Data revolution from the bottom-up

Written by Amanda Glassman, Director of Global Health Policy and a senior fellow at the Center for Global Development.

Data revolutionaries around the world (myself included) are using every forum possible to call for more and better data that is disaggregated, produced more frequently, more open and more useable.  Recently, my colleague Alex Ezeh at the African Population and Health Research Centre wrote me: “We cannot address data system challenges in Tanzania or Nigeria by holding high level meetings in New York or London.”  He’s right: The path to more, better, timely and open data starts with strengthening country governments’ core data collection, analysis and use, whether it’s routine economic statistics or sustainable development goals. Country action should drive the revolution, bottom-up not top-down.

But what exactly might that mean? And how is it different from what we’re already doing? Last month, I attended a meeting that tried to spell out how the data revolution might be realized. My small group was focused on accelerating progress in country systems, and we identified four areas of actions: (1) creating post-2015 quick wins; (2) modernizing National Statistics Development Strategies (NSDS); (3) launching country compacts for better data; and (4) empowering new collaboratives to deal with sticky multi-country issues.

Here’s what we envisioned for each:

-SDG quick wins: Some have advocated more, larger, omnibus household surveys to measure SDG baselines. This is unrealistic; it takes at least two years to design and field a household survey and 2015 – the baseline year – is just around the corner. Instead, our group identified several quick ways that could strengthen existing data so that it could be a baseline for the post-2015 goals, such as: supporting documentation of data sets, particularly administrative data; geo-referencing existing survey and administrative data, mapping key indicators and services at relevant political levels, as has started to be done by the World Bank and Paris21; analysing and visualising existing data more creatively; creating inter-operable existing data sources, perhaps linking survey and administrative data; enhancing usability, accessibility and affordability of existing data sources; and developing and deploying analytical techniques such as small area estimation to produce estimates of marginalized or uncounted groups.

-Modernizing NSDS: While there are some notable examples, NSDS too often take years to develop and are only weakly linked with users and available budgets. Paris21 has helpfully developed new NSDS guidelines, which is a good start. Here are the group’s ideas: reflect the data revolution and the post-2015 agenda in NSDS; modularize and sequence NSDS preparation and implementation; reform legal frameworks where necessary; develop sector-specific plans that reflect both survey and administrative data; and create opportunities for non-governmental and media tracking and feedback on progress.

-Country compacts: Better data requires high-level all-agency commitment. Building on CGD and APHRC’s working group report, our small group recommended a political, financial and technical compact around NSDS goals. But if the NSDS is intended to mobilize resources and coordinate external funders, on which many developing countries’ statistical systems rely, then it’s time to imagine and support a more flexible pragmatic plan that is directly connected to funding. Compacts could be imagined as a souped-up country-level partnership – led at the Presidential level – that would bring together the government, donors, civil society organizations and the private sector to contribute intellectually, financially and technically towards improving national statistics. Compacts could be organized around mutually agreed (modularized) goals on what progress towards good data looks like, disbursing a portion of resources and creating visibility for progress on coverage, quality/accuracy, timeliness and openness of prioritized data. A compact could also provide more flexible, longer-term funding to national statistical systems that would be disbursed against progress on mutually agreed goals in the form of challenge grants. This type of agreement could help leverage commitments from donors while also guaranteeing greater sustainability by requiring counterpart financing from the government.

-New or ramped-up collaboratives:New or ramped-up collaboratives focused on funding or technical support would help address particularly challenging issues, such as improving statistics in poor and small countries, and addressing the need to rebase GDP in many countries. The group also expressed interested in establishing a civil society funding window for data analysis and use to influence debate and policy (developing “intermediaries” to make data digestible), and with a role in “validating” headline statistics. A final area that requires more effort: a more intensive International Household Survey Network that goes beyond collecting surveys and works towards coordination and harmonization of questionnaires and data collection efforts, always with due consideration of country series.

Together, these actions would go a long way to a data revolution, starting from where it counts – the bottom up.

If you’re interested in getting involved in these activities, please email

3 Comments on "Data revolution from the bottom-up"

  1. A-1,FLAT NO-507,MODI BUILDERS,QUTUBULLAPUR,HYDERABAD,INDIA-PIN-500054 | August 18, 2014 at 5:59 pm | Reply

    Data structure data analysis is imperative,how ever the data requirement changes as the country progressess,and the society changes.Also humans prosper from poverty to richness,if they are in services,or if they are professionals.Hence data base,data up-date are a constant things.Yesturday we might have needed how many are with out a motor bike,but today we like how many cars are entering each area so that the roads can be rationed for traffic.Future tools of use could be more night mobilty for short distances by AIr.The Jargon of Motor manufaturers syndicate may be broken with more advanced technologies,by use short air strips,two seater crafts,helicopters,and small flying cabs,etc.
    Hence data aquisition,and assimilation changes over periods,and is a constant requirement.

  2. Extremely high quality Information, and I fully understand what your article was trying to achieve. The purpose of this type of information, as to educate people more fully, with the believe that anyone can drive themselves forward for better educational grades and respected brilliance

  3. The paper condenses all the main issues.

    The PARIS21 NSDS guidelines, by construction, are about the national process of designing an NSDS that would fit the situation and the choices of the country at the time the process is launched; the resulting decisions that make the product of the NSDS design process (the NSDS) are country and time specific, a point in a long term perspective; any compact is therefore an output of the design process. One NSDS only is not going to solve all the problems so a sequence of NSDSs is necessary to progress along a path chosen toward a nationally agreed vision.

    In a genuine bottom-up approach, supporting countries is about assisting them to decide and implement their decisions, not about force feeding with any of the best or advanced practices of the time in data or statistics management. Heads of national statistics units are generally high level technicians, and most of them are very well aware of all the main issues regarding development in their national environment. The major role of external experts is to assist in the delivery of the NSDS: maieutics?

    What external assistance should provide is untied financial resources (preferably grants) but probably more importantly is knowledge that would help countries to best adjust their decisions to the foreseen environment at the time. A wiki populated by all data development partners and presenting that knowledge would be a natural complement to the PARIS21 NSDS guidelines.

    A very critical point is the synchronisation of the NSDS design process, the national development programming and the decision making process about Post-2015 development goals and indicators. It would be rational to revise ongoing NSDSs or adopt new one by the end of 2015 or early 2016, when the countries have collectively endorsed the proposed International goals for development and the related indicators needed for monitoring at national, regional and international level. The PARIS21 reports on country progress in designing and implementing National Strategies for the Development of Statistics (NSDS) could help programming assistance to countries in adjusting their statistics development to the Post-2015 national and global development strategies.

Leave a comment

Your email address will not be published.