Report (draft v.10)

Howard Noble, David White, Joss Winn, Derek Foster, David Blach, Ken Kahn, Richard Hall, Richard Bull
Correspondence to Howard Noble:


The project team would like to thank the JISC Greening ICT Programme team and Rob Bristow and Peter James in particular. We would also like to thank: Derek Foster for helping out at the workshops, contributing with great ideas and suggesting papers we should read. Alex Dutton for his work on (Oxford's open data store) and general enthusiasm and practical endeavor. Josie Fraser for numerous introductions to people and groups and for opening our eyes to the potential of open meter data and energy literacy in schools. Arthur Hjorth for putting us onto the work of Elinor Ostrom and generally helping with the big ideas.


oxford carbon management strategyInstitutional managers can try to reduce energy use in the background. They can install technology that automatically powers down lights and computers and try to steer procurement towards energy-efficient equipment. However, very few changes can be implemented without resistance, and before long managers will need to ask employees to contribute. 

Behavior change initiatives must overcome sizable hurdles if they are to deliver:

Price of energy

It is difficult to persuade employees to keep doing the little things that together add up to huge savings. For example if we all turned off idle PCs every night the savings would be huge. The trouble is we do not think about savings collectively, we think in terms of our own individual effort. Turning a PC off overnight would save about £0.14 (14 hours x 0.100kW x 10p/kWh) which is hardly motivating at home, let alone at work where we don't even see the bill.

As consumers we envisage energy as both a private and public good. We demand that energy is as cheap as air so that we don't have to think about consuming it. We also become very private and obstructive when asked how much energy we consume and what we use it for. This double standard ensures that we continue to treat greenhouse gases and the cost of the wars we fight to secure access to hydrocarbons externalities (as an economist might put it).

Institutional managers and politicians who are interested in climate change share a common problem - they lack a mandate. 

Rebound effect

William Jevons noted that technological advancements that result in improved efficiency (referring to coal) tend to increase rather than decrease the rate of consumption of that resource. The Rebound Effect predicts that even if we manage to tame our desire for a specific resource we will often spend savings on other resource-intensive activities e.g. use money saved on utility bills to buy extra flights. Whilst there is considerable debate surrounding the rebound effect it seems reasonable to say that improved energy efficiency does not necessarily reduce overall consumption. 

Quantified selves

To be sure efficiency leads to more sustainable consumption we need to measure resource use across the spectrum. In particular we need to include the environmental footprint associated with products we buy in from India and China as our own. Price is a very useful lever for controlling how much we consume collectively and we have been trained to take a great deal of notice wherever we are. It is also clear that we live in consumerist and growth obsessed cultures where prices are set to encourage ever more demand. 

We are not accustomed to considering environmental indicators when we make everyday decisions. Most people do not understand what a kWh is, have no idea how much energy everyday electrical devices consume, and baulk at the idea that it is necessary to consider the environmental footprint associated with manufactured goods.

Greenhouse gas emissions

Encouraging people to contribute to sustainable energy initiatives by appealing to their understanding of climate change science is equally fraught. Psychologists have a depressingly detailed understanding of the psychological and social barriers that prevent this approach from working. The table below is copied from the excellent summary paper by Robert Gifford (see below):

General psychological barrier Specific manifestation
Limited cognitionAncient brain; Ignorance; Environmental numbness; Uncertainty; Judgemental discounting; Optimism bias; Perceived behavioural control/self-efficacy
IdeologiesWorld views; Supra-human powers; Techno-salvation; System justification
Comparison with othersSocial comparison; Social norms and networks; Perceived inequality
Sunk costsFinancial investments; Behavioural momentum; Conflicting values, goals and aspirations 
DiscredenceMistrust; Perceived program inadequacy; Denial; Reactance
Perceived risksFunctional; Physical; Financial; Social; Psychological; Temporal 
Limited behaviour Tokenism; Rebound effect 
Dragons of Inaction: Psychological Barriers That Limit Climate Change Mitigation and Adaptation. Robert Gifford, University of Victoria.

Social animals

Our consumer behaviours are largely shaped by the people around us - we are social animals who find it difficult to make rational decisions. Consider the following thought experiment: 

A population is composed of 3 groups: 50% take their consumption patterns from the people nearest them, 30% consume to their heart's content, 20% are are reducing their energy consumption. Over time, will energy consumption reduce, stay the same or increase?

It seems plausible to suggest that energy consumption will increase to close to the level of the 30%. Despite the actions of the 20% a minority group will find it difficult to prevail, especially if their agenda is unpopular (saving energy is not fun). The 30% will therefore influence the 50% more than the 20%. To compound this dynamic the 20% may then feel deflated enough to revert to the evolving norm. 


