This report, commissioned by eleven leading energy trade bodies, presents the findings of a nationally indicative survey of 996 UK adults, including respondents across all political voting affiliations.
The organisations behind this polling represent the breadth of the UK’s clean energy economy – from wind, solar and storage to heat, bioenergy, hydrogen, carbon capture, nuclear, and demand flexibility.
Together, the ADE, BEAMA, the Carbon Capture and Storage Association (CCSA), Energy UK, the Heat Pump Association (HPA UK), Hydrogen UK, the Nuclear Industry Association (NIA), the Renewable Energy Association (REA), RenewableUK, Scottish Renewables, and Solar Energy UK speak for thousands of businesses across the supply chain – developers, manufacturers, investors, operators, generators, and innovators – working to deliver the UK’s clean energy transition at pace and scale.
If you require the report in a different format please email press@energy-uk.org.uk
More about the report
Undertaken by Early Studies, the polling uses two novel methods; social circle surveying and triple tense technique.
Most polling asks people what they think. Social Circle Surveying asks them something different: what do the people around you think? Your family, your friends, your colleagues. It’s a small shift in wording, but it changes what comes back.
Asking someone directly about a complex policy topic puts them on the spot. They generate an answer in the moment, often influenced by what they think sounds reasonable, or by what they feel they should say. Asking them about their social circle takes that pressure off. They are not being judged; they are reporting on what they have picked up around them. What comes back is something more durable: the ambient beliefs actually circulating in households, workplaces, and communities.
The method also picks up things direct polling misses. People often hold back views they sense are unpopular, and communities can end up with a shared but mistaken impression of what everyone else believes. The supplier profit finding on Question 16 is a clear example: social circles in aggregate believe their energy supplier keeps around 40% of the bill as profit, a vast overestimation of the actual share. That belief is not based on personal experience. It is a story circulating in networks, and it has held steady for five years. Direct polling of individuals would miss it; asking about the network brings it straight out.
The approach draws on a growing body of research on how people read their social environments. Asking about social circles produced more accurate predictions of the 2016 US and 2017 French presidential elections than conventional own-intention polling (Galesic et al. 2018). People turn out to be reasonably good observers of those around them, picking up on hints, disagreements, and offhand comments that do not surface in a standard yes/no question.
A note on how we report the findings. Throughout this report, results are written in direct form — “63% of the public believe clean energy makes the UK more secure” — rather than “63% of respondents believe people in their social circle believe clean energy makes the UK more secure”. This is deliberate. Because respondents can report multiple views present in their circle, the aggregate across a representative sample captures the prevalence of each view in the population rather than the single majority view of each respondent. In a representative sample, the share of social circles in which a view is detectable is a reliable proxy for the share of the population holding it. The direct form is readable and accurate; the social circle frame is the route to the finding, and the finding itself is about the public.
Further reading on social circle surveying:
- Ahlstrom-Vij (2022), Electoral Studies. Shows through simulation that the accuracy advantage of social-circle surveys is robust across realistic sources of bias.
- Galesic et al. (2018), Nature Human Behaviour. The headline paper showing that asking about social circles improved election predictions over standard polling.
- Galesic et al. (2021), Nature. A broader argument for “human social sensing” as an under-used research tool.
- Galesic, Olsson and Rieskamp (2012), Psychological Science. Explains why small samples of known others shape how people judge their wider social world.
Tourangeau, Rips and Rasinski (2000), The Psychology of Survey Response. On how respondents construct answers to survey questions, and why those answers can shift.
Questions in this survey were asked three times: what do people in your social circle think now, what did they think in 2020, and what do you expect them to think in 2030?
The 2020 results are not based on previous surveys. The 2026 figure is the real one. It is the only point at which respondents are describing what they observe today. The 2020 and 2030 figures are projections from 2026 — the respondent’s best reconstruction of what their circle used to believe, and their best estimate of where belief is heading.
A single snapshot tells you where opinion sits. A triple snapshot turns it into a trend line: how fast belief is moving, and whether the public themselves think the direction of travel is stable or about to reverse.
The retrospective figures should be read as directional rather than precise; people reconstruct the past in light of the present. The forward projections capture where respondents expect opinion to be heading, not a forecast of where it will land. What matters is the shape of the arc.
Respondents were asked twelve questions about energy infrastructure and the energy transition, each answered across three time perspectives: recalled 2020, current 2026, and projected 2030.
The findings describe how communities understand the direction of travel, not only where opinion sits today.
The headline is that most people understand the link between clean power and better energy security, between international events and energy bills, and most still prioritise tackling climate change.
The public backs long-term infrastructure investment, and wants costs spread fairly, preferring this to paying less now if that incurs costs later on. People recognise storage and flexibility as essential,
back grid upgrades and, in general, are more system literate, more engaged, and more willing to support the transition than the industry conversation typically assumes.
Six strategic insights emerge from the data:
- First, most people associate clean power with improved security, both in terms of energy supply and dependence on international supply chains.
- Second, whilst lowering energy bills is a priority, the public is asking for investment rather than delay; the preference for spreading costs over many years now leads by some way across every political persuasion.
- Third, whilst the cost of energy and energy security remain the top concerns, there is still strong support for tackling climate change and reducing air pollution.
- Fourth, there is majority support for a flexible, storage-led system to accommodate the variable output from renewables – grid upgrades, storage and smart technology to shift when we use power, with only 4% of the public willing to fall back on fossil fuels.
- Fifth, clean energy is increasingly seen as supporting the country’s jobs and industrial strategy, with UK industrial competitiveness the fastest-rising of the seven options.
- Sixth, whilst there is support for, and increasing understanding of the energy sector’s transition, there is a need to improve communications to correct misperceptions about the energy system and its costs.
People have valid and understandable questions about why energy bills are so high, and the sector, stakeholders and governments need to do a better job in communicating the facts.
There is public support for action to decarbonise power, tackle climate change and improve energy security.
What’s missing is an understanding of how this will practically be achieved – what is actually happening, what the impact will be and what the trade offs are.
Downloads
-
June 2026 Watt Communities Want_Six insights – joint energy TA polling
Type: PDF (10 MB)
