Marta Foresti heads the Overseas Development Institute's Politics and Governance Programme.

Sadly, we know a great deal about violence and conflict. Yet according to recent excellent research on subnational conflicts by the Asia Foundation, when it comes to the relationship between conflict and development we often make assumptions that are not always borne out in practice.

National wars are no longer the most common form of conflict. In Asia, local or subnational conflicts are by far the worse form of conflict, affecting half of the region's countries. They are also the most enduring: with the average lasting 45 years, they are among the oldest active conflicts in the world. Finally, and most importantly, they are the most deadly: from 1999 to 2008 more people died in subnational conflicts in Asia than all other forms of conflict combined.

What is perhaps less known is that development does not mean peace. The analysis of conflicts such as Mindanao, Aceh or southern Thailand challenges much of the conventional wisdom on what causes violent conflict and its relationship with development.

In most cases, subnational conflicts occur in areas with functioning, if sometimes contested, systems of government. Not only that, conflicts mostly occur in middle income countries rather than in so called 'fragile states' and often in regions which are not particularly poor, where levels of literacy and infant mortality are remarkably similar to those in non-conflict areas.

The figure below compares average national literacy rates (red line) in selected Asian countries with average rates in regions affected by conflict (blue line) within these countries: in most cases there is no significant difference between national literacy rates and literacy levels in conflict-affected regions. In some cases, literacy rates are even higher in regions affected by conflict than in the country as a whole (eg. Shan and Rakhine in Myanmar, Moro in the Philippines).

Literacy rate: comparison of national average and subnational conflict average (Source: Asia Foundation).

In these countries, issues of inequality, discrimination and identity rather than poverty are often the drivers of sectarian violence and conflict. In many ways, these were also the factors underpinning recent violence and conflicts in relatively stable African countries such as Mali and Kenya.

Yet it is often business as usual for aid donors, who continue to focus on traditional projects for poverty reduction and access to services. Many projects are justified on the basis of 'contributing to peace' but evidence suggests that, at best, development alone does not guarantee peace. At worst, aid risks doing harm by exacerbating existing drivers of conflict or by reinforcing local power dynamics and discriminatory practices.

There is a broader lesson to be learnt from this analysis of Asian subnational conflicts, namely that failing to understand the politics is often at the heart of many types of development failures.

For example, there is growing consensus that current approaches to promote institutional reforms often fail due to a lack of understanding of the political dynamics underpinning them. Similarly, efforts to improve service delivery in different sectors fall flat unless they are able to combine sound technical knowledge with an in-depth understanding of the political and governance blockages that need unblocking to achieve change.

Development is political and getting the politics right is key if aid is to make a difference. We often hear that this is easier said than done but the evidence emerging from the Asia Foundation and other research suggest that the road ahead is clear, if we are brave enough to walk it.

First, it is not enough to say that 'donors need to think and act politically'; we need to spell out what it means and how it can be done. How about focusing on these three steps?:

  • First, do no harm. This requires more and better analysis of initial conditions, including systemic features and incentives underpinning local power dynamics as well as on-going monitoring.
  • Second, prioritise adaptive processes over predetermined outcomes.
  • Third, invest in building networks and coalitions at local level, if necessary through brokers or intermediaries who are often better positioned to address issues of marginalisation through mediation.

We know how it can be done, what we are less clear about is whether donors are best placed to work in this way.

We need to reflect on current thinking and policy priorities on conflict and development. The current mantra is to increase spending in fragile states, based in no small part to research and analysis suggesting that this is where most poor people live or will live in the near future. The evidence from Asia suggests that many states not classified as fragile may actually experience severe, though localised, conflicts. Thus a narrow focus on fragile states might risk excluding large numbers of people affected by conflict.

Conflict dynamics have moved on, and aid approaches need to catch up.