Our solution focuses on the use of freely available on-line tools to support institutional behaviour change initiatives. We draw on the seminal work of Elinor Ostrom who spent over 30 years studying how some communities have successfully managed to appropriate common pool resources (fresh water, fish and wood from a forest) and avoid a tragedy of the commons. She created a set of design principles to help groups build sustainable communities:

 1AUser boundaries: Clear boundaries between legitimate users and non-users must be clearly defined.  
 1BResource boundaries: Clear boundaries are present that define a resource system and separate it from the larger biophysical environment.  
 2ACongruence with local conditions: Appropriation and provision rules are congruent with local social and environmental conditions.  
 2BAppropriation and provision: The benefits obtained by users from a common-pool resource (CPR), as determined by appropriation rules, are proportional to the amount of inputs required in the form of labour  material, or money, as determined by provision rules.    
 3Collective-choice arrangements: Most individuals affected by the operational rules can participate in modifying the operational rules.  
 4AMonitoring users: Monitors who are accountable to the users monitor the appropriation and provision levels of the users.  
 4BMonitoring the resource: Monitors who are accountable to the users monitor the condition of the resource.  
 5Graduated sanctions: Appropriators who violate operational rules are likely to be assessed graduated sanctions (depending on the seriousness and the context of the offense) by other appropriators, by officials accountable to the appropriators, or by both.  
 6Conflict-resolution mechanisms: Appropriators and their officials have rapid access to low-cost local arenas to resolve conflicts among appropriators or between appropriators and officials.  
 7Minimal recognition of rights to organize: The rights of appropriators to devise their own institutions are not challenged by external governmental authorities.  
 8Nested enterprises: Appropriation, provision, monitoring, enforcement, conflict resolution, and governance activities are organized in multiple layers of nested enterprises.  
Cox, M., G. Arnold, and S. Villamayor Tomás. 2010. A review of design principles for community-based natural resource managementEcology and Society 15(4): 38.
Elinor Ostrom. 1990. Governing the Commons - The Evolution of Institutions for Collective Action. (See pp 88-102 for design principles). 


We argue groups should structure their activities around two principles:
  1. Re-conceptualize energy as a common pool resource - consume as if extracting resource from a fishery, forest or fresh water groundwater basin. 
  2. Engage in a process of substitution: (a) reduce use of energy (b) invest savings in environmental projects. 
And use these heuristics to guide the selection and use of on-line tools:
  1. Gather energy meter data and publish it at an appropriate level of granularity, openly, as near to real-time as possible
  2. Clearly represent the availability of the resource by setting quotas/targets and reporting performance
  3. Re-represent performance data in units easily understood by members of the group
  4. Channel any savings into investments to minimise the rebound effect
  5. Display the member profiles and key individuals within the institution 
  6. Compare and contrast performance with other groups and share ideas
  7. Keep a record of the story that emerges (achievements, things that didn't work etc)
  8. Accurately show how the group's collective effort relate to the big picture  


The prototype below illustrates how we have used the two key principles and heuristics to create a sustainable energy dashboard, it is composed of the following elements:

  • Electricity meter data is published openly on-line. This could mean half-hourly readings pushed to the web automatically in real time or weekly readings entered into an on-line spreadsheet manually. Each group can decide what they want to do bearing in mind the savings they want to achieve and the technology they have at their disposal for monitoring and reporting.
  • Performance against a monthly target is visualised in real time.  We have a particularly complicated example, where we show real time performance against a monthly target (projected and actual). A more simple approach to measuring performance could well be adequate.
  • Statistics are presented in a variety of units. kWh are confusing to some and it might be useful to represent energy use in a range of units e.g. number of cars off the road, houses off the grid etc.
  • Financial savings are given in pounds and multiplied by a factor agreed by a sponsor to give an investment fund. Raw savings can be calculated easily based on the price per unit of electricity a group pays. 
  • A brainstorming tool allows people to agree on investments or specific actions that the group wants to tackle. 
  • The name and picture of the institutional energy manager is included in the dashboard. This is the intersection between the group and the institution. This relationship can be mutually beneficial if the institutional lead has a mutual interest in energy efficiency. The group ask the manager for resources, the manager can ask the group for contributions.
  • The groups energy performance is compared with other groups using a league table. Groups may be motivated by comparing their performance with other similar groups.
  • Pictures, videos, blog posts, tweets and as much information as possible is assembled, perhaps on a timeline to record the story of the group activity. 
  • We did not tackle the last heuristic but a good place to start envisaging stories or info-graphics that accurately relate a groups contribution to the big picture would be  
We all have cheap and easy access to a wealth of well-designed tools that we can appropriate to build sustainability dashboards. For example:


