic actors. So while a new approach to fragility will be vetted by the OECD members and used by their agencies, it is important that it is not conceptually constructed around them. This is what partner countries are owed if we are to assist them better to escape poverty and insecurity. 

The views expressed in our blogs remain those of the authors and do not necessarily represent the views or policies of the OECD or its members

Thomas Wheeler is a Conflict and Security Adviser on Saferworld's Policy Team where he focuses global conflict prevention policy.