Social Media-based Public Dialogue: Potential, Theory and Practice

Dr Eric Jensen Department of Sociology University of Warwick [email protected]

This independent report was commissioned by Sciencewise ‘the UK's national centre for public dialogue in policy making involving science and technology issues’. The views expressed in this report are the author’s alone, and do not necessarily reflect the opinions or views of Sciencewise.

Social Media-based Public Dialogue INTRODUCTION

3

FACING UP TO THE CHALLENGES OF DIGITAL ENGAGEMENT

4

OVERSTATING DIGITAL UBIQUITY THE DIGITAL DIVIDE STRUCTURAL BENEFITS AND LIMITATIONS OF THE MEDIUM IN OR OUT OF THE GOLDFISH BOWL OF ONLINE PUBLIC DIALOGUE?

4 5 6 9

SOCIAL MEDIA AND PRINCIPLES OF PUBLIC DIALOGUE

10

REPRESENTATIVENESS AND ROBUSTNESS OF SOCIAL MEDIA-BASED PUBLIC DIALOGUE

21

THE POLITICAL ECONOMY OF FACE-TO-FACE PUBLIC DIALOGUES LEVERS OF TECHNICAL EXPERTISE IN SOCIAL MEDIA-BASED PUBLIC DIALOGUE

22 23

CONCLUSION

26

CONCLUDING RECOMMENDATIONS

27

2

Social Media-based Public Dialogue Introduction Today, rather than meeting in a common physical space to witness an event or to engage in political debate, many citizens meet in symbolic, virtual spaces based on shared interests, needs or identities. Starting from the Sciencewisecommissioned report entitled, ‘In the goldfish bowl: Science and technology policy dialogues in a digital world’ (Latta, Mulcare, and Zacharzewski, 2013), this report explores the possibility of using existing forms of virtual assembly and discourse (i.e. social media) to either supplement or replace the face-to-face public dialogue exercises currently being conducted for UK government departments with support from Sciencewise. The report addresses the following research questions: 

How does the rapid global expansion of social media usage affect our understanding of the available means for conducting public dialogue?



Drawing from extant research and theory on the concept of the online public sphere (and deliberative democracy more generally), what is the potential for public dialogue to be conducted effectively through social media?



What can be learned about the use of social media for public dialogue from existing research on efforts to conduct public dialogues online, and through social media in particular? For example, what is gained or lost from moving public dialogue into this setting? To what extent is such dialogue already occurring within social media?



What are the particular characteristics of social media discourse, how is social media discourse different from in-person discourse and what are the implications of these differing characteristics for public dialogue?

To address the research questions stated above, a critical review of relevant theoretical accounts is employed - with special focus on the online public sphere. The goal is to identify the potential role for social media in a broader public dialogue context, which is likely to continue to include face-to-face engagement. Toward the same aim, this report also reviews extant research literature on online public dialogue - especially focusing on social media, such as Twitter. This report concludes that, if social media options for public dialogue are judged against the current criteria spelled out by Sciencewise, they are likely to fall short compared to face-to-face options. Yet these current dialogue criteria have important limitations that social media options could help to address. To clarify the relative advantages and disadvantages, this report compares social media-

3

based public dialogue to current Sciencewise-supported face-to-face dialogue practices and principles. It should be noted that, while public dialogue practice can certainly be improved, it is important to consider at least one specific case of face-to-face public dialogue as a real-life comparator for what would otherwise be an abstract, ungrounded analysis of social media-based public dialogue. The Sciencewise-supported dialogue selected to provide a comparative case study is on Synthetic Biology. While it not known whether this example represents general practices applicable to most Sciencewise-supported dialogues, there is no reason to assume it is an outlier (for example, it is delivered by a frequent commercial provider for Sciencewise-supported public dialogues). In summation, this project aims to establish a basis for policymakers’ decisionmaking when it comes to the use of social media in public dialogue, as well as highlighting important directions for future research and evaluation. The comparison between face-to-face and social media-based public dialogue running throughout the report is not intended to imply these approaches are mutually exclusive. In fact, I ultimately support using both approaches in tandem. The comparison between the two general approaches is developed for analytical purposes, to help policymakers weigh the potential value of shifting a portion of the formal public dialogue to social media. The report proceeds in three parts. First, a critical review of Latta et al is developed to ground my analysis in discussion with existing arguments on this topic. Second, a comparative analysis of social media versus in-person public dialogue is presented, engaging with the current Sciencewise definition of public dialogue. Finally, extant empirical evidence and theory on social media is presented, along with a discussion of the representativeness and robustness of social media dialogue. This is followed by concluding remarks and recommendations.

Facing up to the Challenges of Digital Engagement At the outset of this project, I was asked to engage with the Sciencewisecommissioned report entitled: 'In the goldfish bowl: science and technology policy dialogues in a digital world' (Latta et al, 2013). My responses to the ‘goldfish bowl’ report’s main points are reflected within this section. I suggest that readers review said report before proceeding beyond this point. The Latta et al report makes some very useful points, including rightfully pointing out the depth of knowledge that the digital world offers those who know how to find it. However, I would argue that they exaggerate the ubiquity of digital technology, overstate the benefits of adopting digital dialogue approaches and understate the potential downsides. I argue for an understanding of the role of online technologies for dialogue in contemporary societies that is better grounded in sociological research. While the limitations of digital public dialogue do not preclude using such approaches – either now or in the future – the limitations must be acknowledged in order to fully inform implementation.

Overstating Digital Ubiquity The report's conclusions about the potential upside of digital engagement

4

approaches are largely sound. However, the ubiquity of digital technology in public dialogue is simply overstated. From the outset, the Latta et al report argues: 'Technology is changing the way that all engagement between institutions and citizens is undertaken' (pg. 1). This exaggerated statement is clearly inaccurate: many institutions engage (or do not engage) exactly as they have always done, impervious to the technological changes going on around them. Some have explicitly decided to continue with their traditional engagement practices to limit the risk of disenfranchising those who are not confident users of digital technology, while others do not have the skills or vision to integrate technology in their engagement work. The report continues: ‘We believe that a digital approach can support good dialogue in two ways: firstly, online engagement around science-based policy will increase the ability of the public to participate in democratic discussion. Secondly, where specific exercises are planned, digital methods can expand the footprint of dialogue, involving more people and broadening the conversation’ (pg. 1). However, this optimistic perspective on the 'potential' for social media to support ‘good dialogue’ underplays empirical evidence on the ‘digital divide’ and the structural limitations of social media as a platform for public dialogue. The authors largely sidestep these limitations through the assumption that in the future (e.g. 2020) the diffusion of digital technologies will be so advanced that the issue of social inequalities in digital access and technical ability will be resolved. This utopian assumption enables the authors to overlook the unpleasant reality of current digital inequalities.

