20 September 2011

EADI-DSA Conference - Day 2: Does International Development Assistance Care About Poverty?

On Day 2 of the EADI-DSA Conference we had two excellent presentations from Ravi Kanbur and Sabina Alkire on "Rethinking Progress and How to Measure it".

Ravi kicked off by noting the stylised fact that poverty is now predominantly a middle income country (MIC) phenomenon, not a low income country (LIC) one (citing work by Andy Sumner of IDS). This is an artifact of the cutoff $1165 GDP per capita above which a LIC becomes a MIC. Indonesia changed status (from LIC to MIC) in 2008, India will in 2012/13 and China did in 1999.

Projecting forward, fewer and fewer countries would qualify for IDA and if IDA flows stayed at the same level some countries like Ethiopia would stand to gain.

He suggested there were 3 possible outcomes in IDA allocation: (a) no change in rules and increases in aid for the remaining LICS, (b) no change in rules, but a reduction in overall aid flows, and (c) change in the rules that keep current levels of IDA focused on poor people, whichever country they live in.

He thought (b) was the most likely outcome, but (c) was the right outcome.

He argued that the moral responsibility of the global community to poor people was analogous the Responsibility to Protect and he proposed the following operational rules for IDA qualification:

(a) keep the standard window at $1165 GDP/cap
(b) add a window that is 2 or 3 times of this threshold, but target it to poorest regions or key human development sectors, and
(c) have a third window to support the production of global public goods

He acknowledged that these proposals are unlikely to be adopted: the LICs will lose out, the MICs who will gain are too different to form effective alliances to argue for the rule change (and in any case, some of those within MICs are happy to graduate from IDA) and it is in donor interests to reduce overall aid flows and the current rules will help with this.

Sabina Alkire's presentation on the multidimensional poverty index (MPI)confirmed the conclusion that only 25-30% of the world's poverty is now in LICs--no matter how you measure poverty. She talked about how difficult it is to get crude measures of progress, like GDP/cap, "dethroned" (as Dudley Seers had put it) because of vested interests and noted how little money is spent on measuring poverty as opposed to expenditure on data measuring economic performance and employment. She presented data which show substantial deviations between $1.25 a day estimates of poverty incidence and multidimensional poverty incidence for some countries (in Ethiopia, multidimensional poverty is almost twice the income poverty rate).

Questions from the floor included: Aren't you assuming that IDA has a positive effect? Does this imply a vision of a global welfare state? How do we include multiple voices in multidimensional poverty measurement? Does IDA undermine Indian citizens' incentive to hold the Indian government to account and increase tax revenues? What about the politics of measurement?

Sabina acknowledged the politics issues and said she was not sure the MPI would survive as a cross-country index. She highlighted a number of possibilities: (a) it would be seen as a good idea that nevertheless peters out, (b) good multidimensional poverty measures would be collected and used to improve analysis and policy, (c) complex or inaccurate multipoverty measures would be collected, and the whole initiative would be discredited with a return to simplistic measures, or (d) conflicts would rage between different measurement schools with no consensus. She thought (d) was the most likely. Despite this she did cite several examples of where governments have adopted the measure as their own and have begun to use it.

Ravi responded by saying that lack of commitment of a government to do anything about poverty could also be a condition for receiving IDA--and in fact this might promote civil society to track the commitment of their governments to reduce poverty (this reminded me of the work from Dolf te Linto and colleagues from IDS on measuring the commitment to reducing hunger). He also said that unless IDA has a zero or negative effect on poverty, that his argument that IDA should focus on poor people rather than poor countries still holds.

A very good session, but I was struck by the resignation of the two presenters that the current measurement powers would prevail. Desmond McNeill pointed out that we need more work on the political economy of measurement--something I would wholeheartedly agree with.

No comments: