Donors and a Data Revolution

February 11, 2014

The High Level Panel on the Post-2015 Development Agenda calls for a “data revolution,” a new international initiative to improve the quality and scope of statistics and information available to citizens and policymakers. Such an initiative is particularly needed in sub-Saharan Africa where core statistical products like censuses, vital registration systems and household/firm surveys are not yet routine or complete, and where data accuracy, timeliness, relevance and availability are perennial problems. (See our own work on this issue here, and a good summary of the problem here)

There have been many “data revolution” meetings and conversations at all levels, and a number of new projects set to test post-2015 goals at country level from a data perspective. Yet one issue has remained outstanding: we’ve collectively avoided taking a hard look at the role of donor funding and spending restrictions in the limited progress on better data in Africa.

On funding, donors don’t spend enough, and countries don’t make up the difference. When they do spend, the amount fluctuates wildly year-to-year.

Aid for statistics comes to a measly 0.16% of total ODA, according to PARIS21. Of this amount, about 27% or $100 million in total went to African countries in 2013. These donor commitments to statistics are volatile, perhaps reflecting the support of one-off data collection efforts.  And yet, Ethiopia and Malawi for example both planned to fund over 80 percent of their total budgets from outside donors, while Tanzania and Kenya receive 54 percent and 36 percent respectively from donors. In many countries, nearly all core data collection activities are funded by external sources

African governments aren’t picking up the slack. Liberia estimated a funding gap of almost US$ 23 million between 2009 and 2013.  The budget for Nigeria’s Federal Office of Statistics demonstrates minimal (if any) relationship between the budget proposed and the actual amount received—and in one year no budgetary capital beyond salaries was provided at all. Nigeria’s national databank faces similar budget constraints; the program received less than half of its requested budget each year between 1999 -2003, and received no funding for two years during that period.  Six African nations are reported to have provided no budget support at all to vital statistics registration, while 23 more have been identified as having an inadequate budget for performing these activities.

If donors want better data, they’ll need to recognize the magnitude of funding gaps, create stronger incentives for country financing, and increase and smooth funding to national statistical systems.

On spending, donors’ restrictions on how money is spent don’t help with good data either.

Donors routinely spend millions for micro-oriented survey fieldwork and one-off impact evaluations while core statistical products like censuses and vital statistics go undone for years.  According to a recent UNICEF report, only 60 countries in the world have complete vital registration, and none of these are in Africa. This means that routine administrative data that is the basis for day-to-day funding allocation decisions remains inaccurate and unchecked. Further, donors don’t like paying salaries, instead paying for per diems, computers and field work for specific surveys.

As my colleague Justin Sandefur notes here, government statisticians earn per month what external consultants earn in a day. Increasing take-home pay by chasing donor-funded per diems via workshop attendance, training and doing survey fieldwork is the order of the day, and it is not surprising then that core national statistics products and quality are not a priority.

If donors want better data, they’ll need to fund national statistical systems differently – prioritizing core statistical products, and supporting national statistical organizations in ways that empower them to recruit and retain qualified staff. This doesn’t mean abandoning the special surveys and evaluations, but it does mean making sure that the core statistical products aren’t forgotten in the process.

What Next: Donors for Better Data

Donors need to fund more and differently, but how to start? Defining shared metrics for “good data” – that is accurate, timely, relevant and available – is a first step.  Tying progress on those metrics to increasing and flexible funding is a promising second step. 

For instance, more flexible donor funding in a set of countries could be tied to annual improvements in the World Bank’s Bulletin Board on Statistical Capacity score. The Bulletin Board could likewise be enhanced by adding metrics on accuracy and transparency of data, transforming existing ‘yes/no’ indicators into continuous variables, and omitting more administrative measures.  Or our own work on data discrepancies –where administrative data is contrasted to household survey data – could serve as a measure of accuracy in some sectors.

Alternatively, donors could link increased funding to progress on coverage and accuracy of core, under-supported products such as vital statistics. There is precedent for this approach; in a survey of donor practices, my colleagues Alan Gelb and Julia Clark found several cases where donors funded the expansion of national registries on a per capita basis. Further standardized tools exist that would allow for evaluation of vital statistics and cause-of-death statistics on an on-going basis (see for example:

Of course, there’s much that African countries themselves need to assess and accomplish to improve data, and we look at some of these issues in our forthcoming report on Data for African Development.  But growing momentum around a ‘data revolution’ provides a not-to-be-missed opportunity for donors to reassess how they engage around data in Africa.  By tying progress on data to increasing and flexible funding, donors could create incentives for country’s to fund their own national statistical plans and priorities. And donors get their cross-nationally comparable post-2015 MDG measurements in better shape too. Most importantly, it will help solidify the underpinnings of a true data revolution that can be led and sustained in the region. 

Thanks to Justin Sandefur and Jenny Ottenhoff for comments. 


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.