While the global development data community has been talking about the data revolution for one year, not everyone agrees upon what it means. For some, it has come to mean more open data to help citizens hold government accountable, or a major survey to build a baseline for the post-2015 goals. For others, it means a revolution in the way we employ technology for data collection, dissemination, and use. Finally, and particularly for PARIS21, it can mean increased capacity for national statistical systems – the institutions that still collect the vast majority of statistics on human development.
With the aim of furthering this discussion, PARIS21 in association with Center for Global Development and African Population and Health Research Centre, hosted a seminar on the morning of 2 April 2014, “Debating the Post-2015 Data Revolution: will technology or better institutions make more of a difference” This debate brought together statisticians, researchers, and other interested stakeholders to discuss the relative importance of these two critical aspects of the data revolution. For example, do we need more funds for statistical capacity, or should we encourage the use of innovative technologies like mobile call logs? While framed as a debate for the purposes of encouraging discussion, all participants recognized that there was no right answer – the question was how to find a balance between the two.
Starting off in support of institutions, Amanda Glassman, Director Global Health Policy, Center for Global Development, argued that we need to spend money on building statistical capacity, instead of focusing on “cute apps when the basics aren’t right.” The amount of funding available for national statistical systems (NSS) remains small, volatile, and is often tied to specific projects.
Trevor Fletcher, Project Coordinator at PARIS21, followed up in support of technology, arguing that institutions have failed us, and statisticians slow the process with their methodologies and quality frameworks. Fletcher suggested (though fully admitting he was arguing the extreme case) that NSOs be replaced with call centers and server farms, everyone be provided with a smartphone, and real-time links be created so we can collect data on everything we need at any time.
The voices from the floor in favour of institutions were strong, with one representative from a national statistical office (NSO) arguing that “the role of statisticians must be defended to ensure that science is behind statistics” and that there must be an equilibrium between absolute quality of statistics, and fit for purpose. A representative from a regional organization argued that many policy-makers unfortunately do not pay full respect to statistics when it comes to decision-based policy making, with many saying that “statistics are good, but they’re in the way”.
On the role of technology, participants were more cautious, suggesting that it is both useful and inevitable, but cannot be used effectively without strong institutions. One participant from an international organization suggested that technology can result in “data indigestion” as too much data can be overwhelming, and it is the role of the institutions to define how to leverage that data. Along the same lines, another participant argued that “technology is an irreversible trend, not only in statistics but in every field of development,” yet that NSOs eventually realize that technology needs a teacher, and as the technology develops more capacity is needed so that it can be used in a meaningful way.
Examples from countries suggested that technology is already being used, although the results vary. In Philippines, handheld devices were successfully used in the 2007 census, and agricultural statistics use mobile phones for gathering price data. However in Uganda, a halt was put on mobile phone trials as so many were being sponsored by different groups, that with a lack of standardization it was out of control. Finally, when Nigeria did a poverty study in 2010, the NSO came up with a poverty rate of 63%, which was perceived as too high by many people. A polling company using online surveys and telephones came up with a figure of 20%, which was perceived to be too low. As with other examples, this points to a need to find a balance between institutions and technology.
In conclusion, the participants agreed that statisticians are beginning to react to technology in a positive way. While some were initially threatened by big and open data, many are now adapting and seeking to determine how they can become brokers of this data – give it a quality stamp and make it useable. At the same time, they will have to adapt their own data collection methods in order to compete. This is, and will continue to be, a huge challenge for institutions.
Of course, enhancing the role of NSOs is vital for the production of rigorous statistics needed for democratic functions. If there's money available for household surveys and technology, there must be a way to transfer that money to national statistical systems in a way that allows them to invest in core statistical functions. It was agreed that we can’t have more demand for statistics and less money at the same time.
The debate was based on the assumption that both technology and institutions mattered to the data revolution. What became apparent over the course of the morning, through informed and lively discussion, was that the key is how they can best exist together. Without technology, national statistical systems will be left behind as the world moves on. However without effective institutions, new technologies are not sustainable or able to produce statistics that are considered robust. Ultimately the best approach, as with most things in life, is all about finding the right balance.