Image by Ghullam Abbas Farzami / World Bank
  • Report
  • 19 September 2019

Failing to reach the poorest: Subnational financing inequalities and health and education outcomes

This report from DI and ODI looks at how effectively subnational financing targets regions with the worst health and education outcomes.



Amy Dodd, Marcus Manuel, Zach Christensen

Investments in health and education play a vital role in improving lives, reducing inequalities of opportunity and outcome, and developing a country’s human capital, which contributes to broader economic benefits. But, as the latest UN report on the 2030 Sustainable Development Goals (SDGs) notes, access to quality healthcare and education is not equally distributed across the world. And, as previous research by DI and by the Overseas Development Institute (ODI) showed, sufficient financing is not being mobilised to address these global inequalities.

Previous research conducted jointly by ODI and DI found that subnational financing is not well targeted on the poorest regions. Access to quality, affordable or free health services and education is vital. In some countries, the number of teachers and health workers in the very poorest areas is as much as ten times less than the national average, increasing the risk of leaving the poorest behind and continuing inequality.

This paper builds on that research, assessing how well subnational financing of health and education in all 82 low- and middle-income countries (including all LDCs) targets regions with the worst health and education outcomes. Many countries do not publish sufficient recent and comprehensive data, and analysis was possible in only 16 countries.


  • There is little evidence of a progressive allocation of health and education resources by governments to the regions with the worst health and education outcomes. Instead, in most countries, the trend is for slightly less financing to be provided to areas with the worst outcomes – typically 12% below national average for health and 7% below for education. The consistency of the findings across different indicators – whether under-five or under-18 mortality, years of schooling or secondary completion rates – suggests that poor and/or inconsistent targeting is a broader problem.
  • There is strong evidence that donors do target aid for health to regions with the worst under-five mortality rates – typically 29% higher. Results are more mixed in terms of under-18 mortality rates.
  • There is no consistent trend in donor targeting of aid for education. There is strong positive targeting in some countries, and the reverse in others. The evidence that governments that do publish data do not target the neediest is all the more concerning as those countries that do not publish subnational expenditure data are likely to be even less concerned about targeting spending at the neediest regions.


If we are to ensure that we leave no one behind and build human capital for all, we will need to ensure that the worst-performing regions receive consistently more, not less, resourcing. The evidence from this paper suggests that there is a clear case for:

  • governments and education donors to learn from why donor aid for health seems to be so much better targeted;
  • government and education donors to commit to improve the targeting of their financial resources;
  • at a minimum, ensuring at least equal allocation of government spending between regions, and;
  • above-national-average spend for education aid in worst-performing regions.

Better targeting on its own will not eliminate inequality. But it is a necessary and enabling precondition for much faster progress.


The report was authored by: Amy DoddMarcus Manuel and Zach Christensen

Photo: Ghullam Abbas Farzami/World Bank, 2014. Chemistry experiment at a school in Herat City, Afghanistan.