A data standard for transaction-level spending data

Deprecated: This specification is deprecated. See here for an updated technical specification for government budget and spending data.

We think it's time to create an internationally applicable, technical standard for the publication of transaction-level government spending data.

Why now?

The release of transaction-level data (i.e. information about individual disbursements or contract spending) is a relatively new idea, compared to the release of higher-level accounting information or budget overviews. The availability of such data allows for fine-grained analysis and oversight of activities and will, in the future, enable anyone inside or outside of government to reconstruct key reports from raw data. In order to perform these types of analysis, it is often necessary to combine spending information from several sources - either for completeness or comparison.

Civil servants cannot be everywhere. They cannot monitor the day-to-day running of every publicly funded project and cannot personally engage in discussions about all of the possible ramifications of the decisions made by government. This is where civil society can play an incredibly important role as infomediaries in establishing a foundation for trust and evidence-driven public discourse. What they need is comprehensive and clear data on the spending of the state. The standard would also facilitate two-directional information flow, not only from governments but to governments who could collect meaningful and structured feedback, as well as benefitting from the tools and expertise of those outside government.

There is also an unprecedented amount of auxilliary information being released including off-budget information such as aid revenues, geo-data and data on contracts and companies. A standard for transactional level spending data would make it easier to establish how these piece slot together and gain a complete picture of the financial health of a nation.

Goals

  • Tool for advocacy: as more and more spending data is released, we should document best practices for government data formats.
  • Tool for government: give interested parties a simple story for releasing financial data that captures international best practices.
  • Converge on the scope of data release, i.e. what needs to be in a published spending database.
  • Lead to greater efficiency - much time and effort is currently dedicated to converting between data-formats especially non-machine readable formats to machine readable.
  • Allow for greater comprehension of financial data - much time is currently also spent disambiguating terms and what is and isn't included in a given dataset
  • Enable scale - a basic format would enable tools to be built to work with the data
  • Long-term goal: enable comparability of the data through aligned classifications and taxonomies.

Challenges

Too often, standardization in this context appears to be supply-driven: every publisher wants to express the full range of data they hold and are willing to release. Necessarily, such an approach leads to a standard that is the superset of all the systems that feed into it.

Such designs tend to be of little use to the intended consumers, as they raise the barriers to understanding the information considerably. Our approach therefore is to generate demand-side use cases first, ensuring that everything that is done will generate value for data users.