Towards standardised, action-focused carbon footprint measurement for UK government services

The cross-government, in-person gathering of teams that have measured the carbon footprint of specific services at Dell Technologies HQ in London.

This week I was lucky enough to attend a bit of a landmark event — a first in-person gathering of teams from different UK government departments who have measured the carbon footprint of individual services.

Our goal? To agree a standardised, cross-government approach to measuring the footprint of services, and feedback on a proposed framework for grading services according to the extent and rigour of that measurement as well as how far a set of principles for green ‘service design and operations’ have been followed.

The outcome? By the end of the day, we had reached a high level of consensus on the path forward, with agreement on the idea of a ‘maturity framework’ for measurement that focused on continuous improvement as well as the formation of a working group to drive this work!

Organisation and attendance

The impetus for this collaborative session came from Ishmael Burdeau, Sustainability Lead at the Department for Work and Pensions, and Mark Butcher and Louise Bunting from Posetiv, who facilitated the workshop. We were hosted by Dell technologies at their London HQ, and in attendance were representatives from Home Office, Department for the Environment, Food and Rural Affairs (Defra), Government Digital Service (GDS), Cabinet Office, as well as the Department for Work and Pensions (DWP).

Why is measuring the carbon footprint at the service level important?

Figures on the carbon footprint of services that we can have real confidence in are vital for delivering on our net zero goals. This data reveals whether assumptions about which approach to software development, server set-up, user journey design or data policies might be the most energy and carbon efficient hold up. It can also highlight the way that reducing carbon footprint is often aligned with making a service more efficient, cheaper, accessible and inclusive for users. In turn, these insights can empower informed decisions that ultimately lead to reduced carbon footprint. They allow us to be as efficient as we can be within the boundaries of what we can control on a given service, and prompts conversations about the impacts of those things that might sit outside our direct control.

Mark Butcher and Louise Bunting from Posetiv talking about ‘Why Carbon Counts’ when developing services.

But carbon is not the ‘be all and end all’ from a sustainability perspective.

At the same time though, when we look from the broader sustainability or ‘ESG’ (environment, society and governance) perspective, its clear that carbon footprint is only one environmental impact amongst many. Ultimately, ensuring that our services are genuinely ‘sustainable’ will mean going beyond optimising them in terms of energy use and carbon footprint and measuring for the full range of impacts pointed to the Sustainable Development Goals. Looking across the whole lifecycles of services, these could include things like water use, impacts on human and ecosystem health, resource depletion, eutrophication, and land use change amongst others.

However, given that current net zero legislation focuses on reducing our greenhouse gas emissions (usually measured in carbon dioxide-equivalence) and its central importance in combatting climate change, it seems a more than reasonable starting place!

Energy use (and associated carbon footprint) may be the principal environmental impact when a service is ‘live’, but if we look across the whole lifecycle then we see a much wider array of effects coming the the material extraction, manufacture, distribution and end-of-life activities associated with digital systems.

‘We want to take this away from being an abstract subject, to create things that are practical for service teams to use, so they can understand the impact of their decision points and be accountable for them, as well as be clear what they can influence and what are other peoples responsibilities’. (Mark Butcher, Posetiv)

What benefits would a consistent, cross-government approach to the measurement of service carbon footprint bring?

It was suggested that a consistent approach to what and how we calculate the carbon footprint figure for services across government would mean that:

• We are able to compare ‘apples with apples’, and have a clear sense of how well one service may be performing relative to another, and why.
• We can much more easily learn from one-another’s approaches to reducing emissions and share insights between departments, saving us all time and resource and accelerating our collective work on this front.
• We can foster a culture of transparency that, given Scope 3 emissions reporting includes the ‘upstream’ impacts of our supply chains, leverages the collective ‘buying power’ of the UK public sector to drive transparency and consistency in the reporting from suppliers, whether data centre, software or hardware providers.

The ordering of the workshop

The proceedings kicked off with a review of the different approaches to the measurement of service carbon footprint made in DWP, Defra, Government Digital Service and HMRC, with the appropriate personnel providing the background on methodologies and explanation of what each included.

Our focus then fell on the main task for the day — examining and providing feedback on a starter-for-ten, ‘straw man’ proposal for a set of criteria to categorise government services into 3 tiers — gold, silver and bronze. This framework featured two principal ‘pillars’: the first focused on the extent and rigour of carbon footprint measurement, ranging from those figures calculated based largely on assumptions and proxies (e.g. cost) through to those that were baselined, end-to-end and real-time emissions counts. The second considered how ‘green’ the service itself might be, based on how far a set of principles for ‘service design and operations’ had been followed during development.

In breakout groups, we discussed the merits of this proposed 3-tiered approach, focusing on how understandable the criteria were, how they could be evolved and developed, and whether any reframing was required. Then we fed back to the group.

The ‘strawman’ proposal for dividing a number of carbon reporting and measurement criteria into a 3 tiers, labelled gold, silver and bronze.

Some questions that emerged from the discussion on the framework itself:

• In the place of the gold, silver and bronze tiers (which it was suggested might perhaps both encourage conflict and some degree of complacency once ‘gold’ has been reached) would more of a maturity matrix model with a percentage score generated for service work better?
• Could the framework focus on continuous improvement by highlighting ways that a service is a work progress, helping teams feel empowered with what they have done but also drawing attention to what they can do to increase their score?
• Can we produce a scoring for the level of accuracy or confidence in figures for carbon footprint?
• Should service support be measured as part of the service, or considered separately?
• Could the framework be linked to and used at the various gateway points at which services are assessed, such as Service Assessments and Spend Controls?

And then questions about the broader challenge in front of us:

• How can a set of criteria for measurement cater to widely differing services, such as some which have significant numbers of offline steps, and those which are fully digital, or those exclusively with users in the UK versus some others accessed across the world?
• How can we document the decision points in projects, demonstrating with the right data when there are win-wins with other project priorities (cost, accessibility, security etc) and when there are tradeoffs made, how these have been decided upon?
• What about difference in applying measurement for brownfield versus greenfield? Does that change what we decide to measure?
• Should we use the term ‘measurement’ at all when we are using models or proxies to estimate carbon? Is an efficiency measure more appropriate than carbon footprint, as it recognises that there is only so much we can do with a given service, given the limitations of what we can influence?
• What is the right way to communicate this work to senior stakeholders? Risk and efficiency might be the right frame? Appropriate framing can help us move from a culture where the focus is often ‘carbon accountancy’ to one driving at putting action-focused measurement at teams’ fingertips?
• What else do teams need to support them to take the right action? Tools, training and guidance?

Wrapping up the day and next steps

The day concluded with a vote on a series of questions that reflected some of the main themes emerging from the group discussion:

• The question of converting the proposed gold, silver and bronze tiers into a ‘maturity matrix’ style framework with weighted scores received a unanimous vote from the room!
• There was broad agreement that the criteria on the ‘measurement and reporting’ pillar of the straw-man framework seemed in a more mature state than those on the ‘service design and operations’ side. The room voted to develop both going forwards, with the proviso that some key actors who had worked on the principles/standards front were not present on the day and would need to be looped in.
• There was a clear majority in favour of converting the present meeting into the first of a working group to meet at regular intervals and drive this important work forwards!