Department of Engineering and Public Policy Carnegie Mellon University 1 Risk Analysis and...

1Department of Engineering and Public Policy

Carnegie Mellon University

Risk Analysis and Communication: What can we learn from research?

M. Granger MorganHead, Department of Engineering and Public PolicyCarnegie Mellon UniversityPittsburgh PA 15213 USAtel: 412-268-2672e-mail: [email protected]



First, Two Slides on Carnegie Mellon's Department of

Engineering and Public Policyhttp://www.epp.cmu.edu

A department in the Engineering college at Carnegie Mellon University.

Faculty: Total of 41. Include true 50:50 joint appointments with all five engineering departments as well as joint appointments with four different social science units in three other colleges.

Undergraduate double major degrees with traditional departments (574 BS graduates to date).

Graduate program is a research-oriented Ph.D. focused on problem in which the technical details really matter (current enrollment 45, 119 Ph.D. graduates to date).



Research Four major areas:

1. energy and environment.2. risk analysis and communication.3. telecommunication and information policy.4. technology policy.

In the context of these four areas, we also work on issues in technology and development (China and India) and on issues in dual-use technology, arms control, and defense policy.

EPP currently has several large collaborative group efforts:• Center for Integrated Study of the Human Dimensions of Global Change.• The Electricity Industry Center.• Green Design Initiative.• Center for Energy and Environmental Systems.• Center for the Study and Improvement of Regulation.• Brownfields Center.• IT and telecommunications policy.• Risk analysis, ranking, communication.



Today I will talk about:• What is risk?

• Basic ideas in risk analysis.

• A few details on the characterization and treatment of uncertainty.

• Basic ideas in risk management.

• Basic ideas in risk communication.

I've also included some slides which I will not show which summarize a few results from four different recent research projects which I'd be happy to discuss individually.

• Some recent topics of research.



Risk

risk - n. 1. The exposure to the chance of injury or loss; a hazard or dangerous chance: he decided to take the risk.

Note:• negative outcome;• uncertain;• not just the probability of loss.

My dictionary defines risk as follows:

While this looks relatively simple, in the real world, things get more complicated…



Let's perform a little thought experiment

Suppose I have a new product.

I've done careful market research and know:

I could sell Q devices at a price P, for total revenues QP.

I'd make a profit R.

BUT, the product will have a net impact on national mortality of D excess deaths/year.

When would I be justified in introducing this product?



Some of the things that may matter...- A few people bear the risks and many get the benefits.

- The product is frivolous.

- The deaths do not occur immediately.

- I can identify the individuals before/after the deaths.

- The deaths all occur at once/are spread out.

- The people are socially related.

- D = N-M where N is deaths caused and M is deaths prevented.

- The effects are uncertain, <P(n)> = N.

- In addition to mortality, there is morbidity; environmental impact; etc.

In short, risk is a "multi-attribute concept…



Risks can be reliably sortedin terms of such factors...

Source: Slovic et al., in Readings in Risk, 1980.

QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.




A useful framework forthinking about risk

Exposure processes

Effects processes

Perception processes

Evaluation processes

benefits and costs

human activity

nature

Objects and systems exposed to possibility of change

Changes that occur

Perceived changes



A specific example

Transport, diffusion, oxidation

and deposition

Effects processes



benefits and costs

coal-fired power plant

natural air pollutants

Material in atmos. Plants and people exposed

Sunsets get redder; crops grown in sulfate poor soil increase yield; people develop respiratory difficulties.

Perceived changes

weather culture, political climate



The role of values

Exposure processes

Effects processes



benefits and costs

human activity

nature


Changes that occur

Perceived changes

Value-free analysis possible in principle.

Value-free analysis inherently not possible.



Today I will talk about:

• What is risk?

• Basic ideas in risk analysis







A highly simplified taxonomy of tools for risk assessment

risks from continuousexposure

risks from discreteevents failure mode

and effects analysis

fault treeanalysis

transport and dispersion models

dose-response models

SO2 SO4

wet and dry deposition



Transport and Dispersion Models


Source: Morgan and McMichael, Policy Sciences, 1981.



Dose Response Functions

Resp

onse

(e.g

. ca

nce

r in

cidence

)

Dose (e.g. concentration times time)

A B

threshold

Resp

onse

(e.g

. ca

nce

r in

cidence

)


A B

Linear, without (A) and with (B) threshold.

Non-linear, without (A) and with (B) threshold.



Complications

Re

spo

nse

(e

.g. c

an

cer

inci

de

nce

)


timeco

nce

ntr

ati

on

time

conce

ntr

ati

on

Healthy

Dead

C x t = D

C x t = D

Variable sensitivity amongindividuals.