Not all groups will have the IT skills necessary to create their own dashboards. We envisage an iGoogle-like platform that helps groups build fairly generic dashboards from a small set of widgets:
  1. Brand dashboard with a name and icon
  2. Put group and the individuals within the group onto a map
  3. Visualize meter data
  4. Visualize energy performance against a target
  5. Visualize the performance of several groups together
  6. Visualise relevant national and regional energy trends
  7. Display survey poll statistics (brainstorming and other activities)
  8. Forum to help share information
  9. Member names, images and key roles
  10. Display Mozilla badges (investment and other achievments)


We asked for funding for this project because we were very skeptical about two popular ideas in this area of inquiry:
  1. Showing people how much electricity they consume (in whatever form) leads to improved energy efficiency. 
  2. Improved energy efficiency results in a reduction in overall consumption. 
We asked for funding to:
  1. Give us time to learn from a wide range of literature
  2. Run participatory design workshops to learn from peers
  3. Explore a range of technical ideas with respect to building a different kind of energy saving dashboard. 
Our hunch was that we can draw on online game research to find ways to motivate people to tackle complex tasks. Online game designers are particularly good at representing quantitative measures of performance as qualitative indicators of social capital. In other words, getting people to slay dragons so they can be rewarded with points/badges they can show off. We set out to ask whether similar similar techniques could be used to help people cooperate on slaying a different set of dragons

We completed the following project tasks:

  1. Learn from a wide range of academic literature
  2. Run three participatory design workshops 
  3. Pilot the design of an open data store that holds half-hourly meter data
  4. Create compelling visualisations of electricity meter data
  5. Create a prototype internet use patterns for groups wanting to improve energy efficiency
  6. Organise a capacity building event for designers and programmers interested in this area of research
  1. We have been less successful than we hoped at created an open data store and advanced visualisations. We were hampered by the difficulty in scraping data from existing energy meter databases. This is not to say these are difficult avenues for other groups to adopt but technical development was not the focus on this project and we felt our effort was better spent in prototyping an innovative ideas for developing energy dashboards 
  2. We decided to run a second larger workshop in Oxford in place of one at De Montfort
  3. We gave a papers at 
    1. UKERC Energy and people: futures, complexity and challenges conference  
    2. Digital Engagement 2011 conference 


We wanted to read as widely as possible around the main questions that underpinned the project. This helped to design the workshops and arrive at the final set of heuristics. Please see reading list in appendix A

The workshops  were composed of three sections:
  1. Survey
  2. Common goods game
  3. Semi-structured discussion
10-15 participants attended each of the first two workshops (Oxford and Lincoln) and 50 people took part in a third shortened workshop convened by the JISC Greening ICT programme. See appendix B for more details.

Take ideas from academic research and the workshop data to create heuristics and a prototype energy efficiency dashboard.


The main deliverable at the start of the project was the workshop (see appendix B ). To design the survey we had to quickly ascertain what questions we wanted to ask workshop participants, and what we could realistically do with a survey. We decided to focus on:
  1. The level that electricity metering was conducted and whether group performance statistics could reduce the need for fine-grained metering i.e. at building rather than room level.
  2. Getting participant reactions to specific visualisation that have been developed by other research groups
  3. Probing attitudes towards openness with respect to electricity meter readings
With research papers and workshop data to hand we then went through a brainstorming phase to arrive at the heuristics and prototype internet use patterns. 

We wanted to demonstrate our ideas by building an open data store and showing how it can be used to visualise data in compelling ways. We built two connectors to scrape data from an Ion Enterprise web server and the Scottish and Southern Energy billing database. Whilst the connectors work and we have a pilot service ( we have not had time to create a robust solution - if the the data providers change their systems the connectors will break. 

We then set about exploring how easy it is to visualise university meter data using and by implementing a Google gadget. We managed to create very simple implementations but we didn't get far enough to share code other groups could use. We are satisfied however that these are two of the better approaches for creating interesting visualisations: for coders and Google Docs spreadsheets for everyone else. 

We spent quite a bit of time deliberating on the surprisingly complicated subject of how to set targets. There are at least two main considerations when setting targets: (1) their ongoing psychological influence and (2) how easily they are to implement. For instance, it might be easy to understand and implement a target that is simply 5% reduction on the reading at the same time the previous year, complications may well arise: (a) what does it mean to the target if more people start using the building, (b) what will happen to psychological impact performance against the 5% target once the easy efficiency gains are achieved, (c) how will the equation that calculates the absolute target at any moment cope with missing data, (d) how to go about calculating a projected performance against a target in real-time. 

The main underlying problem has been that electricity meter data is seen as being owned by energy utility providers and service providers who create applications for institutional energy managers. In both cases, the data has never been seen as something that should be freely available for others to use as they please. This creates significant problems for groups that want to take ownership of an energy efficiency initiative as they will find it difficult to understand energy and report on performance. 

