BALTIMORE, MD. – As thousands were meeting at the American Geophysical Union fall meeting in San Francisco (see complete list of Forum coverage here), some 800 researchers, practitioners, and policymakers from more than 30 countries met in Baltimore to discuss “Risk Analysis for Better Policies.”

Risk communication and climate change were among the main subthemes at the 33rd annual meeting of the Society for Risk Analysis (SRA). The challenge of communicating climate change, in particular, was considered from different perspectives, and levels of analysis ranging from specific case studies to big-picture overviews. That approach echoes SRA’s objective of seeking to work across disciplines and at multiple levels.

Thirty-Three Years of Risk Analysis

SRA imageThe Society for Risk Analysis (SRA) was formed in 1980 by an interdisciplinary group of researchers and practitioners who wanted to better understand how humans assess and manage risk, and who wanted a place to publish work that did not fit neatly into established science and social science journals.

A year later the society began publishing Risk Analysis, first as a quarterly, then as a bi-monthly, and now as a ~200-page monthly. Risk communication quickly became a subspecialty of the journal.

The first articles on climate change — studies exploring the public’s mental models of global climate change through interviews and surveys — were published in 1994. In December 2005, Risk Analysis published a special issue on “Defining Dangerous Climate Change.” Another special series of articles, on “Climate Change Risk Perception and Communication,” was published in June 2012. (British social psychologist Nick Pidgeon edited or co-edited both of these special collections.) Recent issues have included articles on projected sea-level rise in Washington, DC (November 2012), on flood risk modeling for New York City (May 2013), and on the role of values in the acceptance of anthropogenic climate change (early view).

Communicating Specific Risks

It’s clear from the Risk Analysis series and recent articles just highlighted that the risk posed by climate change manifests itself in and through other risks: flooding (which can be the result of extreme precipitation events, storm surges, and/or sea-level-rise), droughts, heat waves, invasive species (which may include new bacteria or viruses), etc.

As a result, communicating climate change can involve communicating risk under and about very specific conditions. In several sessions of this year’s SRA meeting, presenters reported on research that tested messages (or message elements) crafted for very specific risks: hurricanes, flooding, nuclear energy, disease vectors like mosquitos and ticks, etc. How did wording, these researchers asked, affect the public’s response? What was the best way to express the probability, or the uncertainty, of something happening?

In the case of hurricanes, for example, researchers working for or in conjunction with the National Center for Atmospheric Research found the following:

  • Dramatically-worded warnings got the most attention, but they also aroused the most suspicion.
  • Individualists were less likely to evacuate in response to warnings than were those with hierarchical, egalitarian, or communitarian leanings.
  • The National Weather Service’s “cone of uncertainty” caught the public’s attention but delivered decidedly mixed messages; some respondents referred to it as “the cone of death” or “the cone of confusion.”

When Risks Shift or Intensify

Prior experience with a risk affects how people respond to messages about that risk. But under climate change, risks are shifting and intensifying. Some individuals may not know how to interpret messages about unfamiliar risks; others may not heed messages about familiar risks they think they understand. (The effort to understand the social factors that increase or decrease perceptions of risk is referred as “the social amplification of risk framework.”)

Reporting on hurricane warning messages, Evans School of Public Affairs professor Ann Bostrom noted that Floridians were surprised that New Jersey gas stations were not equipped with generators; that was common practice in Florida. In the audience was Rutgers behavioral ecologist Joanna Burger, who reported on interviews she had conducted after Hurricane Sandy. Most New Jerseyans, she explained, had had no previous experience living without power for that long.

A dramatic example of the risk posed by inadequate responses to risk was provided by doctor and journalist Sheri Fink, who briefly recounted the tragic story she chronicled in Five Days at Memorial: Life and Death in a Storm-Ravaged Hospital, a greatly expanded version of her feature story in the August 30, 2009, issue of New York Times Sunday Magazine. The hospital had served as a secure shelter when Hurricane Katrina first struck New Orleans, but in the aftermath, surrounded by flood waters and without power, it became a trap. According to Fink, “it appears that at least 17 patients were injected with morphine or the sedative midazolam, or both,” when hospital staff judged that these patients, weakened by dehydration after days in the 100-degree heat, would not survive long-delayed evacuations by helicopter.

A featured speaker in one of the conference’s plenary sessions, Fink then pointed out the parallels between the system failures at Memorial and the many flaws Hurricane Sandy revealed in New York City hospitals’ emergency plans and preparations.

General Rules of Risk Communication?

The conflicted outcomes of such examples, and the different results of so many specific studies, can paralyze practitioners. Aware of this, other SRA panels synthesized current understanding of risk analysis and communication through systematic literature reviews or by revisiting and updating key articles and studies.

SRA imageAs a practical summary of their extensive, multidisciplinary review of literature on communicating uncertainty, public health professors Cindy Jardine and S. Michelle Driedger produced a pamphlet for health communicators that “busted” five myths:

  • Myth 1: Uncertainty is all the same — Differences matter!
  • Myth 2: People will not accept uncertain information — People can handle uncertain information . . .
  • Myth 3: Admitting uncertainty undermines credibility — Honesty Matters!
  • Myth 4: Information should be withheld until there is reasonable certainty — Timing matters!
  • Myth 5: The media never get it right — Work with the media!