Response depends in a time-dynamic way on exposure (i.e., no one-to-one mapping between cumulative exposure and response).



Examples of real dose-response functions



Source: Morgan, IEEE Spectrum, 1981.



We could go through a similardiscussion for discrete events

Event trees:

Fault trees:

Source: Rasmussen, Ann. Rev. of Energy, 1981.






• What is risk?



M. Granger Morgan, Max Henrion, with a chapter by Mitchell Small, Uncertainty: A guide to dealing with uncertainty in quantitative risk and policy analysis, 332pp., Cambridge University Press, New York, 1990. (Paperback edition 1992. Latest printing (with revised Chapter 10) 1998.)






Probability

Probability is the basic language of uncertainty.

I will adopt a personalistic view of probability (sometimes also called a subjectivist or Bayesian view).

In this view, probability is a statement of the degree of belief that a person has that a specified event will occur given all the relevant information currently known by that person.

P(X|i) where:

X is the uncertain event

i is the person's state of information.



The clairvoyant test

Even if we take a personalist view of probability, the event or quantity of interest must be well specified for a probability, or a probability distribution, to be meaningful.

"The retail price of gasoline in 2008" does not pass this test. A clairvoyant would need to know things such as:

• Where will the gas be purchased?

• At what time of year?

• What octane?



Does a subjectivist view mean your probability can be arbitrary?

NO, because if they are legitimate probabilities, they must

• conform with the axioms of probability

• be consistent with available empirical data.

Lots of people ask, why deal with probability? Why not just use subjective words such as "likely" and "unlikely" to describe uncertainties? There are very good reasons not to do this.



The risks of using qualitative uncertainty language

Qualitative uncertainty language is inadequate because:

- the same words can mean very different things to different people.

- the same words can mean very different things to the same person in different contexts.

- important differences in experts' judgments about mechanisms (functional relationships), and about how well key coefficients are known, can be easily masked in qualitative discussions.



Mapping words to

probabilities

Probability that subjects associated with the qualitative description

0.00.20.40.60.81.0

Almost certain

Probable

Likely

Good chance

Possible

Tossup

Unlikely

Improbable

Doubtful

Almost impossible

range of individual upper bound estimates

range of individual lower bound estimates

range from upper to lower median estimate

Qua

lita

tive

des

crip

tion

of

unce

rtai

nty

used

Figure adapted from Wallsten et al., 1986.

This figure shows the range of probabilities that people are asked to assign probabilities to words, absent any specific context.



Ex Com of EPA SAB

SAB members:

Key:

likely

someth

ing betw

een

likely

and not l

ikely

not lik

ely

Other meeting participants:

1.0

0.9

0.7

0.5

0.3

0.1

0.0

5

0.0

1

0.0

01

0.0

00

1

0.0

00

01

0.0

00

00

1

Probability that the material is a human carcinogen

The minimum probability associated with the word "likely" spanned four orders of magnitude.The maximum probability associated with the word "not likely" spanned more than five orders of magnitude. There was an overlap of the probability associated with the word "likely" and that associated with the word "unlikely"!

Figure from Morgan, HERA, 1998.



The bottom lineWithout at least some quantification, qualitative descriptions of uncertainty convey little, if any, useful information.

Here are two examples from the climate assessment community:

Schneider and Moss have worked to get a better treatment of uncertainty incorporated into the past and current round of IPCC. Progress is uneven, but awareness is growing. Individual investigators are pushing the process along.

The U.S. National Assessment Synthesis Team gave quantitative definitions to five probability words and tried to use them consistently throughout their overview report.



In doing risk analysis, we must consider two quite different kinds of uncertainty

1. Situations in which we know the relevant variables and the functional relationships among them, but we no not know the values of key coefficients (e.g., the "climate sensitivity").

2. Situations in which we are not sure what all the relevant variables are, or the functional relationships among them (e.g., will rising energy prices induce more technical innovation?).

Both are challenging, but the first is much more easily addressed than the second.



Uncertainty about quantities

From Morgan and Henrion, Uncertainty, Cambridge, 1990/99.



One should…

…use available data, and well-established physical and statistical theory, to describe uncertainty whenever either or both are available.

However, often the available data and theory are not exactly relevant to the problem at hand, or they are not sufficiently complete to support the full objective construction of a probability distribution.

In such cases, one may have to rely on expert judgment. This brings us to the problem of how to "elicit" expert judgment.



Expert elicitation takes time and care

Eliciting subjective probabilistic judgments requires careful preparation and execution.

Developing and testing an appropriate interview protocol typically takes several months. Each interview is likely to require several hours.

When addressing complex, scientifically subtle questions of the sorts involved with most problems in climate change, there are no satisfactory short cuts. Attempts to simplify and speed up the process almost always lead to shoddy results.