We presented a paper titled Open Meter Data to the most senior committees at the University of Oxford and there was unanimous agreement that it was desirable that electricity data should be freely and openly available on the web. Unfortunately there is still a significant amount of technical work to make this a viable option. It is also apparent there are considerable political issues in terms of persuading groups who have hitherto believed they own the data to relinquish control. The main points we made in the paper were:
  • It may no longer be necessary to pay for expensive software that essentially provide little more than a layer of authentication over a database of meter data and graphics tools.
  • Departments and colleges would be able to make use of freely available tools for reporting, monitoring and visualising the meter data in ways that they choose. 
  • People across the University (and beyond) will be able to use the data to create innovative new tools to help the institution improve energy efficiency. 
After the end of the project we worked with Oxford Brookes University to submit a proposal to the EPSRC to run an action research project that would use the heuristics to improve energy efficiency in a range of public buildings. Although unsuccessful it was a valuable exercise in terms of sharing and improving the ideas.  


The main outputs and results for this project are:


We have focused on describing heuristics and internet use patterns to help groups re-envisage energy as a common pool resource. We argue that we need to shift the focus of energy dashboards away from being simply an information display, towards being a community building tool. Elinor Ostrom research convinces us that there is a lot more to avoiding a tragedy of the commons than monitoring resource appropriation. Real change comes when individuals form groups and take ownership of a problem. Group members agree rules and sanctions and sufficient levels of trust that a minimal amount of conflict resolution needs to happen. 

We have a long way to go. Most people still think energy is either an invisible or cheap private good. It seems almost blasphemous to suggest that we fight wars to secure energy supply or even that energy use results in pollution. The evidence suggests otherwise of course, but if we are to believe psychological research we should not be surprised: cognitive dissonance is a distinctly human trait. 

Energy underpins everything we do, the change we need to make is immense and it is unlikely we will solve it with technology alone. We need leaders but leaders need mandates. We recommend acting collectively now so that we can learn hands-on what needs doing and which leaders to select. 


The next phase of this line of inquiry is to conduct action research to:
  1. Observe and record what happens within groups as people try to govern a resource.
  2. Gather evidence to improve and extend the heuristics
  3. Work out how to quickly build different kinds of dashboards using the heuristics
  4. Map out how to set up and maintain sustainable energy campaigns 


Behaviour change:
Politics, philosophy and economics:


You can download the survey questionsvisualisations participants looked at when completing the first part of the survey, and the final survey data.

In the second part of the session we ran a game where participants were given 10 sweets and asked to contribute some to a central bank each round. The amount contributed was increased by a small percentage and divided equally amongst the players. At the end of the game everyone revealed how many sweets they had amassed - those that had 'cheated' by giving little and received the benefits of the cooperative bahaviour of others would be the richest. 

In the third and final part of the session we opened up the discussion to wider topics such as whether energy could be seen as a common pool resource, attitudes towards targets, and the Rebound Effect. 

In summary, we learnt the following from the workshops:
  • Most people are happy for energy data to be made openly available but everyone assumes it there must be a rule against this being so
  • Whilst people prefer energy data to be as granular as possible (office rather than University level) there is some evidence to suggest people can be persuaded to act within larger groups by for instance using pledge data. This is significant because it means organisations might be able to avoid installing more expensive metering devices than they otherwise might
  • Whilst it is obvious that people vary with respect to the visualisations of performance they personally find most engaging, this on reflection, might be irrelevant if the visualisations are made available on-line. It is easy to imagine a dashboard that makes a range of visualisations available and people can choose the one they most prefer. 
  • A more significant problem is one where people have strong preferences for the channel they prefer to use to receive information (email, posters, web pages etc). Catering for everyone's needs will just mean more work pushing information to all the relevant channels. 
  •  There was a striking difference between the participants perception of the value of competition. For some competition seemed the natural and obvious way to motivate action (i.e. pitting groups against each other), whilst for others this was seen as exactly the opposite to what would work best. Since we recommend openness there is the risk that groups could be thrust into competition whether they like it or not. On reflection we think the tone should be cooperative if not least because comparison between groups is fraught because of the thorny issue, in the field of energy efficiency at least, that it is difficult to quantify the value any one group gets from their energy use (electricity used for research might result in new clean sources of energy for instance). If the tone is cooperative then groups will be more like to exchange advice on how they achieve improved efficiency. 
  • The Rebound Effect was a difficult concept for most, and in general we think it might lead some to feel dispondant and less likely to contribute to an initiative. Never-the-less, if we want to reduce global CO2 emissions local efficiency must result in global net reduction otherwise there seems little point. We tackle the Rebound Effect subtly in the heuristics by saying savings should be ring-fenced. We hope groups will decide to invest their savings in things that result in a net energy reduction e.g. shares in a wind farm. 
Howard Noble,
Mar 21, 2012, 6:34 AM
Howard Noble,
Mar 21, 2012, 6:21 AM
Howard Noble,
Mar 21, 2012, 6:23 AM