Tune in to the 2012 CAPE Conference to hear Andrew Rogerson speak on the future of aid (14 & 15 November 2012).

Unpacking the recent debate on where the world's poor live and where they will live in future pays interesting dividends in the present.

There are two basic stories in play. One of these accounts, associated with Andy Sumner, focuses primarily on income, and says that the big dividing line is between poor and middle -income countries. The other one, associated with Homi Kharas (and me), focuses more explicitly on politics: its dividing line is between fragile and stable countries. (Don’t take my word for it: this is how The Economist (1 September 2012) characterised them).

These stories have much in common, perhaps surprisingly. Firstly, in terms of the numbers, because so many poor are already concentrated in middle-income but fragile states like Pakistan or Nigeria, which are not predicted to make rapid improvements. So, the two accounts already overlap to a considerable extent.

Secondly, and more importantly, it is politics that drives both stories, albeit in slightly different ways. If we take ‘fragility’ as the dividing line, this highlights countries with chronically weak or missing institutions. These face formidable barriers of capacity and legitimacy. They stagnate while others with comparable endowments make more rapid progress; they cannot turn global opportunities to their lasting advantage. For fast-growing middle-income countries, a rather different political challenge is apparent: they increasingly have in their grasp the means to tackle poverty, but may lack the political consensus to do so.

These crossovers imply that a re-thinkingof both the whys and the hows of international aid is necessary.

Let’s start with the whys. Using the ‘fragility’ lens, the case for external intervention (always with costs – not purely financial) can be compelling: the prospect of tolerating long decay, collapse, weak recovery and frequent relapses is both hard to accept in humanitarian terms, and dangerous in terms of contagion in the region and beyond.

In more stable contexts, using the ‘income’ lens, the rationale for aid is more subtle and contestable. It involves often unstated, and maybe unrealistic, expectations of political influence on the part of outsiders, where the problem is attributed to absence of political will far more than of capacity. Few self-assured sovereign governments would welcome aid on the latter basis, and, money being fungible, they can usually neutralise attempts to earmark resources for priorities that do not jell with their own.

And what about the hows? The sophistication of institutions in stable environments means that, here, the track record of aid-funded investments and results is usually easier to substantiate. In fragile contexts, however, it is a different matter. As per-capita aid levels for fragile states rise, and as poverty falls elsewhere, it is time for serious assessment of the ingredients that make for lasting success in tackling fragility.

Although this certainly implies a lot more soul-searching by the international community in future, we should be thinking and acting on these whys and hows right now, not wait until 2025. The CAPE conference at ODI this week will be a good place to start.