The next eight slides (which I will skip because time is short) talk about the important issues of overconfidence and the impacts of "cognitive heuristics." These are critically important topics for anyone who actually plans to do expert elicitation. <Skip to 37.>



Over Confidence

Source: Morgan and Henrion, 1990/99.

Extra slide - will not show



Over Confidence

Source: Morgan and Henrion, 1990/99.




Cognitive heuristics

When ordinary people or experts make judgments about uncertain events, such as numbers of deaths from chance events, they use simple mental rules of thumb called "cognitive heuristics."

In many day-to-day circumstances, these serve us very well, but in some instances they can lead to bias - such as over confidence - in the judgments we make.

This can be a problem for experts too.

The three slides that follow illustrate three key heuristics: "availability," "anchoring and adjustment," and "representativeness."




Cognitive bias

from Lichtenstein et al., 1978.

Availability: probability judgment is driven by ease with which people can think of previous occurrences of the event or can imagine such occurrences.




Cognitive bias…(Cont.)

Anchoring and adjustment: probability judgment is frequently driven by the starting point which becomes an "anchor."

from Lichtenstein et al., 1978.




Cognitive bias…(Cont.)

I flip a fair coin 8 times. Which of the following two outcomes is more likely?

Outcome 1: T, T, T, T, H, H, H, H

Outcome 2: T, H, T, H, H, T, H, T

Of course, the two specific sequences are equally likely...but the second seems more likely because it looks more representative of the underlying random process.

Representativeness: people judge the likelihood that an object belongs to a particular class in terms of how much it resembles that class.




Expert elicitation…(Cont.)In all our elicitation studies, we've focused on creating a process that allows the experts to provide their carefully considered judgment, supported by all the resources they may care to use. Thus, we have:

- Prepared a background review of the relevant literatures.

- Carefully iterated the questions with selected experts and run pilot studies with younger (Post-doc) experts to distil and refine the questions.

- Conducted interviews in experts' offices with full resources at hand.

- Provide ample opportunity for subsequent review and revision of the judgments provided.

All of these efforts have involved the development of new question formats that fit the issues at hand.




Expert elicitation …(Cont.)Over the past two decades, my colleagues and I have developed and performed a number of substantively detailed expert elicitations. These have been designed to obtain experts’ considered judgments. Examples include work on:

Health effects of air pollution from coal-fired power plants.

• M. Granger Morgan, Samuel C. Morris, Alan K. Meier and Debra L. Shenk, "A Probabilistic Methodology for Estimating Air Pollution Health Effects from Coal-Fired Power Plants," Energy Systems and Policy, 2, 287-310, 1978.

• M. Granger Morgan, Samuel C. Morris, William R. Rish and Alan K. Meier, "Sulfur Control in Coal-Fired Power Plants: A Probabilistic Approach to Policy Analysis," Journal of the Air Pollution Control Association, 28, 993-997, 1978.

• M. Granger Morgan, Samuel C. Morris, Max Henrion, Deborah A.L. Amaral and William R. Rish, "Technical Uncertainty in Quantitative Policy Analysis: A Sulfur Air Pollution Example," Risk Analysis, 4, 201-216, 1984 September.

• M. Granger Morgan, Samuel C. Morris, Max Henrion and Deborah A. L. Amaral, "Uncertainty in Environmental Risk Assessment: A case study involving sulfur transport and health effects," Environmental Science & Technology, 19, 662-667, 1985 August.



Expert elicitation…(Cont.)

• M. Granger Morgan and David Keith, "Subjective Judgments by Climate Experts," Environmental Science & Technology, 29(10), 468-476, October 1995.

• Elizabeth A. Casman, M. Granger Morgan and Hadi Dowlatabadi, “Mixed Levels of Uncertainty in Complex Policy Models,” Risk Analysis, 19(1), 33-42, 1999.

• M. Granger Morgan, Louis F. Pitelka and Elena Shevliakova, "Elicitation of Expert Judgments of Climate Change Impacts on Forest Ecosystems," Climatic Change, 49, 279-307, 2001.

• Anand B. Rao, Edward S. Rubin and M. Granger Morgan, "Evaluation of Potential Cost Reductions from Improved CO2 Capture Systems,"Proceedings of the 2nd National Conference on Carbon Sequestration, Alexandria, VA, May 5-8, 2003.

Climate science, climate impacts and mitigation technology:

• M. Granger Morgan, "The Neglected Art of Bounding Analysis," Environmental Science & Technology, 35, 162A-164A, April 1, 2001.

• Minh Ha-Duong, Elizabeth A. Casman, and M. Granger Morgan, "Bounding Poorly Characterized Risks: A lung cancer example," Risk Analysis, in press.

