Science & Technology CommitteeWritten evidence submitted by Professor Nick Pidgeon (Risk 21)

Background

1. The Understanding Risk research group at the School of Psychology Cardiff University1 is a centre of expertise for the study of public risk perception, risk communication, and public engagement with science and technology. The work of our group is independent of stakeholders, and has been funded over the years through grants from academic sources including the Leverhulme Trust and the Research Councils (ESRC, EPSRC, NERC, USNSF).

2. We are currently conducting a major project (2011–12) under the UK Energy Research Centre (UKERC) programme, investigating public acceptability of whole energy system change to 2050—including all major elements of energy supply infrastructure (renewables, carbon capture and storage, and nuclear).

3. We have particular expertise in attitudes to nuclear power, having built up unique empirical data sets through studies of British public responses to nuclear energy, summarised as follows.

Major nationally representative surveys conducted for us by Ipsos-Mori in 2002, 2005 and most recently 2010.

Parallel qualitative studies (the earliest in 2002) of how the public talk about nuclear power in relation to climate change and energy security concerns.

A major study from 2003–08 on “Living with Nuclear Risk” with residents living close to the Bradwell, Oldbury, and Hinkley Point power stations.

We are the only social sciences centre in the UK to have systematically studied British public attitudes to nuclear power, both empirically and theoretically, over the past 10 years.

Key Factors Influencing Public Risk Rerception and Tolerability of Energy Infrastructure Facilities and Projects

4. Risk perception and risk communication research2 is a well-established field of inquiry, dating back to work in the mid 1970’s.

5. A first general finding of this work is that it is a misnomer to talk of the public as a single undifferentiated entity. In reality many different “publics” exist in Britain as elsewhere—comprising a myriad of attitudinal positions, cultural diversity, interests and concerns regarding risks.

6. A second clear finding is that—despite what is often assumed—concerns about risk do not stem from any simple deficit of knowledge about the science or an “ignorance” of what uncertainty and probability might mean (the so-called deficit model of science communication). There is now ample evidence to show, and the House of Lords “Science and Society” report of 20003 makes clear, that simply providing more knowledge does not automatically bring greater acceptability of a technology or risk issue. Rather, the relationship between greater knowledge and attitudes is complex: typically quality of argument improves as people gain more information, while views also become more polarized with stronger opinions voiced on all sides to a debate.

7. Research has shown that public concerns about technological and environmental risks are based upon a set of factors not ordinarily incorporated into traditional risk assessments. In particular:

(a) Qualitative characteristics of the hazard: eg whether a risk is seen as uncontrollable and involuntarily imposed, the perceived catastrophic nature of “worse case” accidents, and perceived lack of knowledge about future impacts.

(b) Trust and concerns about risk governance: in particular whether the responsible decision maker and/or regulation is competent, fair and caring; whether unintended consequences of complex and rapidly moving scientific enterprises will be controlled; and worries about the social commitments that a technology entails (eg the form of society and organisation required to keep a technology safe).

(c) The historical context within which a hazard arises; eg links between civilian nuclear technology and its military uses, and the culture of secrecy.

(d) Media reporting, often difficult to anticipate, which can at times attenuate, and at others amplify, a range of these “non-risk” aspects4.

(e) The affective properties of a hazard—that is, whether it promotes a general feeling of insecurity or fear. It is now recognised that good decisions and reasoning processes require both “analytic” (appraisal of risks, benefits etc.) and “affective” components.5

8. Recognition of the above cautions against relying solely upon communicating engineering concepts of risk (probabilities, damage estimates, expected fatalities etc.) when engaging people and communities, as these are unlikely to meet their actual concerns.

9. What is also clear from four decades of research is that a technology with potential risk but also a very clear and visible personal benefit (eg personal benefits of energy use) will be more acceptable to people compared to one which does not have such a visible benefit. Equally, a risk is sometimes less acceptable where risks and benefits are inequitably distributed—eg benefits believed to go to others but the risks to ourselves.