The Digital Divide The authors of the Goldfish Bowl report make an excellent point in highlighting the fact that online dialogue about current scientific issues is a continuous process: ‘On any issue, people are already talking about policy, sharing experiences and anecdotes, and spreading information of varying quality. Every dialogue exercise, whether digital or not, is undertaken against a background of digitally-mediated information and conversation’ (p. 4). Therefore, policymakers’ efforts to curate a public dialogue on current scientific issues will always be interjecting in a conversation that began before their arrival and will continue after their intervention. This then raises the idea of radically re-conceptualising the methods used for gathering public opinion (if this is a key aim of public policy dialogues). For example, for issues that are already widely discussed, why not simply commission a systematic analysis of online conversations already taking place, rather than constructing an artificial context for dialogue? Here it is worth considering who the public dialogue should be aimed at. Specifically, can it be representative of the broader population, or will it just

5

feature those who are politically engaged and technologically adept? The Goldfish Bowl report does acknowledge that digital engagement ‘can disadvantage non-users and unconfident users’, but does not pursue this point any further to identify its implications. The authors cite the general statistic that 73% of the British population currently use the Internet, and that 80% of under25s use social media weekly (pg. 2). The use of general, undifferentiated statistics fails to capture the ways in which people are using and engaging with social media, and fails to capture the full effects of the digital divide. Research conducted by Pippa Norris (2001) highlights three distinctive forms of digital divide: ‘global’ (between rich and poor nations), ‘social’ (inequality within a nation) and ‘democratic’ (between those who use digital technology for political, or ‘public’ purposes and those who do not). When debating the potential value of social media for public dialogue, it is important to consider these three digital divides. However, for the purposes of this report, the third is especially relevant. The claim that ‘it is impossible to ignore the trends that the Internet is becoming central to all of our lives’ (sic) is simply inaccurate; for a variety of economic, social and cultural reasons, the Internet plays little or no role in the lives of many individuals. How Internet users make use of the democratic opportunities the web affords is what characterises the democratic divide. Just because people are online or given the opportunity to engage in public dialogue does not mean they will take up this opportunity1. There is rational disengagement with, and intentional ignorance of, the political system. Moreover, there are sociological and political reasons for non-engagement amongst Internet users. For example, there may be systematic bias within the online public sphere regarding who is sufficiently networked and encouraged to participate in democratic dialogue. Any future use of social media to supplement face-to-face public discourse must take this into consideration.

Structural benefits and limitations of the medium Beyond my concerns with the overstatement of digital engagement’s benefits and the understatement of its limitations, it should here be noted that the headline claim, 'dialogue is an approach, not a delivery mechanism,' is problematic at best. ‘Approach’ and ‘delivery mechanism’ are not opposing ideas: Digital public dialogue methods can be both an approach and a delivery mechanism at the same time. As such, my critique of the stated limitations and benefits does not differentiate between the two. Below, I critically engage with each of the averred benefits and limitations identified in the report. Where it is possible, I note both the positive aspects and the negative aspects of the authors’ arguments. BENEFITS

1

An anonymous Sciencewise reviewer suggested that the CLEAR model developed by Professor Lawrence Pratchett could be helpful for informing the design of online engagement in a way that encourages wider participation.

6

• Digital engagement fits with the way science works today: Scientists are digitally adept. The authors claim may be true compared to the mean level of digital participation in the general population; however, the move from using digital technologies to serve technical purposes to using them for dialogue purposes may not be so straightforward for all scientists. I am not aware of any evidence, nor do the report’s authors provide any, suggesting that scientists have a greater propensity to use digital communication methods than other professionals. • Digital methods allow time-shifting: Allows greater public involvement by enabling flexibility about when and where a person contributes. I find this is a particularly good point. The potential flexibility surrounding digital methods enable those who are currently blocked from participation due to constraints on their schedule, location or physical disability to contribute at their convenience to public dialogue. It should be noted that not all approaches to digital engagement allow for time flexibility as some involve synchronous communication. Nevertheless, these approaches still benefit from location flexibility and elongation of the period of participation, both of which could substantially extend public involvement. • Digital information can be infinitely deep. Digital communications straightforwardly allow for a layered communication approach, which allows people to opt in to getting further and more detailed information. Again, this is a particularly valid point. However, as will be addressed further on, this infinite depth is not without its practical limitations. While individuals may opt into getting more information, encouraging them to do so may be harder in a digital context than it would be in-person. • Digital methods are more direct and personalise complex issues. The idea that social media offers 'disintermediation' (bypassing traditional media to directly connect policymakers, scientists and publics) is important, and correct. However, the examples cited to make this point in the report are not the most relevant to dialogue, as they do not demonstrate the twoway communication that this medium affords. They instead return to the traditional idea of only broadcasting ideas outward, stating that: 'Celebrity scientists who tweet, researchers who blog, and academics who develop outreach projects are all disseminating directly to the public; this makes research personalised and personal' (p. 5). This orientation towards one-way communication is also replicated in the authors’ response to Part 1 of my critique. They note that ‘all universities […] disseminate knowledge online’ to back up their claim that digital engagement practices have been universally adopted by institutions, saying ‘all universities […] disseminate knowledge online’. I would strongly resist the implication that institutions that ‘use computers or the internet’ are necessarily ‘engaging’ or promoting public dialogue. Indeed, far too many scientists and institutions are being drawn into a public relations approach to online communications, which

7

involves speaking but not listening. The benefits of disintermediation are greatest when two-way dialogue can occur. • Digital tools can allow you to understand your participants better. I agree that social media offer the potential for policymakers to develop a greater understanding of the perspective of individuals outside the policy context. The authors, however, signal an interest in identifying individuals who are powerful online in order to give their views more weight: 'social media profiling tools such as Klout and PeerIndex are crude but improving means of assessing the influence and authority of social media users' (pg. 6). Gerlitz suggests that these tools can be opaque, reductionist and potentially deleterious. Aside from their inherent methodological flaws, the suggestion that their use could guide digital engagement practices is concerning. Using such tools would suggest a problematic understanding of the nature and purpose of public dialogue, where the focus is on what is popular among the popular, and not on the wants, concerns, needs and values of a broader public. • Digital engagement can open the conference centre doors. The authors argue that making expert deliberations more visible online is an inherent good that automatically ‘creat[es] broader links that increase openness, approachability and transparency’ (pg. 6). This argument, although valid in principle, does not acknowledge the complexities of making expert deliberations visible in a way that the public can engage with. Without knowledge of where to find these deliberations online and how to apply some mechanism for responding to what they see, it is hard to pinpoint the benefits to most people of this 'openness, approachability and transparency'. Moreover, if this engagement process is managed in ways that are inaccessible or difficult to access, the overall impacts from the exercise could ultimately be negative. LIMITATIONS / RISKS 

Information quality. The authors state that, 'Information can easily go “viral” – particularly where inaccurate information has a strong simple story, and the reality is complex.' (pg. 6). The authors are correct here. As with traditional journalism, simple stories and explanations tend to get the greatest attention, and the digital sphere allows for ever quicker spread of rumour and false information. However, this is primarily a concern if one is working in a public relations context, trying to push a particular message out. Within a two-way dialogue, there are opportunities to clarify one’s meaning. In such a two-way dialogue, scientists and policymakers should focus at least as much attention on ensuring they understand public perspectives as on concern over their own views being oversimplified or misunderstood. Understanding why online publics hold specific viewpoints would open the space for negotiating a shared understanding on the topic.



