Skip to content

Report of the Digital Government Review

Open Performance Data: creating a meaningful context

So far, there are no comments on this section. Jump to comments

Consider the current approach to releasing information about what the government spends money on.

Both government departments and local authorities already release their spending data. Yet this does not take place in a format that can be linked either across departments or easily linked to public services or outcomes. Spending data is thus not seen in the context of what the spend produces.

This risks creating ineffective and unproductive public debate. It moves us towards a state that simply spends less, not a more effective one nor one that is more aligned with people’s needs.

For an example from our own research, more of which can be seen in Appendix C, we used the FOI process to gather IT asset data from a number of local authorities. Linking this to spend data on suppliers was extremely challenging but this linkage could provide valuable insight on value for money and into the varying costs across different authorities.

The review determined that open performance data need to be placed in parallel with open spending data, allowing spend data to be seen in the context of the performance produced.

This should apply to performance data for all public services regardless of who is providing it, i.e. whether the service is directly delivered by the public sector, delivered in conjunction with partners, or delivered by an outsourced partner.

Imagine the debate in Spring 2014 over the performance of the UK Borders Authority (UKBA) passport service [35] if citizens could have actually seen a day-by-day or near real-time view of performance data – for example the average processing time for a passport. Instead, a key part of the political debate revolved around whether a photo of a pile of passport applications was a real backlog or not [36]. It was farcical to have this discussion without public performance data.

A wholly new urgency, sense of responsibility and measure of accountability would have been in place. Similar effects can be imagined in other sectors: the attendance levels in schools, performance in hospitals, performance of major delivery projects, performance of local authority services and so on.

We can be reasonably certain that UKBA does know how many passports it is issuing in near real time as their unique serial numbers go live. When a service is in trouble and the subject of political debate it becomes profoundly undemocratic that only the government knows this information, not the people who are suffering and want to hold the government to account. Knowledge, as they say, is power.

Performance data is the type of information that ought, of course, to be made available to people working within those public sector organisations – for example, it will provide better information and tools for policy makers, organizational leaders and people managing outsourced contracts.

But we must not stop there. By releasing data to the public as well we can help create a newly informed public debate and support more informed decision-making by citizens on a day-to-day basis.

For example, if a citizen can see that the queue for a passport is 8 weeks or the average time to register to vote in a particular local authority is 6 weeks, they will be in a more informed position when prioritizing when to fill out a passport and voter registration form.

We will need to be careful not to fall into the same trap as spend data. Releasing data without appropriate context can be counter-productive, so an approach that works with citizens to understand changing needs and determine which data is valuable should be established and followed.

This should not be seen as letting government “off the hook” for providing a poor level of service. The open nature of the information will generate an informed debate around the cause and accountability for any poor performance but, in the meantime, people will be in a more informed position.

Both the Government Digital Service and some local authorities have made some progress in the direction of releasing performance data [37] [38], which is to be applauded, but the information that is currently available lacks context; can be difficult to understand for the average citizen; and often only contains information on the elements of the service that have been digitized or moved to the Internet.

What is needed as well as the underlying data is a measure of the performance of the full end-to-end service.

Few citizens are interested in how many people are currently on a given webpage. This information is useful for the people delivering the service (and it is fascinating to some of the more technically minded of us outside of the public service) but it is the performance of the end-to-end service that truly matters to people. How long will it take from filling out my application form before I can vote? How long is it taking to deliver passports? How much are different schools improving the long-term performance of children?

Performance data should be open data. Releasing spend data without context can be damaging and dangerous, it is difficult to prevent a descent into unhealthy and ill-informed debate when only part of the picture is available. Releasing meaningful performance data will improve the debate and improve our democracy.

Recommendation 8

Priority: High

Improve accountability by releasing public sector performance data as open data

Performance data is what is meaningful to people and communities outside government. It can help them make choices and decisions. Meaningful performance data can be determined by iteration or by ongoing consultation and collaboration. Spend data should be released alongside performance data to improve debate and accountability. The combination of the two should be tied to strong accountability mechanisms to enable people to hold their elected representatives to account

“Analysts and policy makers must understand the limitations associated with the use of massive largely unstructured data sources and ensure that they derive evidence based policies from them in a way that is both scientifically and statistically correct, fair and ethical to contributors and non-contributors to those databases alike” – Professional body


This page reformats automatically when printed. Print this section

Please note that comments left here are public - you can also make a private submission.

Your email address will not be published. Name, email address and comment are required fields. Please note we may moderate comments.