One of our most interesting study findings was that most government officials perceive the greatest demand for data outside the government. In other words, most of the government staff which collect, produce and use data see external users – such as academics, NGOs, and the private sector – as their primary data clients. Why? Because these users frequently ask government officials to give them data. Obvious unmet data needs exist outside the government. In posts one and two in this series, we explore what it takes to make data more useful for government officials. To wrap up, we discuss how the development community can help the government make its data more useful for everyone else.
How does the government share data?
First, we note that government officials did not mention difficulties accessing or finding relevant data from other government agencies. Most reported sharing data via informal mechanisms. For example, if someone in the Ministry of Education needs health data, he or she calls a contact in the Ministry of Health. But by all accounts, this doesn’t happen very often anyway – most government staff only collect, aggregate and report data in their own sectors.
While government officials don’t seem to struggle with these informal mechanisms, external users do. Some data is available publicly via the CBS website or sector websites, but many users report frustration with the format, timeliness and coverage of this data. As a result, people who can go directly to government contacts when they need data. But this approach obviously leaves a lot of unmet demand. So if the greatest perceived government data demand is outside of the government, then what can be done to make government data more accessible to external actors?
What does “accessible data” actually mean?
Data accessibility refers to two things: how easy a user can access the data (e.g. find it, download it), and how easy the user can manipulate the data (understand, join up and analyze it). The idea of accessibility is simple in principle, but our study indicates that it’s rather complicated in practice. In particular, government data accessibility is limited due to available staff time – not having sufficient time/ability to post data in various different formats – as well as by physical resources (e.g., limited server space). While serious, these barriers could be addressed rather directly via government or donor prioritization of resources for such important initiatives. With some internal and external support, the government could make their data resources more widely and easily available to the general public, feeding growing demand for government information from academia, donors, civil society, the private sector and even other government agencies.
Apart from data infrastructure improvements (e.g. procuring more server space), accessible data also refers to the format in which data is posted and available. The following are but a few recommendations of how data formats can help to shape to a more reliable government data environment:
- Post data in .csv or other soft-copy formats – data that meets the criteria in standard definitions of “open data” – such that users can work with these data products and produce others
- Post more historical datasets, so as to allow for adequate comparisons over time
- Provide release schedules and regular updates to help further the quality of data comparisons
- Ensure that clean, high-quality datasets are made available to help promote further analytical products that are based off of sound datasets
Governments should use data to make resource allocation more effective. But governments also have a responsibility to make data available to the academics, companies, CSOs and donors that also need data to improve development decision-making. Direct support to help government make data more clear, open and accessible is an important step for enhancing the development data environment in Nepal.