Translating complexity into public debate. The authors state that:

8

'simplifying complex findings without losing accuracy is fundamental to supporting open debate' (pg. 6). Certainly, the ability to speak in plain language about the technical details of a topic is important to public dialogue, both online and offline. However, competent jargon-free science communication does not always translate into a better or more open debate. Particular social media platforms for dialogue can make substantive scientific debate more difficult. Twitter, for example, places strict limits on message length. This limit can compel users to (over)simplify their messages, which in turn increases the importance of the background knowledge individuals bring to the conversation. Dialogue on a site such as Twitter can inadvertently create parallel conversations because of the limited space for explanations. Therefore, it is important to ask for clarification when ambiguous statements are made. 

Creating a ready audience. The authors state that: 'Enhancing dialogue with digital tools should aim to give participants a baseline of knowledge, not a baseline of opinion'. This point seems to be based on the idea that the public must first reach a knowledge threshold deemed 'informed' before being worthy of engagement (this issue is discussed in more detail later in this report). This can be problematic as everyone is to some extent informed and ignorant about different topics and different elements. Instead, it is important that policymakers and scientists are honest about the boundaries and limitations of their available evidence on a topic.



Building trust, online and offline. 'The views of a trusted member of a chatroom, online group or social network may command more weight than that of an official spokesperson or expert scientist' (pg. 7). Indeed, this is a general feature of social reality (offline as well as online): The views of individuals in people’s immediate social circle will often carry greater weight than an official or expert. When members of the public (often rightly) suspect that officials and experts do not understand them or their situations, it is understandable that expert views are not given paramount consideration. Experts should work to make themselves trustworthy and to maintain an on-going humility about the boundaries of their knowledge. They should also recognise the value of non-experts’ experience, and expertise. Again, this applies both online and offline.

In or Out of the Goldfish Bowl of Online Public Dialogue? In summarising this review, it is perhaps useful to look toward the framing device used to ground the report. The authors provides very little elaboration of the 'goldfish bowl' metaphor that is included in the title. They simply articulate the metaphor as follows: ‘The core benefit of digital engagement as an enhancement to dialogue is that it allows dialogue to take place in a “goldfish bowl” – visible to the outside world but separate from it. In the world outside the fishbowl, separate discussions and communications take place that boost the impact of the exercise.’ (pg. 13).

9

The authors are certainly onto something with this metaphor, but it has important limitations. Firstly, it is important to consider who becomes the focus of attention (i.e. who is inside the goldfish bowl) in a public dialogue, including one taking place online. Contrary to the authors’ recent response to Part 1 of my critique, I do not think the significance placed on digital technology is the most important limitation of their report. Rather, it is the differentiated nature of digital exclusion that is not acknowledged, and which could most undermine the credibility of online public dialogue. The 73% of the UK population who currently use the Internet are not evenly distributed across the population: the poor, elderly and less educated are less likely to have access to it. Therefore, whose voices are heard? Which discussions are included, and why? These kinds of inclusion/exclusion dimensions are essential to understanding the validity of online public dialogue in practice (i.e. not just its potential). Secondly, we cannot take for granted the permeability of the membrane between what has been deemed the site of online public dialogue on one hand and the grand cacophony of voices that is the broader social web on the other. Simply saying that discussions outside of the boundaries of the online public dialogue will ‘boost the impact of the exercise’ begs the question of how precisely a two-way stream of communication can be maintained in this circumstance. Policymakers wishing to engage with online public dialogue should consider whether it is truly necessary to create a separate space or pathway to inform social policy, or whether they should go where the dialogue is already happening on the web. This ‘naturalistic observation’ of incorporating public perspectives on specific issues could easily be more inclusive in terms of the number of participants. If rigorous social scientific methods of analysis were employed, the exercise could be more systematic than current face-to-face practices. It would also place effectively no demands on participants’ time, and be far cheaper and more straightforward to conduct. There are certainly limitations to such an approach, especially when attempting to gather feedback on issues that are not yet widely known. However, given the greater cost of managing a dialogue ‘within the goldfish bowl’, the burden of proof lies with those advocating such a managed approach to demonstrate its superiority. Of course, it would be feasible to employ both approaches within a single digital public dialogue initiative. Qualitative analysis of naturalistic observations of on-going social media discussions could inform the issues raised in a managed public dialogue. The findings could also help place the managed deliberative exercise within a broader context.

Social Media and Principles of Public Dialogue In looking to answer the questions stated at the outset of the report, and posed during the critique of the Goldfish Bowl report, it is useful to draw upon a comparison between current Sciencewise-supported face-to-face public dialogue activities and the possible use of social media-based public dialogue. As I suggested in my discussion thus far, there are two main points that need to be taken into consideration when looking specifically at social media. Firstly, in principle, the expansion of social media means that it is a busier, more diverse 10

and more pervasive setting for dialogue than ever before. Secondly, in practice, there are still many limitations to relying upon social media as the sole, or even primary, venue for public dialogue. In order to make comparisons, let us begin by thinking about what we are trying to achieve with public dialogue, and then weigh up the relative advantages of social media or conventional face-to-face approaches for achieving those goals. This can then serve as the jumping-off point for a more in-depth critique of the two methods, based upon theory and extant empirical research. In the analysis that follows, the Sciencewise definition of public dialogue is used to assess whether social media-based public dialogue is commensurate with face-to-face public dialogue. Given that any method of facilitating public dialogue will have limitations, I present this assessment (Figure 1) as a comparison with current face-to-face public dialogue approaches.

11

Figure 1: Principles of Public Dialogue: Comparing Face-to-Face and Social Media Approaches Topic Gaining participants for public dialogue from key backgrounds

Dialogue Principle ‘Members of the public, policy makers, scientists and other expert stakeholders’ are identified as the key players.

Face-to-Face In terms of inclusion of policymakers and experts, this is a key strength for the current model of public dialogue supported by Sciencewise. In terms of public participation, the level of inclusion may vary. Active, and often paid, recruitment required. Must live near enough to the event to feasibly participate. Current Sciencewise-supported practices are likely to exclude rural populations2. Problematic sampling practices (viz. convenience / quota sampling) often used to obtain easiest access to individuals for events (e.g. Synthetic Biology and Science Policy Horizon Scanning dialogues).

