By Pietro Gennari, Chief Statistician, and Dorian Kalamvrezos Navarro, Statistician in the Office of the Chief Statistician, FAO

A 2009 report to the UN Statistical Commission on the status of agricultural and rural statistics ominously predicted the following scenario: “with data not being adequately used, resources are cut, leading to a vicious cycle of a reduction in content and data quality.”

Fortunately, this doomsday scenario has not yet panned out. An unexpected silver lining in the COVID-19 pandemic has been a resurgence in the public’s fascination with numbers, an upgrade of statisticians’ seats at the decision-making table, and a political establishment that is more receptive to data and statistics than ever before.

The Cape Town Action Plan for Sustainable Development Data and the UN Secretary-General’s Data Strategy have echoed the message that the UN must better support data use.

Yet, the 2009 report’s warnings about the consequences of poor data use should not be ignored. Poor use of food and agriculture data and statistics has a detrimental effect on evidence-based policymaking, the accountability of national governments to the goals of eradicating hunger, malnutrition, and achieving sustainable agriculture, and business decisions by the farmers, fishers, and the other actors of the food system — all weakening a country’s prospects for achieving the SDGs.

Too often, poor data use has been chalked up to the poor quality of the data that are produced in the first place. This is a legitimate concern, particularly where sub-optimal methods of data estimation are used, or where traditional survey tools and sampling methods limit the granularity and timeliness required for providing actionable intelligence. However, data quality – or perceptions thereof – is by no means the end of the story.

A second key obstacle hampering data use is the fact that these data are not disseminated in ways that would facilitate their use. For instance, the vast majority of data on food and agriculture collected through national surveys or censuses remains locked in institutional silos, and is often made available only in paper or PDF formats that limit or prevent re-use, particularly by the research community.

Such “data graveyards” stem from a weak dissemination culture pervading National Statistical Systems, as well as a misunderstanding of the concept of data protection and data confidentiality. An important tool to help avoid such situations is the recently unveiled FAO Food and Agriculture Microdata (or “FAM”) Catalogue, which aims to facilitate a wider dissemination of agriculture census and survey microdata[1]. 

A third key factor is a general malfunction of the data ecosystem. This malfunction has variously been described as a “market failure” between demand and supply, a “broken link” in the data value chain, or a “stunted feedback loop problem” between producers and users of data. At one end of the data value chain are data producers that struggle with collecting, curating and disseminating data. At the other end are data users that may want to apply the data to make informed decisions, but may not be doing so because they mistrust the data, lack access to the data in useful formats, or lack the basic conceptual tools for properly interpreting the data.

Tackling these issues and promoting data use is one of the greatest challenges faced by statisticians working either in national or international organizations. This was recognized by the Joint Inspection Unit’s 2016 Evaluation of the UN’s statistical capacity development support for the MDGs, which concluded that the UN does rather well in supporting data production, but less well in the “difficult task” of supporting data use, which should therefore become a priority. Several global strategic documents, such as the Cape Town Action Plan for Sustainable Development Data and the UN Secretary-General’s Data Strategy, have since echoed this message.

Why is it harder to support data use than data production? While data producers consist of entities that are easy to identify and target, data users are a scattered, heterogeneous group, and are often isolated from decision-making processes. Moreover, there is a persistent communication gap between data producers and data users, in general, and more particularly between the statistical community and the policy-making community.

Thus, a tailored approach is needed for targeting specific data user groups, with a particular emphasis on the middle stages of the data value chain. The middle stage is where various data intermediaries – such as civil service analysts, journalists or academics – take the data coming from data producers and churn them into actionable information through analysis and research.

Data use by such intermediaries – (“infomediaries” as coined in the World That Counts report) – can be stimulated in a number of ways. For example, the 50×2030 Initiative, a partnership between FAO, IFAD and the World Bank aiming to strengthen the capacity of 50 low-income and lower-middle-income countries to improve agricultural sector data, recently launched a data use grant competition for researchers in five developing countries.

But as long as countries wish to maintain a crisp division of labor between data producers and intermediaries, we may never be able to fully repair the broken link in the data value chain. The data published by data producers will continue to come out too raw even for savvy infomediaries to digest.

A greater effort is therefore needed to bolster the analytical capacity of National Statistical Offices. Most NSOs still believe that their main mandate is merely to produce data – not to analyze and present them in ways that are accessible to other users – and they generally lack the capacity to disseminate key messages to inform policy decisions.

NSOs need to break their traditional confines of disseminating figures and statistical yearbooks to also become data intermediaries themselves, by producing analytical reports, headline indicators, and press releases with policy-oriented key messages, all of which add value to underlying data. They also need to be brought into the fold of the policy-making arena so that the type of reports they produce – and the timing of release of these reports – are in tune with contemporary political debates, while retaining their institutional independence and their essential role as trusted, impartial providers of statistical data.

International agencies stand ready to support NSOs in this journey. This includes helping NSOs embrace open data, deal with data protection and privacy in a way that also allows for data dissemination, and establish platforms that can facilitate communication between data producers and data users.

International agencies are also beginning to integrate data literacy and data analytics training components into large statistical capacity development programmes. For example, a recently issued Guide to Promoting Data Use under the 50×2030 Initiative lists a number of actions aimed at helping data producers better communicate and present data products, tailored to different audiences. Yet there is still ample scope for international agencies to join forces and work closer together on promoting data use and supporting the analytical capacities of NSOs, in the spirit of the UN Secretary-General’s Data Strategy and with a view to further supporting countries in achieving the 2030 Agenda.

[1] The International Household Survey Network (IHSN) has similarly already been disseminating household survey microdata for several years.