International Development CommitteeSupplementary written evidence submitted by Andrew Dorward, Professor of Development Economics, School of Oriental and Africa Studies (SOAS)

As I mentioned immediately after the session, my comment that I needed three months to think about a livelihoods indicator was not correct, I am not sure what I was thinking of. As I said in answer to your initial question, a goal and indicator on livelihoods and income is in my view the best way to address “employment” as I think employment is really about the achievement of secure incomes. In our original submission to the Committee we suggested (in paragraph 15(ii)) a “food expenditure ratio” as a theoretically sound and practicable measure of food prices relative to incomes of households in different income deciles. It could also be described as a theoretically sound and practicable measure of real incomes, and for the first income decile would focus specifically on the incomes of the poorest 10% of households, allowing for their vulnerability to food price changes. It meets the holism and equity principles in its focus on the poor, its consideration of vulnerability to food price changes, and in its linking of real incomes to economic growth and to households’ potential expenditures on, for example, education, housing, and nutrition. Measurement and estimation could be improved, but rudimentary estimation is possible with current data sets. The measure is introduced in section 3 towards the end of the attached briefing paper (although it is introduced more as a measure of food prices than of incomes). (For the record, further details on the derivation of the measure and examples of estimates are available in annexes B and C of a working paper available at )

There were two other comments that I wished to make to the committee and did not manage to fit in during yesterday’s session.

1. Subsidiarity

This would/could allow some flexibility in opting in/out of particular goals. This might be seen as a disadvantage in letting governments “off the hook” on controversial but critical goals (like human rights or climate change). Alternatively, however, it could offer advantages in allowing flexibility in negotiating, and while it might appear to let governments “off the hook”, their (formal) wanting to be “off the hook” on any particular goal/target would itself be a major public statement with political and accountability implications.

2. Implementation targets

There was a final question regarding impact and implementation targets. My understanding of John McArthur’s response was that he was suggesting that there should not be implementation targets. I think that this needs to be unpacked a little, as it depends on what one means by implementation targets. In my view his answer was appropriate for a narrow definition of “implementation goals” and “implementation targets”.

In our submission, however, we used the terms “implementation goals” and “implementation targets” a little more loosely to distinguish between goals about fundamental objectives (such as improved literacy, knowledge and skills from education) and goals about service delivery and access as means for achieving impact (for example the goals for universal primary education with enrolment and completion targets). We argued that these different types of goals and targets had sometimes been rather mixed up in the MDGs. I would suggest the need for a distinction between impacts (desired changes in achievement of fundamental objectives), outcomes (results of actions), outputs of actions, and inputs (resources into activities).

These distinctions and the flow from resource inputs to activities to outputs to outcomes to impacts are standard parts of development policy and intervention planning and management (eg, inputs of finance and human resources etc. for school development activities to generate outputs of increased school places leading to outcomes of greater school enrolment leading to impacts of increased literacy and life skills). According to this way of looking at things, post 2015 goals should all be concerned with impacts (eg poverty, incomes, nutrition security, communication, health status, life skills, sustainable resource use and maintenance, population growth rates, etc). For many of these there will be useful “outcome” targets (eg school enrolment, km of roads per person, health service access and utilisation, etc). Such outcome targets should be measurable and there must be clear understanding of the requirements for their actually contributing to impact achievement (for example the need for quality, complementary resources and activities).

How these outcomes (and other outcomes) will be achieved and the resources required for their achievement should not be part of the international goals and targets (but they will be important in national and sub-national planning and in resource mobilisation and acquisition).

October 2012

Prepared 21st January 2013