Diverse participation

‘Diverse mix of public participants with a range of

There are major limits to the diversity of faceto-face public dialogues. The framing of this target around

Social Media Lowers barrier for participation, thereby making it easier for interested parties to get involved The ‘lowered barrier’ could decrease quality of public dialogue that takes place Individuals can more easily claim to be experts without really being experts There may be a less influential platform for expertise in the dialogue Passive recruitment possible Must have technology and skill

The internet as a site for public dialogue raises concerns primarily based on class. Due to certain

Following a comment on this point by an anonymous Sciencewise reviewer (denying that home location was an exclusionary factor, saying ‘not true or an issue’), I selected four additional cases using a random number generator to check for locations of face-to-face public dialogue events. This further research supports my original point. The Synthetic Biology dialogue events took place in: London, Llandudno, Newcastle-upon-Tyne and Edinburgh. The further locations were London and Newcastle (Animals containing human material dialogue); London, Cardiff, Bristol, Newcastle and Edinburgh (Stem Cell Dialogue); Swindon and Manchester (Open Data Dialogue); London, Manchester and Cambridge (Science Policy Horizon Scanning Dialogue). 2

12

views and values’

Diverse range of views and values represented

‘Diverse mix of public participants with a range of views and values’

‘diverse views and values’, rather than diverse people representing different social and ethnic backgrounds downplays what is a major limitation in current practice.

ethnic and immigrant communities facing particular economic hardship and discrimination, the class basis for the digital divide can spill over into a lack of ethnic diversity.

The lack of detail provided about precisely how and where people are recruited for Sciencewise-supported public dialogues (e.g. Synthetic Biology dialogue) clouds this issue considerably.

The comparative question of whether this lack of diversity in social media participants is more acute than the formalised face-to-face dialogues that have been conducted under the auspices of Sciencewise is an empirical one that has not been assessed to date. An initial comparison between social media audiences in general and face-toface dialogue events could be conducted if more consistently detailed demographic data were provided in Sciencewise-supported dialogue reports.

It is unclear to what extent diversity is being achieved in terms of the range of participants. There is certainly diversity in the views and values expressed by public participants during Sciencewise-supported public dialogues.

There is a real risk that the range of views and values represented in a social media-based public dialogue activity will be restricted, given the unrepresentative demographic characteristics of social media users. As above, it not currently known whether the potential for a restricted range of views is more acute on social media than the current Sciencewise-supported face-to-face options.

The Synthetic Biology public dialogue suggests some basis for concern on this criterion as details of precisely where

13

participants were recruited from and the demographic make-up of the full participant list are not provided in the full report or any of the methodological appendices. There is also reason for concern based on the thin ways in which participants’ views are summarised in the report). Public participant learning takes place

[Public participants] ‘learn from written information and experts’

Face-to-face public dialogue allows for the demarcation of expert and public, enabling the latter to learn from the former following an organised format. As the information/expertise is controlled by those delivering the public dialogue activity, there is a risk of systematically biasing the dialogue findings through this ‘learning process’ Some face-to-face public dialogue exercises make a concerted effort to include a diverse range of views and values by drawing upon activists or other groups with diverse views, defined as stakeholders or experts. However, from the Sciencewisesupported dialogue reports that I have seen, scientists’ perspectives are by far the most heavily weighted.

There is a risk in online public dialogue that expert and public are not delineated. This can break down the ability for experts to air their views in a specified time frame and therefore make it less likely public participants will undergo as much learning from institutionally verified experts. On the other hand, participants are more likely to draw their information and influences from a broad range of sources, thereby reducing the risk of systematically biasing the discussion through the selective presentation or overweighting of scientific expertise.

14

The Synthetic Biology public dialogue, for example, included ‘expert’ views from scientists and social scientists. However, there were no NGOs or other stakeholders involved as ‘experts’. Views are effectively exchanged and developed

[Public participants] ‘listen to each other, and share and develop their views’

Depending on the format of face-to-face dialogue, public participants are strongly encouraged to listen to the views of others. If the dialogue workshop is effectively facilitated, this deliberative process will lead to mutual understanding as participants’ views on the topic evolve.

Final conclusions are wellconsidered

[Public participants] ‘reach carefully considered conclusions’

Many forms of face-toface dialogue strongly encourage participants to develop their views after considering a range of perspectives and information sources. Indeed, facilitating a careful consideration of all the major issues relevant to a topic is probably what face-toface public dialogue is

As much social mediabased dialogue is conducted on the basis of an imagined audience, online participants may well construct their comments on the basis of an imagined, rather than known, listener3. However, as online discussions are conducted at home, or at least not in public, participants are, in principle able to consider their views at greater length. Whether or not this actually occurs is an open question. In most forms of online dialogue, individuals have a much greater degree of anonymity than in face-to-face interactions. This means that there is likely to be a range of both ill-considered and well-considered conclusions represented in social media-based public dialogue. Finding ways to encourage deep consideration and

See: ‘I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and the Imagined Audience’. (2010). New Media and Society. (http://nms.sagepub.com/content/early/2010/06/22/1461444810365313). Of course, as an anonymous Sciencewise reviewer pointed out, ‘there are exceptions (e.g. ePanel, online citizen jury)’. 3

15

best suited to deliver (at least in principle).

deliberation within the constraints of social media would be an obvious next step if this new platform for dialogue were to be adopted.

Sharing conclusions to inform policy

[Public participants] ‘communicate those conclusions directly to inform Government’s decision making’.

Structured face-to-face public dialogue exercises supported by Sciencewise are directly connected to government decisionmaking processes. This is essential to ensuring their conclusions inform government decision-making.

Unless there is a prejudice in government against insights derived from public dialogue conducted through social media, there is no reason to believe it is an inferior option on this criterion. Indeed, the potential for much larger numbers of participants and greater transparency may lend added credibility to conclusions from policymakers’ perspectives. On the other hand, the lack of formalised processes for framing the dialogue around scientific expertise may undermine the findings from the perspective of the more technocratic government departments.

Better understanding of public aspirations and concerns provided to Government and policymakers

‘Good public dialogue’ should ‘help policy makers and Government to […] gain a rich understanding of public aspirations and concerns that goes beyond media headlines or focus groups’

This largely depends on the competence of the agency/people selected to conduct the public dialogue activities, as well as the resources and restrictions they have been given to complete the task.

