Central banks report sharp improvement in data governance

Data governance at central banks appears to be improving, with a significant increase in institutions reporting to have it in place over the past year.

For the first time, the majority of central banks responding to Central Banking’s big data survey said they have a “clear governance structure” in place to handle data, including big data.

In the 2019 survey, 63% of respondents said they had a clear structure and 37% did not.

The figure marks a sharp turnaround. In the preceding three years, the proportion of central banks reporting a clear governance structure had gradually diminished, to a low of 38% in 2018.

Does your central bank have a clear governance structure for the management and collection of data (including big data)?

The pattern of responses may reflect the considerable efforts that many central banks have gone to in overhauling their approach to data. Many respondents to this year’s survey said they had completed projects or had initiatives under way to improve data governance.

Some central banks reported having structures in place on a departmental level, but said they were working towards enterprise-wide approaches. Many said they made use of “data stewards” – individuals nominated in each department to ensure data is collected and organised correctly.

Several central banks said they were working towards measures to harmonise aspects of data governance, with clearly defined rules for data owners, sources and producers, as well as dictionaries, metadata and taxonomies. Often, the IT or statistics departments were tasked with overall data oversight.

One central bank in Asia said it worked on the basis of “clear role division”, with different people in charge of collecting, processing, analysing and storing data.

What are your central bank’s main areas of focus when it comes to big data investments?

Some commented on the tensions inherent in getting data – and data scientists – to the right places.

“I think the central bank would benefit from a centralised ‘big data research unit’ that brings together staff with data science skills and big data,” said a respondent from Oceania. “But there is also merit in keeping people with data science skills dispersed throughout the organisation.”

A challenge for many was ensuring the right level of security, without excluding people with a legitimate need to access data. A central bank in Africa said it was working on refining the “authentication process” for users, aiming to give each person their own profile with the set of time series they were allowed to access.

Data is clearly a priority for most central banks. None of the 58 respondents said their data budget had decreased in the past year, and the majority (63%) said it had increased in real terms. Budgets were unchanged for the remainder.

Central banks report that improved software and human resources are key areas for investment. Similarly, they say data quality and staff expertise are their top two challenges, followed by governance and budget.

Smaller central banks were more likely to say staff expertise was a challenge, while larger institutions were more concerned about data processing and governance.