Public Perceptions and Nuclear Energy

10. Historically, many of the above factors have come into play in shaping perceptions of nuclear power. Concerns about catastrophic accidents (and associations with atomic weapons), the invisible and long-term nature of radiation, involuntariness of exposure, and strong negative affective associations. Regulation and management of nuclear power is also distrusted by many people, in part because of visible failures (TMI, Chernobyl) in the face of past reassurances of safety, but also because of secrecy and lack of accountability. Openness and transparency are now watchwords for regaining confidence.

11. In more recent times, but pre-Fukushima, opinion polling had indicated a significant reduction in opposition amongst the public in Britain,6 as compared to the very high levels of opposition (up to 80%) reached after the Chernobyl disaster. This in part reflected the arguments being advanced regarding nuclear power’s possible contribution to combating climate change and to delivery of future energy security, but also the fading collective memory of Chernobyl.

12. A closer look at the national data shows a more complex picture, however, with a large proportion of recent support remaining conditional—a “reluctant acceptance” at best.7 While many more in Britain have indeed come to support nuclear power over the past decade they do so while viewing it only as a “devil’s bargain”, a choice of last resort in the face of the threat of climate change.8 Given the choice individuals still show very clear preferences for renewable electricity generation.

13. Views become more complex at existing nuclear locations, many of which are being asked to host nuclear new-build. What we know here is that the response of people in such communities does not always mirror that obtained from national samples. A common assumption is that people in these locations will be overwhelmingly positive about nuclear power, because of long-standing experience with the local station and local economic benefits. While it is true that surveys conducted at such locations tend to be somewhat more positive about nuclear power in aggregate compared with samples living elsewhere, detailed research again suggests a more complex picture and the need to look beyond the headline statistics.

14. Geography certainly matters, as when a station contributes economically or in other ways to nearby communities, but not to others slightly further away who might also nevertheless perceive themselves to be at risk. The detailed history (for example of organised protest over the years) and socio-economic location (degree of dependence upon the station for local jobs) is also important in understanding local views. In this respect the existing UK nuclear sites vary enormously in social, economic and historical circumstance.

15. In our own interview research in 2004–07 at Oldbury and Bradwell, we found that many nearby local residents did express confidence in site activities. For most of the time people saw their existing local station as both a familiar and unremarkable feature of the locality, and confidence in plant activities had also built up over time. However, almost everybody we interviewed could also recount instances (news of the Chernobyl disaster, the London terrorist bombings, a friend being diagnosed with cancer) where the “extraordinary” risks of nuclear power, and with this very real personal anxieties, had been brought home to them in a powerful way.9 People at such locations also deploy various forms of humour and irony in their talk about nuclear power and its potential risks, in order to express uncertainties and anxieties which might otherwise be difficult to voice.10

16. Whatever their position on nuclear power, the vast majority of residents (84%) we surveyed in 2008 at Oldbury and Hinkley Point agreed that the industry and government should fully involve them in plans for siting new nuclear power stations locally.11 Many also had concerns about radioactive waste (77%).

17. It is too early to reliably judge the full impacts of the Fukushima Disaster on public perceptions in the UK or internationally. Polling internationally has shown large declines in support in many countries (eg Germany, France, Japan). It is genuinely puzzling that in the UK (and the USA) there remain as many people in favour as are opposed to nuclear power in such polls.12 This may be due to Fukushima’s spatial distance, and/or because people here attribute the primary cause to an overwhelming natural disaster, or because climate change and energy security discourses remain important for British people. We do know, however, that accident risk is more salient now for people, and that there is also some evidence of a widening gender gap (women are more concerned) in attitudes to nuclear power. Detailed empirical work with identical survey items and methodologies replicating previous studies is now needed to provide robust answers to this important question.

18. The impacts of Fukushima on existing UK nuclear communities are likely to be even more complex. As argued above, anxieties always exist below the surface at such sites and external events such as Fukushima have the capability to bring them to the surface, and powerfully so, for many people. Communication, dialogue and engagement with such communities are all likely to become much more difficult—both practically and ethically—as a result.