If we assume maximum competence and a reasonable investment in resources without onerous restrictions imposed on the dialogue, then the social media option is likely to be inferior on this dimension. A lack of depth and sustained engagement is evident in most social mediabased discourse, which

The process of distilling ideas discussed in face-toface public dialogues

16

for reporting to government and policymakers could result in a poor understanding of public aspirations and concerns. On the other hand, a competent delivery, analysis and reporting of a public dialogue exercise following robust social scientific methods could deliver greater insights than ‘media headlines or focus groups’. For example, based on my extensive experience conducting, reviewing, and critiquing focus group research, it is clear to me that the Synthetic Biology report falls far short of the depth that could be achieved with a competent focus group study.

would militate against ‘rich understanding’. Likewise, this option would most likely not outperform focus groups in terms of depth. However, where the analytical competence evident in the reporting by those delivering the public dialogue is limited (e.g. the Synthetic Biology dialogue), the sacrifice of depth with social media may be negligible. Where the level of depth allowed in the reporting of results is restricted for any reason, the sacrifice of depth involved in choosing a social media-based approach would similarly be mitigated.

On way a robust focus group study would differ from the Synthetic Biology example is by displaying richer detail about public participants’ thinking and any implicit patterns that underpinned such thinking.

The analysis above suggests that social media-based public dialogue is an inferior option in some respects and a superior option in other respects. As such, those commissioning any particular public dialogue exercise or analysis would need to decide which limitations were most acceptable when deciding whether a social media or face-to-face approach would be most appropriate. One key factor that is hard to predict in advance is the distinction between which option is best ‘in principle’ versus ‘in practice’. If we assume highly competent delivery for either face-to-face or social media approaches, then a formalised approach along

17

the lines of current practices is likely to be superior, based on the public dialogue goals outlined by Sciencewise. If, however, we assume that the relatively thin analysis and problematic recruitment strategies represented in the Synthetic Biology4 public dialogue would be the standard competence level in practice, then the putative advantages of a face-to-face approach using current Sciencewise-supported practices quickly evaporate. That is, the relative strength of face-to-face public dialogue in principle is depth; however, if this strength is not delivered in practice, then much less expensive options, such as using social media, would produce comparable results. This, and the other conclusions indicated within this portion of the report, are explored in greater depth in the following sections.

An Empirical Case of Social Media-based Public Dialogue In this section, I discuss an empirical analysis of online public dialogue as a way of further examining the benefits and limitations of such an approach. Ingham (2013: 1) conducted a mixed methods study of venues for online public dialogue, including the Facebook page of the European Parliament. The results showed that the online discussions were ‘more competitive and argumentative than deliberative’ (Ingham 2013: 1). It also found that ‘the degree of interactive debate was very low with users showing little respect for each other where they interacted at all’ (Ingham 2013: 1). This finding highlights the value of a facilitated approach to public dialogue (such as the approach followed by Sciencewise-supported face-to-face dialogue activities), which can be structured to enable mutual listening and understanding. Such a facilitated approach may be transferrable in some respects to a social media context, but this has yet to be demonstrated. From the perspective of a public dialogue model influenced by German social theorist, Jürgen Habermas (i.e. the Sciencewise model), a particularly negative finding is the lack of ‘reasoned debate with justifications for arguments’ (Ingham 2013: 1) within the unmoderated social media discussions. While Ingham found no indication that debate participants shifted their positions because of their social media-based interaction, such interaction can nevertheless inform and influence the opinions of others viewing, but not directly participating in, the debate5. In sum, ‘the rules of the sites, the overall structures in which debate takes place and the cultures of the participants which value free expression and the ability to state a diversity of views over deliberative ideals’ (Ingham 2013: 1) would not fit well with the Sciencewise definition of public dialogue. This study highlights the dominance of ‘agonistic pluralism’ in social mediabased public dialogue ‘where users go to declare views in a competitive environment’ (Ingham 2013: 1). This kind of non-facilitated social media discourse is likely to be viewed as problematic based on the current Sciencewise model. In principle, however, a well-structured formalised public dialogue 4

http://www.bbsrc.ac.uk/web/FILES/Reviews/1006-synthetic-biology-dialogue.pdf

See: http://www.academia.edu/1220569/Social_Media_in_Higher_Education_A_Literature_Review_a nd_Research_Directions 5

18

exercise conducted face-to-face could instead produce more constructive and cooperative forms of discussion. Ingham has examined a very specific, relatively informal and unstructured form of online public dialogue. Therefore, just as structured and formal face-to-face dialogue may address Ingham’s conclusion, so may more structured online dialogues.

Social Media-based Public Dialogue in Theory Moving beyond the empirical findings above, it is important to critique the underlying theory driving current public dialogue practices, and their implications. It could certainly be argued that social media-based public dialogue would give a greater voice to those whose views and values are already taken seriously by news media and politics. However, this is currently the case with many face-to-face public dialogue activities supported by Sciencewise, which have an unrealistically high burden for participation. Large swathes of the population do not live near the selected urban centres for the workshops and thus cannot possibly join in6. Indeed, this kind of problem highlights one of the inherent limitations in Habermas’s notions of the public sphere and the ‘ideal speech situation’, which implicitly undergird the notion of public dialogue advocated by Sciencewise. In a nutshell, the framework Habermas propounds is not feasible in the real world, nor is it really desirable when the goals of public dialogue, as framed by Sciencewise, are put into practice in real-world contexts. Habermas’s vision is a ‘public of private people making use of their reason’ (Habermas 1989: 51) and discussing issues from a position of cool detachment to enable the triumph of the better argument. However, there are numerous critics of Habermas’s influential view, which also apply to Sciencewise’s current framing of public dialogue. Understanding the limitations of the theory underpinning Sciencewise’s dialogue model helps to reveal an alternative approach, based on pluralism, which would be better served by social media-based dialogue. Critics focus on how Habermas’s model tends to favour those with the education and resources to be able to express their views in rational terms and with a measured tone (e.g. Eley 1992: 304). Habermas’s model has been further challenged by Ryan (1992: 285) because it tends to lead to ‘spatial or conceptual closure[, and] constrain[s]…the public to a bounded sphere with a priori rules about appropriate behavior’. This critique clearly applies to Sciencewisesupported face-to-face public dialogue activities, which limit governmentrecognised public dialogue to particular physical spaces and formats. In contrast, Ryan advocates a ‘plural and decentered concept of the public’ based on the principle that ‘notions of interest and identity need not be antithetical to public good’. Finally, she emphasises the inadequacy of a public realm dominated by bourgeois values, pointing out that those most disenfranchised and excluded from policy discussions tend to express their views in a manner that is viewed 6

In the case of a public dialogue exercise that did actually specify precisely where participants were recruited from, it was the streets of the cities where the workshops took place (no indication of which streets, which parts of town, etc.). This recruitment method means that those not physically in the city at the time of the recruitment activity could not be selected for participation.

