Defining the use of data in policy making is contentious, and attempts to measure it are sparse. However, enough evidence suggests that more work is required to reinforce data use in policy making. For example, a 2018 evaluation of the World Bank’s support for data and statistical capacity notes, “weaknesses in promoting data use have been a major issue for the past 10 – 15 years”.

Not only is uninformed decision making common, but it is a waste of resources not to use data that are expensive to produce (and generally, they are[1]) and, in many cases, at the expense of resources being put to use for other priority areas. Also, by not using data, policy making is akin to flying blind!

To understand how to improve the use of development data in public policy making, we would first need to explore why it is that data are underutilised. To start, here are a few apparent reasons that come to mind:

• Data are missing
• Data are not fit-for-purpose
• Data are hard-to-access

One of the obvious barriers to the use of data is that the required data could be missing altogether (for whatever reason). But even when data are available, if they do not meet the timeliness, granularity requirements and methodological rigour, they may not be fit for purpose and consequently not get used. In addition, difficulties in access could also be why potential users may not use the available data. But when the required data are available and easily accessible, the following could still hinder its use.

• Limited data literacy
• Perverse incentives
• Mistrust in data

Weak or lack of data literacy, limiting analytical capabilities on the part of potential data users, may adversely affect the use of data. In addition, perverse incentives could also lead to the outright rejection of politically inconvenient data. And mistrust in the data (not necessarily due to poor quality) may also negatively affect its usage. Furthermore, there could be some cross-cutting institutional issues too why data use may get obstructed, such as:

• Siloed data
• Inadequate financing for development data
• Misaligned priorities

Data in isolated and siloed information systems across the public sector pose multiple challenges to their flows and uptake. Also, inadequate investments in the data ecosystems perpetuate a vicious cycle in which governments and development organisations invest too little in data systems. In turn, the situation results in poor data (affecting data use) that warrants further investments. The cycle continues. In addition, it is possible that global and local development objectives may not always align and are sometimes in conflict, ultimately affecting data use.

So, what can we do to improve data use?

We can consider mitigating actions by strengthening both the ‘demand’ and ‘supply’ sides of data use. On the demand side, we could think of the following:

• Improving the data literacy of policy makers
• Fixing the incentives

Policy makers may have different levels of skills in appreciating data and statistics. Hence the first step could be to assess data literacy needs at various levels of the decision-making machinery. The assessment could include capturing factors around understanding data and using it to inform required actions to strengthen data literacy. Also, with more precise knowledge of what level of communication would suit different data users, data producers too can identify and reinforce relevant activities to produce data and ensure use.

Public scrutiny and debates on government policies can stimulate data demand and use. Hence, strengthening the data literacy of humanitarian, civil society organisations, journalists, and citizens could be helpful. Also, creating a sense of competition among policy makers by comparing public policies could be instrumental in aligning incentives.

On the supply side, we could think about the following:

• Producing demand-driven data which are timely and adequately disaggregated, known and easily accessible and understandable

Production of demand-driven ‘fit-for-the-purpose’ data is vital. Data must be known (their existence and adequacy) and trusted. They must be accessible (such as through easy-to-use data portals) and understandable (such as through data visualisations) to reinforce use when needed. Also, promoting the reuse of existing data could free up valuable resources.

Besides demand and supply, stress must also be on strengthening governance issues of data ecosystems, such as:

• Measuring data use
• Strengthening the culture of evidence-based decision making
• Ensuring adequate funding is made available to bolster the data ecosystem (including an eye to digitalisation)

Measuring data use could be a good start. Some proxy indicators (such as measuring references to statistics in national policy documents) could help. In addition, while governments must strengthen the culture of data-driven decision making, including prioritising data demand, adequate funding should also be made available to enhance the data ecosystem and infrastructure. Finally, promoting the reuse of existing data could free up valuable resources too.

With the knowledge of the above blockers and enablers, we could hope to improve the issue of inadequate data use in decision and policy making.

---

[1] A 2015 research from the Sustainable Development Solutions Network estimated that among 77 of the countries receiving the most aid, the average cost per household survey was US$1.4 million