On the final day of the conference, 10 of the 36 contributors to Effective Risk Communication — a just-published book that revisited key concepts, models, and theories — participated in a roundtable discussion with the goal of highlighting lessons learned.

Thomas Webler said that the traditional model of the relationship between experts and citizens errs on two counts: “Analysis isn’t just what scientists do, and deliberation isn’t just what citizens do.”

Jardine, one of the contributors to this volume, warned against self-fulfilling prophecies: “Preparing for adversarial situations can create adversarial situations.” Joe Arvai, a co-editor of the volume, concurred: “If the intent becomes ‘believe my message,’ then communication quickly goes off the rails.”

And Lauren Fleishman-Mayer noted that scientists are often out-communicated because “their opponents use the lay wording of the public [while] scientists and activists produce messages that are not salient for the public.”

Rising Odds Against Risk Communication?

The challenge posed by these “opponents” was addressed in two sessions that provided historical context for risk analysis and communication, especially regarding climate change.

In an “Authors Meet Critics” session, sociologists Aaron M. McCright and Ortwin Renn, the incoming and first non-US president of SRA, summarized a book they co-authored with the late Eugene A. Rosa. In The Risk Society Revisited — the title alludes to The Risk Society published in 1986 by German sociologist Ulrich Beck — Rosa, Renn, and McCright reviewed the work of four social theorists who had each concluded that the environmental risks humans face are now largely the consequence of human actions. “Climate change” does not appear in Beck’s book, but it provides a further example of human actions reaching such a scale that they alter the environment in which humans live. Such human-made risks, Beck and the other theorists argued, pose special challenges for democratic societies.

The new challenge faced by risk analysts and communicators is that in one democratic society in particular this understanding of risk is now hotly contested. This was made clear in a second presentation by Aaron McCright, in a separate session on “Social Aspects of Climate Change Governance.”

There McCright previewed a paper he co-authored with Riley E. Dunlap and Chenyang Xiao, a paper just accepted by Weather, Climate and Society for publication sometime in 2014. Previous research, McCright noted, had found correlations between political ideology, trust in science (or perception of scientific consensus), and acceptance of anthropogenic climate change. Surveys and polls also documented an increasing gap between Republican and Democratic views on climate change since 2008 — when, in McCright’s words, “climate change denial went on steroids.” He and his colleagues wondered whether party identification was now playing a bigger role in public opinion on climate change.

After systematically analyzing Gallup survey data from 2006 and 2012, the researchers found that the effect of party identification on the perception of scientific consensus and acceptance of climate change had doubled. “The denial of anthropogenic climate change,” McCright concluded, “is becoming integrated into Republican identity.”

McCright offered no reasons to hope that renewed communication efforts by scientists and/or activists — or even Republicans’ own personal experiences with extreme weather — could change minds. Some Republican leaders had publicly accepted climate change, he acknowledged, but these were not current opinion leaders in the party.

Working from a different starting point, Roger Kasperson, professor of geography at Clark University and a founding member of SRA, came to a similar conclusion. In his talk on “Social Trust and Fracking,” Kasperson included a graph that tracked “Public Trust in Government” from 1958 to 2013. Pointing to the jagged but unquestionably falling line, Kasperson highlighted the gulf between the highest and lowest points. From more than 75 percent in 1964, public trust in government had fallen to 19 percent by the fall of 2013.

Public Trust in Government graphic
Public Trust in Government, 1958–2013 by Pew Research, Oct. 2013.

It is very hard to restore trust, Kasperson said, and without trust, top-down or command-and-control risk communication cannot work. One must begin with bottom-up deliberations — when and if one can actually get the parties talking. In other words, theories of risk communication formulated during periods of higher trust, as many were, might not now be applicable.

‘Do Not Generate Unnecessary Risk’

SRA image
The keynote speaker on the final day of the SRA meeting was Norman Loayza, director of the World Bank group that drafted Risk and Opportunity, the World Development Report for 2014, which framed development as a problem of risk management. After recounting positive and negative examples of risk management by developing countries, Loayza distilled the report’s recommendations into five policy principles:

  • Do not generate uncertainty or unnecessary risks.
  • Provide the right incentives for people and institutions to do their own planning and preparations, while taking care not to impose risk or losses on others.
  • Keep a long-run perspective for risk management by building institutional mechanisms that transcend political cycles.
  • Promote flexibility within a clear and predictable institutional framework.
  • Protect the vulnerable, while encouraging self-reliance and preserving fiscal sustainability.

As it happened, the top item in the news on that particular December 11th morning was the deal announced the evening before by Democratic Senator Patty Murray of Washington and Republican Congressman Paul Ryan of Wisconsin. The House approved the budget agreement the following day, the Senate a few days later. Washington, it seemed, had chosen to follow the World Bank’s first guideline; it would not generate unnecessary risk by shutting down the federal government — as it had in October.

The American political debate about the nature of risk — one side believes the federal government is a necessary actor in efforts to reduce risks, including climate change; the other side believes a too-powerful federal government is itself the risk — has unquestionably affected the global economy and international efforts to address climate change.

The current highly polarized climate of American politics has, in effect, become a risk factor. How best to analyze and manage that particular risk is surely a worthwhile subject for additional future research. But then a new question might arise: To whom should the results be communicated?

Topics: Communicating Climate