19

with disdain by government technocrats (cf. Baumann 1996). Indeed, the protests of the disenfranchised are often perceived as ‘loud, coarse, and, yes, abrasive’ (Ryan 1992: 286). Young (2000: 178) further highlights the difficulties facing the disenfranchised: It often takes considerable organizing, dramatic action, and rhetorical shrewdness for people whose concerns are excluded from that agenda to break through and gain access to [means of]…disseminat[ing] their issues so that state institutions eventually deal with them. These criticisms of Habermas’s theory also apply to the ascendant form of public dialogue supported by Sciencewise, as well as the empirical study by Ingham reviewed above. In a similar vein, Benhabib (1992) criticises Habermas’s (1989) narrow and exclusionary conception of public discourse as being limited to critical-rational debate over ‘public’ issues. She points out that it is completely legitimate to make arguments based on one’s position in society rather than cold, rational deliberation. Moreover, Habermas’s theory ignores those with ‘voices that are mute’ within public discourse: ‘those who do not or cannot speak in public, who from inarticulateness, fear, habit, or oppression are removed from participation in public life’ (Gould 1996: 175-176). Gould (1996: 176) attributes this exclusionary current in Habermas’s theory to his ‘exclusively discursive’ conception of rationality and his reduction of ‘communication’ to ‘rational verbal discourse’. Clearly, this entire line of critique focusing on Habermas’s theoretical model also applies to the Sciencewise framework for deliberation through faceto-face public dialogue exercises. The practice of Sciencewise-supported public dialogues is also subject to this line of critique. Insofar as the current practices are mapped onto social media-based public dialogue activities, there may be similar issues affecting social media dialogue as well. In principle, however, engaging with the natural flows of informal dialogue taking place on social media would be more pluralistic and open, and in many ways would address the critiques of Habermas stated above. Social media participants could be empowered to bring in different points, images, etc. to contribute to the discussion. Political theorist Nancy Fraser (1992) criticises Habermas’s emphasis on consensus developed through reason as the primary objective of deliberation. She argues that his normative model of a single, consensus-oriented discussion would result in the suppression of dissent and the homogenisation of public policy discussions. While the Sciencewise framework does not explicitly require consensus, the presentation of findings in Sciencewise-supported public dialogue reports often imply uniformity of viewpoint by not displaying dissent from the discussions. Indeed, the Synthetic Biology public dialogue report7 largely provides singular summaries, giving little hint of pluralism in most of the report.

7

http://www.bbsrc.ac.uk/web/FILES/Reviews/1006-synthetic-biology-dialogue.pdf

20

This line of theoretical argument recalls classic social theorist Hannah Arendt’s (1958: 220) contention that modernity views plurality as a threat to order and profitability because of its inherent ‘haphazardness’ and perceived ‘moral irresponsibility’. Structures, such as the Sciencewise public dialogue framework, impose order on the haphazardness of public views from start to finish. The current practices of formalised face-to-face public dialogue restrict the range of participants and controls the information provided to participants. The structure of the discussions are pre-determined and the depictions of public views are limited to those provided by the (usually) market research companies hired for this purpose. This practice appears to be designed to limit the extent of uncontrolled pluralism in public perspectives that gets funnelled through to decision-making by government departments. Key theorists, such as Fraser, favour an alternative framework in which pluralism infuses public debate with a rich variety of voices and perspectives mirroring the diversity of publics in modern societies. In practice, this requires allowing public dialogue to be infused with the widest possible range of perspectives. This pluralistic model could be advanced by the wider pool of potential participants in a social media-based public dialogue. Fraser (1993: 14) argues that ‘arrangements that accommodate contestation amongst a plurality of competing publics better promote [Habermas’s] ideal of participatory parity than does [the] single, comprehensive, overarching’, rational framework that Habermas himself proposes. In many ways, social media offers a platform for public dialogue that is much better aligned with a pluralistic model. Yet, without active facilitation, a social media approach to public dialogue would fall well short of the requirements of Habermas’s model and the concomitant definition of public dialogue employed by Sciencewise.

Representativeness and Robustness of Social Media-based Public Dialogue One concern that is likely to be raised when considering the value of social media as a new site for public dialogue is whether public views are gathered in a way that is representative and robust. On this methodological issue, it is again instructive to compare the social media option to existing formal UK public dialogue exercises supported by Sciencewise. In principle, participant recruitment in face-to-face public dialogues could develop a much more diverse and inclusive pool of participants than social media-based public dialogue, where the level of control over who participates is likely to be much lower. Therefore, this criterion is potentially a major strike against the idea of moving aspects of policy-linked public dialogue onto a social media platform. This potentially supports current Sciencewise dialogue practices, but, as with other such dialogue criteria, it may not hold up to close scrutiny of actual practices. To take the example of the Synthetic Biology public dialogue exercise8, the method of gathering participants was described as follows:

8

http://www.bbsrc.ac.uk/web/FILES/Reviews/1006-synthetic-biology-dialogue.pdf

21

Participants were identified through qualitative free-find techniques, with the following recruitment procedures: After providing details of the research, individuals were asked to undertake a short screening questionnaire to assess eligibility and ensure that the designated quotas are accurately filled. If they met the requirements they would be invited to participate in the research. (Synthetic Biology Dialogue; p. 5). Details of participant recruitment (i.e. stating where precisely participants were found) are not described anywhere in the report or the methodological appendices. The actual demographic breakdown of participants for all relevant categories is not specified anywhere either (the interim evaluation only provides this detail for the variable of gender). Similarly, the claim that robust qualitative analysis has been conducted is not born out by the cursory and thin descriptions of public views expressed during the dialogues, which are provided in the Synthetic Biology dialogue full report9. The use of quota sampling in this dialogue project (and others) was inadequate as a recruitment procedure. While the sampling plan sounds representative and diverse on the surface, quota samples (as in the Synthetic Biology dialogue) tend to be drawn from conveniently accessed locations, thereby undermining the diversity of the recruited participants. If other formal face-to-face public dialogues have similar limitations in their recruitment practices, the gap in representativeness and inclusivity between face-to-face and social media-based options may in fact be relatively small.

The political economy of face-to-face public dialogues Buried in the interim evaluation report on the Synthetic Biology public dialogue exercise is this statement, which addresses the tendering process that led to the selection of a large market research company to deliver the dialogue: ‘Some concerns about the tender process were expressed by one member of the Steering Group, and it was suggested that the process excluded small academic groups’ (http://www.bbsrc.ac.uk/web/FILES/Reviews/synbio-dialogueinterim-evaluation.pdf; p. 12). This concern seems to have been warranted given the ultimate selection of a large commercial provider for this particular dialogue. Indeed, an admittedly hasty survey of the public dialogue exercises described on the Sciencewise website (as of March 2014) indicates that this same well-connected market research company (TNS-BMRB) was awarded at least 6 public dialogue contracts. The major market research company, IPSOS MORI, was awarded at least 12 public dialogue contracts. Three affiliated mid-size consultancies, OPM (n=6), 3KQ (n=6) and Dialogue by Design (n=6), were awarded at least a further 18 public dialogue contracts between them (as well as evaluation contracts). Another market research consultancy was awarded at least three dialogue 9