19. While the current UK policy focus is on the existing nuclear sites, the upper end of potential expansion of UK nuclear power—to 40GW or more—would almost certainly require development at sites without any history of nuclear operations. It is my personal opinion that the Fukushima disaster has made such development almost impossible in the short-term. The social and governance implications of attempting to site in locations with no history of nuclear operations have received no attention to date, and yet some of the more ambitious nuclear scenarios appear to depend upon this.

20. It is worth noting here that there is now an extensive literature on local acceptability and siting of new energy facilities—some in relation to nuclear, but more recently drawing on detailed empirical studies of renewable energy siting.13 A clear conclusion is that the term NIMBY (Not In My Backyard) is an inaccurate description of the ways that such communities are responding when they raise questions about new developments that might affect them. It both marginalises and denigrates many legitimate local concerns, including: threats to local identity and community fabric; distrust of the motives of powerful outside organisations; concern about the degree local involvement and control; worries about disruption during construction activities, or despoliation of valued landscapes; alongside any potential risks and long-term uncertainties. Again, this work cautions against relying solely upon communicating (engineering) concepts of risk when engaging such communities.

How are public risk perceptions taken into account in the planning process for energy infrastructure?

21. At the moment there is no legal obligation to take such perceptions into account in the planning process. However, in practice they can and do impact the outcomes of planning inquiries through the legitimate activities and representations made by interest groups. In extreme cases they can disrupt the planning process entirely, when developers are persuaded to withdraw plans in the face of what they believe to be concerted local opposition. Perceptions can be very real in their consequences!

22. The Health and Safety Executive has introduced the notion of “societal risk” to risk policy vocabulary14—indicating risks which can lead to high numbers of fatalities in any one incident. They argue that there may be a case for stricter safety limits/investments applied at planning and elsewhere if the public are averse to such “scale” events (as compared to the equivalent number of fatalities occurring in individual accidents). However the evidence for such “scale-aversion” as defined by HSE is mixed.15

23. One of the paradoxes of contemporary planning inquiry processes is that they exclude many of the legitimate issues that local people often want discussed. Can we trust the developers and regulators in what they are saying? Will local people have a degree of control over outcomes? Will this development lead on to further (possibly worse) developments in the future? What about any uncertainties that probabilistic risk assessments cannot estimate? For this reason, risk and safety can become vigorously contested domains, as the only potential vehicle to raise such issues.

How effectively does local and central Government communicate risk and could it be improved?

24. Currently there is no unified approach to risk communication either within central or local government in the UK. Efforts have tended to remain fragmented and tailored to the needs of individual Departments. For example, the Department of Health published useful guidelines on risk communication in 1997, including a “risk comparison” scale for placing risks in context.16 The Health and Safety Executive have sponsored research on risk communication in the workplace,17 and with respect to major accident hazards,18 while in 2002 the Cabinet Office published general guidance for Departments on handling risk, including advice on communication.19 The Treasury has also issued guidance on managing risks, including a “concern assessment framework” for gauging perceptions.20 The Food Standards Agency—set up in the wake of the BSE debacle—were also proactive under their first Chairman Lord Krebs, especially in relation to risks where evidence was sparse or highly uncertain (their philosophy, generally regarded as successful, was to be as transparent and open as possible about uncertainties). I make suggestions to overcome this fragmentation at national level in Paras. 29–31 below.

25. An important issue here is to define what we mean by risk communication. Current thinking is that it should be oriented towards dialogue with people. The US National Research Council has defined risk communication as:

“an interactive process of exchange of information and opinion among individuals, groups and institutions. It involves multiple messages about the nature of risk and other messages, not strictly about risk, that express concerns, opinions, or reactions to risk messages or to legal and institutional arrangements for risk management” (National Research Council, 1989, p21).21

26. Risk communication is now a very mature field of research, with the UK at the forefront of international activity in this field. The core lessons of this research are that:

(a) To succeed such communication should encompass a dialogue rather than progress in a one-way (“expert” to “public”) fashion;

(b) Given the right support and risk representations people can grasp and deal with complex probability information. Here, aids such as visual frequency displays or comparisons with other (qualitatively similar) risks can help, but all techniques have to be used with care and with reference to the evidence on their effectiveness;

(c) While it is important to “get the numbers right”22 communicating risk is about far more than this—enabling trust, exploring divergent values of varied people, meeting concerns about uncertainty or governance arrangements etc. As such, the public should not be characterised as “irrational” in their responses to risk estimates when they are concerned about these other matters.

