The trouble with targets: what would happen if we won all our campaigns?

May 4, 2011

     By Duncan Green     

recent briefing, Jessica Hagen-Zanker and Anna McCord at ODI ran the numbers for five countries from sub-Saharan Africa (Ethiopia, Kenya, Malawi, Mozambique and Uganda) and came up with some worrying results. ODI Africa social spending targetsFirst they identified six sectoral spending targets (above). Then they looked at the budget documents of the five countries to examine actual performance against the targets in 2006/7 (the most recent year with comparable figures). Here’s what they found (click on the table to expand). ODI Africa social spending targets 2And their summary: “Total government spending (including on-budget ODA) varies significantly, ranging from 20% of GDP in Ethiopia and Uganda, to around 24% of GDP in Kenya and Mozambique, and a high of 27% in Malawi. Government sectoral expenditure (including on-budget ODA) was not high enough in any of the five case study countries in 2006/07 to meet the social protection, water and sanitation, or infrastructure targets. Table 1 shows that only Malawi and Ethiopia met the 10% target for agriculture, and only Malawi met the 15% target for health. Ethiopia, Kenya and Mozambique managed to achieve (or exceed) the 20% target for education, but overall, only seven of the 30 different country sectoral spending targets were met. In most cases, the shortfalls were significant. An analysis of 2006/07 data indicates that it would not be possible to finance the six targets simultaneously in any of the case study countries, given existing government expenditure. The total cost of all six sectoral spending targets as a percentage of government expenditure ranges from 98% in Kenya to around 120% in Ethiopia and Uganda. If all government resources were directed to these six sectors, four of the five case study countries would be unable to allocate expenditure to any other government functions, and would still face a sizeable target funding shortfall. Hence fully financing all six targets from existing (2006/07) budgets would not be feasible. The only options for meeting the sectoral targets simultaneously would be either to increase government expenditure (for example through increased revenue or on-budget aid) or through the utilisation of off-budget aid. However, as government expenditure increases, so too do the costs of the targets as three targets are calculated as a proportion of government expenditure, so the total cost of meeting the targets is itself a moving target. Government expenditure would need to increase by between 87% (Malawi) and 154% (Uganda) to meet the targets while also retaining current levels of expenditure on other government functions. [what about getting there by increasing aid?] Total on-budget ODA would need to double to achieve the sectoral spending targets in these six sectors. In the current environment of fiscal consolidation in donor countries, such a significant increase is unlikely in the short to medium term and even the maintenance of 2006/07 ODA levels may be a challenge.” So what should campaigners do – stop asking for the moon? Not necessarily – campaigners are not civil servants or accountants; if they are not pushing the boundaries of the feasible, they are probably not doing their job. And they are not in charge – they aim to be the antithesis in the dialectical process of policy change, not the final synthesis. But especially if they are likely to be successful, they should be aware of the cumulative impact of their demands, they should think more about how governments should enhance revenue, as well how they should spend it, and they should talk to each other more, rather than each go off and just invent new spending targets. Otherwise the danger is that both campaign asks and government promises become devalued currency – hardly the way to strengthen the social contract between citizens and state. But these are just preliminary thoughts – I’d be interested in hearing the reactions of advocacy wonks and campaigners on this one.]]>