http://www.bbsrc.ac.uk/web/FILES/Reviews/1006-synthetic-biology-dialogue.pdf

22

contracts. Meanwhile, just one public dialogue contract appears to have been awarded to a university to be led by an academic social scientist10. This kind of de facto skew in (or perhaps preceding) procurement towards commercial providers has consequences for the quality and robustness of public dialogue activities. As far as I am aware, not a single public dialogue conducted by a commercial provider has had results published in any peer-reviewed journals or in any other similar, quality-controlled, venues. Moreover, the methodological robustness and theoretical engagement evident in the work is of poor quality. From my perspective as a social scientist teaching social research methods at the University of Warwick, the public dialogue reports produced by TNS are exceptionally weak examples of qualitative research. This low quality level means that there is little evidence of the superior depth that is supposed to develop from using a costly and intensive face-to-face public dialogue approach. If similar providers continue to be used for social media-based public dialogue, there would be little distinction in depth between the two formats in terms of the visible outputs in the form of reports. Such limitations in current practices are directly relevant to considerations of new approaches, including the use of social media. If the current practices were achieving all that was promised, in terms of depth of analysis and insight, then there would be a much stronger case for avoiding such a radical shift as moving aspects of formal science policy dialogue online.

Levers of Technical Expertise in Social Media-based Public Dialogue Aside from issues of who is performing the research, there is also the issue of who is participating in the research. ‘Members of the public, policy makers, scientists and other expert stakeholders’ are identified as the key players by Sciencewise. Given the power imbalances that exist within and across these four categories, the way in which power and expertise is managed in public dialogues relating to science and technology is vital to the question of just how deliberative these exercises are. This issue of the most effective relationship between expertise and non-experts has been subjected to intense academic scrutiny and debate in the field of science and technology studies. A major intervention in this debate came from Collins and Evans, who advocated an approach that privileges expertise. Collins and Evans’ (2002) favour using expertise as the only legitimate basis for participating in ‘technical decision-making’, by which they mean political deliberations about aspects of techno-scientific development, such as those covered by Sciencewise supported public dialogues. Their thesis is that rather than aiming for a radical democratisation of science governance ‘by dissolving the distinction between expertise and democracy’, we should preserve expertise as the primary focus. This would involve ‘recognizing and using new kinds of expertise emerging from non-professional sources’ (Collins and Evans 2002: 269). As such Collins and Evans would be supportive of the kind of structured dialogue that Sciencewise co-funds, in which the public is required to learn from experts in order to become informed, and therefore legitimate, sources of insight 10

http://www.sciencewise-erc.org.uk/cms/aims-and-objectives-3

23

for public policy. Although it is not explicitly acknowledged, the standard processes that Sciencewise prescribes for public dialogue are implicitly based on a permutation of this notion that technical expertise is required for legitimate participation in deliberation about technical issues. The requirement that participants in a public dialogue should ‘learn from written information and experts’ implies that they do not already have a legitimate basis for participation prior to the formally organised deliberative exercise. This exercise is implicitly intended to create an ‘informed’ (by experts) public that can be a legitimate participant in the political process. The video on the front page of the Sciencewise website11 features Sciencewise ‘steering group chair’ Kathy Sykes discussing public dialogue. She explains: ‘The key thing about doing dialogue well – and what we [Sciencewise] mean by doing public dialogue is […] for there to be time for the public to become really informed about an issue so [it is] deliberative. People have a chance to speak to scientists and to hear other perspectives: Hear from journalists, from ethicists, from many different points of view. So they really have the chance to become informed. That matters if we want people to be able to make wise comments that are going to be useful for policymakers’. There are two points to be made here: One general and one particular to the topic of social media-based public dialogue. 1. The current Sciencewise-supported approach to public dialogue exercises runs the risk of inducting public dialogue participants into the experts’ ways of seeing the issues. Because this educative dimension of formalised public dialogue exercises takes place prior to the ‘listening’ element, it risks reducing the diversity of perspectives represented in conventional public dialogue exercises. It is noteworthy that, while publics are supposed to learn from experts in formal public dialogues, there is no requirement for experts to learn anything from publics. This reinforces a position of power for institutional expertise. Given that the theory underpinning public dialogue calls for everyone to be able to participate in deliberation on an equal footing, the current model of face-to-face public dialogue supported by Sciencewise fails in practice to meet this important criterion of deliberation. 2. Social media-based public dialogue would probably need to follow a credo of ‘come as you are’ in terms of public participation, with relatively limited opportunities available for policymakers to ensure that dialogue participants are ‘informed’. While this is a problem from the perspective of the current model of public dialogue advocated by Sciencewise, it is likely to be a more democratic and pluralistic framework that has some support in theoretical literature on democracy and the public sphere. 11

Last accessed 4 April 2014, http://sciencewise-erc.org.uk/.

24

However, unless active recruitment procedures of some kind were employed, there is a risk that only those with an existing interest or view on the topic would join in.12 Most participants in public dialogue on social media are unlikely to possess technical experience or expertise on the topic of the public dialogue sufficient to be considered ‘informed’ by experts. Thus, the majority of social media-based dialogue participants would be marginalised from the discussion following Collins and Evans’ (2002) framework for public deliberations over scientific issues. Their model – like that of Sciencewise – privileges scientific rationality and expertise as ‘the proper, “natural” frame[s] of reference’, while only valuing non-expert, technically inexperienced laypeople’s perspectives, insofar as they can demonstrate they are scientifically ‘informed’ (Wynne 2003: 404). Indeed, Collins and Evans’ (2002) uncritical privileging of technical expertise is highly questionable: Expertise is constituted within institutions, and powerful institutions can perpetuate unjust and unfounded ways of looking at the world unless they are continually put before the gaze of laypersons who will declare when the emperor has no clothes. (Jasanoff 2003: 398) This more powerful role for laypersons is diminished by a model of public dialogue that requires that they first become inducted into a scientific understanding of an issue before considering their own perspectives. Indeed, it could be argued that in face-to-face public dialogues - such as those supported by Sciencewise (represented by the Synthetic Biology exemplar) - the interests of those with the most expertise, institutional support and symbolic capital are best serviced, while the democratic discourses of the underprivileged and socially excluded are invisibly redacted or comparatively marginalised. In sum, the current approaches employed by Sciencewise-supported dialogues are more in the republican tradition of allowing avenues for democratic participation, while spurning outright democracy in all its messy diversity. There is good reason to argue that social media-based public dialogue would be more in keeping with a radical democratic agenda based on a pluralistic model of deliberation. That is, opening up an avenue for the kinds of discussions members of the public have about an issue on social media allows for them to make their way into policy discussions. Furthermore, such expanded participation need not take the form of the anarchical, irrational discourse that seems to be implicitly feared by Collins and Evans and the current Sciencewise framework. Social media offers possibilities for facilitating dialogue, whilst allowing for much greater breadth of participation than many forms of formal face-to-face deliberation. Social Media-based Public Dialogue for Viewpoint Pluralism 12

In practice, because of the recruitment procedures that are used, the current public dialogue practices of Sciencewise-supported exercises are much more exclusionary than they could be in principle. These problematic practices include convenience and quota sampling conducted by market research companies (which limit the likelihood of including the voices of the poor and disenfranchised). To ensure greater inclusivity, there would need to be a concerted effort to reach out and involve poor and disenfranchised participants.