(d) Above all there is a need to continually evaluate the impacts of communications as related to the original purpose of a communication programme. Such evaluation can range from simple measures of information transmission, to satisfaction with deliberative or participatory processes, to increased trust in decision makers.

27. Risk communication becomes particularly difficult when there is sparse or uncertain evidence (the “unknown-unknowns”). Under such circumstances the best advice is to be as open and transparent as possible about what is and what is not understood, how uncertainties might be resolved, and what precautionary measures might be adopted in the meantime.

28. Strategic capacity in risk communication is sorely lacking in the UK as elsewhere. Individual proposals for a risk information centre have been raised in the past (eg by the Hazards Forum) but there is no one entity in the UK dedicated to research, communication and policy support for risk.

29. In a paper this year in Nature Climate Change with Baruch Fischhoff of Carnegie Mellon University, completed before the events in Japan, we argued for a strategic approach to risk communications23 as applied to climate change. As this argument is a generic one, applicable to many complex, uncertain, and socially divisive risk issues, it applies equally to energy risk communications. We argue that the proper goal of risk communication is in supporting decisions—whether this be a government decision to proceed with a piece of infrastructure, or a local community debating the implications of this for them. What is communicated is then dictated by the requirements of the decision problem at hand.

30. A strategic approach to risk communication comprises two elements: (1) strategic listening—an approach which treats communication as a genuine dialogue, and seeks to thoroughly understand intended audiences and their decision needs prior to communication design; and (2) strategic organisation.

31. The range of skills needed for such an effort would include natural scientists, decision scientists, social scientists and communications specialists, through to programme designers and evaluators. It should aim to meet basic research needs in risk and uncertainly analysis, risk perception, and risk communication as well as immediate policy goals—in effect operating as a “boundary organisation” between academia and public policy.24 It should be resourced so as to provide continuity of career progression for its scientists, alongside responsiveness to emerging risk communication needs. We suggest that good models for such an interdisciplinary boundary organisation might be the RAND Corporation (US), IIASA (Austria) or the Tyndall Centre (UK). If this seems challenging then we should not forget that risk communication has become central to a number of critical public policy issues, not just energy or climate change.

To what extent can public perceptions be changed by improving risk communication? (please provide examples)

32. In some more straightforward circumstances (such as health protection) we know that risk communication can work well if it focuses upon first understanding where a group holds key misunderstandings, or has gaps in knowledge, and targeting those with appropriate risk representations and messages.25 If behaviour change is also the goal, such communications also need to be supplemented by information on what practically one can do to avoid the risk.

33. Where legitimate value conflicts exist (as with many questions of energy infrastructure) the issue of “changing” perceptions becomes more problematic. Here one can easily cross into areas that look (to the recipients of any communication) like propaganda. Under such circumstances people may reject a message and/or question the motives of the communicator. Again, this is why it is better to think of risk communication as a process of participation, dialogue and decision support rather than a simple one-way transfer of information to persuade.

How do risk perception and communication issues in the UK compare to those of other countries?

34. We know that the very broad principles of risk perception, communication, and the development of risk controversies are likely to operate in similar ways across different nations—at least in Europe and North America where much of the key research has been conducted. However, as has been stressed throughout, context does matter. Therefore the details of the ways different issues manifest can differ significantly by country. For example, high societal trust in technical expertise has been one reason why in France nuclear power has historically enjoyed greater levels of acceptance compared to in the UK (a situation which may now be changing). Remaining sensitive to such nuances of context is one of the key aims of “strategic listening” noted in Para. 30 above.

Interest Statement: I currently serve as a member of the Chief Scientist’s Science Advisory Group (SAG) at the Department of Energy and Climate Change. The opinions here are, however, offered solely in relation to my professional academic position.

December 2011