25

The Habermas-inspired (1989) model of public dialogues, including only ‘disinterested’ publics making rational arguments, risks excluding the views of those who cannot physically participate or who are not asked to participate by those conducting the dialogues. Fraser (1992), Ryan (1992), Benhabib (2002; 1992), Young (2000) and others emphasise the contribution of interested individuals and groups who have been marginalised or excluded from political decision-making. These theorists advocate a ‘pluralistic public sphere’ built upon principles of open access and raucous, agonistic discourse in which everyone has a voice. Far from Collins and Evans’ (2002) demand for swift closure of the ranks of possible participants in public debate over scientific issues (see Rip 2003: 423), the ideal of a pluralistic public sphere calls for open participation and inclusion at all stages of decision-making. To avoid the ‘peculiar evil of silencing the expression of an opinion’ (Mill 1859: 24), public dialogue in science should be a ‘more inclusive social debate over [scientific] knowledge and its proper grounds and human purposes’ (Wynne 2003: 408). Wynne (2003: 408) calls on governments to ‘open up spaces, now colonized by existing scientific culture, to collectively negotiate questions of public meaning’. If joined with face-to-face exercises designed to extend beyond the already enfranchised, well educated and well resourced, social media-based public dialogue might be effective at charting a new, more pluralistic course between the ‘Scylla of public disillusion and the Charybdis of technical paralysis’ (Collins and Evans 2002: 272). Social media offers the possibility of preventing the marginalisation of certain voices within public dialogue by ensuring that participants are included regardless of their relationship to a pre-existing power structure or proximity to a market research company. Yet, with creativity and effort, it also offers the possibility of facilitating dialogue at a distance to bring social media-based public dialogue more closely in line with the current Sciencewise framework for faceto-face public dialogue.

Conclusion There are a number of objections that could be made about moving public dialogue into a social media setting, and the first part of this report did just that. However, after analysing these objections, this report has found that the limitations of such a shift towards social media-based public dialogue are different to, but not greater than, current practices in formal face-to-face public dialogue exercises. Objections to social media-based public dialogue could include:

Objection The digital divide means that whole swathes of the population would

Response Indeed, this is a major limitation of social mediabased public dialogue, which means that alternative engagement methods are needed to reach the poor and disenfranchised. However, 26

not be able to participate in such a dialogue.

current approaches to public dialogue using face-to-face methods typically fall foul of this same objection when their methods of recruiting participants are scrutinised.

Social media-based public dialogue would yield findings that are not representative or robust.

As discussed in the body of this report, the representativeness and robustness of the most recent practice in face-to-face public dialogues is dubious. Like face-to-face dialogue exercises, the key will be the quality of the facilitators or analysts commissioned to run the exercise, as well as how the dialogue online is structured or facilitated. Trialling different techniques for social media-based public dialogue in advance of a full-scale formal departmental public dialogue would be prudent. However, such trials must employ robust social scientific evaluation methods if they are to yield accurate insights.

The voice of experts will be drowned out online.

Certain forms of social media dialogue may fail to distinguish between experts and the public. Firstly, however, this addresses the imbalance of current face-to-face dialogue practices supported by Sciencewise, where only experts are allowed to set the context for the debate. But, secondly, as I have mentioned, social media dialogue could be structured to avoid entirely ignoring expert views.

Broader participation in dialogue through social media will not yield more informed policy conclusions.

Social media-based public debate, if effectively facilitated, could lead to decisions that are better informed and more robust than current Sciencewise-supported practices are producing. It could mitigate pitfalls, such as constraints on the range of factors being considered, which can affect formal, face-to-face public dialogue exercises, but are less likely to apply to social media-based debate.

Concluding recommendations A recent blog post by a former public participant in a Sciencewisesupported face-to-face public dialogue (now a citizen representative for Sciencewise) wrote, ‘Going forward […] I would like to see […] more cost effective means to engage with the public and make the process more open, involve a wider population and create the environment where more citizens’ are motivated to contribute in a new informal, fresh and

27

supported way’13. My two primary inter-linked recommendations align with this vision: Recommendation 1: Move the current broad-based public dialogue initiatives for gathering feedback on public policy onto social media and other online mechanisms. This offers a larger number of individuals the opportunity to participate, as well as greater efficiency and faster answers for policymakers. It also lowers the barriers to participation, thereby allowing broader (if thinner) engagement on an issue. Finally, it rebalances the power dynamic in public dialogue exercises away from institutional expertise and towards lay publics. Clearly, pilot dialogue projects through social media paired with good research on their strengths and weaknesses would be needed to develop best practice in this new domain. Recommendation 2: Target the use of face-to-face public dialogue activities to reach poor, disenfranchised and other underrepresented groups in public dialogue. These groups are not currently being effectively involved in the dialogue projects reviewed for this report (and presumably many others), but face-to-face approaches should, in principle, be the best way to reach those currently disconnected from public engagement with scientific issues. The methods and principles of formalised public dialogue exercises supported by Sciencewise should be adapted to this new purpose, including rebalancing the power dynamics that currently favour institutional experts. Effective, inclusive recruitment strategies and creative design of the dialogue exercises themselves will be crucial factors in extending the reach of public dialogue beyond the limited range of publics that is typically engaged by formal (and informal) dialogue exercises. Finally, I would recommend that Sciencewise develop ways of ensuring greater academic research involvement in public dialogue projects (particularly from social scientists interested in science and technology studies) in order to enhance the quality of delivery, analysis and evaluation. Regardless of whether social media-based approaches are adopted going forward, the significant investment of UK government funding in public dialogues (£334,000 in the case of the Synthetic Biology dialogue14) is yielding relatively limited depth of analysis and insight, when compared to the most obvious alternatives, such as social scientific focus groups and analysis of social media. With greater academic involvement, the quality of the research aspects of public dialogue, such as qualitative analysis of dialogue discussions, could be expected to improve. Greater general contributions to knowledge through further publications would also be more likely. In sum, in order to make facilitated social media-based public dialogue work effectively, it will require even greater finesse and expertise than face-to-face approaches.

13 14

http://www.sciencewise-erc.org.uk/blog/?p=2236 http://www.bbsrc.ac.uk/web/FILES/Reviews/1006-synthetic-biology-dialogue.pdf

28