Bounding uncertain health risks:



Warming for 2x[CO2]

1

234

56789

10111213141516

Temperature response given 2x[CO ] (K)

0-5-10 5 10 15 20

4 with "surprise"

exp

ert

2

with state change

0-5-10 5 10 15 20

2Source: Morgan and Keith, ES&T, 1995. …and, lest you conclude that most of these

experts are in basic agreement…



Pole to equator temperature gradient for

2x[CO2]122

56789

10111213141516

Meridinal temperature gradient given 2x[CO ] (K)

3530 40

30 35 40

4

ex

pe

rt

na

na

na

3

w/climate state change

2

Source: Morgan and Keith, ES&T, 1995.



Biomass in Northern Forestsw/ 2xCO2 climate change

1

2

3

4

5

6

7

8

Expe

rt

9

10

11

A. B.

Change in soil carbon in minimally disturbed Northern Forests between 45°N and 65°N under specified 2x[CO 2] climate change.

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.40.4

Change in standing biomass in minimally disturbed Northern Forests between 45°N and 65°N under specified 2x[CO2] climate change.

North America

Eurasia

North America

Eurasia

Eurasia

North America

"trivial"

w/permafrostw/o permafrost

North America

Eurasia

North America and Eurasia E of the Urals

Europe west of the Urals

0.6 0.8 1.0 1.2 1.40.40.2

North America

Eurasia

North America

Eurasia

w/permafrostw/o permafrost

North America and Eurasia E of the Urals

Europe west of the Urals

North America

Eurasia

Change in standing biomass Change in soil carbon

Source: Morgan et al., Climatic Change, 2001.



Biomass in Tropical Forests w/ 2xCO2 climate change

A. B.

1

2

3

4

5

6

7

8

Expert

9

10

11

0.6 0.8 1.0 1.2 1.40.4

Change in standing biomass in minimally disturbed Tropical Forests between 20°N and 20°S under specified 2x[CO2] climate change.

0.6 0.8 1.0 1.2 1.40.4

Change in soil carbon in minimally disturbed Tropical Forests between 20°N and 20°S under specified 2x[CO2] climate change.

Change in standing biomass Change in soil carbon

Source: Morgan et al., Climatic Change, 2001.



Uncertainty about model formOften uncertainty about model form is as or more important than uncertainty about values of coefficients. Until recently there had been little practical progress in dealing with such uncertainty, but now there are several good examples:

• John Evans and his colleagues at the Harvard School of Public Health (Evans et al., 1994).

• Alan Cornell and others in the seismic risk (Budnitz et al., 1995).

• Hadi Dowlatabadi and colleagues at Carnegie Mellon in Integrated Assessment of Climate Change (ICAM) (Morgan and Dowlatabadi, 1996).



Energy & Emissions

Atmospheric Composition &

Climate

Demographics & Economics

INTERVENTION

Impacts of Climate Change

StructureInputs Outputs

To run the model: 1 - Double click on INPUTS to set up the scenario inputs; 2 - Double click on STRUCTURE to set up the model; 3 - Double click on OUTPUTS and evaluate the indicators.

Regional ²T

GHG Models

Aerosol Model

ForcingChange in Long Wave

Forcing

Change in Short Wave

Forcing

Elicited Climate Model

RITS Conc

ICAMIntegrated Climate Assessment Model

See for example:Hadi Dowlatabadi and M. Granger Morgan, "A Model Framework for Integrated Studies of the Climate Problem," Energy Policy, 21(3), 209-221, March 1993.andM. Granger Morgan and Hadi Dowlatabadi, "Learning from Integrated Assessment of Climate Change," Climatic Change, 34, 337-368, 1996.

A very large hierarchicallyorganized stochasticsimulation model builtin Analytica®.



ICAM deals with...

…both of the types of uncertainty I've talked about:

1. It deals with uncertain coefficients by assigning PDFs

to them and then performing stochastic simulation to

propagate the uncertainty through the model.

2. It deals with uncertainty about model functional form

(e.g., will rising energy prices induce more technical

innovation?) by introducing multiple alternative

models which can be chosen by throwing "switches."



ICAMThere is not enough time to present any details from our work with the ICAM integrated assessment model. Here are a few conclusions from that work:

• Different sets of plausible model assumptions give dramatically different results.

• No policy we have looked at is dominant over the wide range of plausible futures we’ve examined.

• The regional differences in outcomes are so vast that few if any policies would pass muster globally for similar decision rules.

• Different metrics of aggregate outcomes (e.g., $s versus hours of labor) skew the results to reflect the OECD or developing regional issues respectively.



These findings lead us......to switch from trying to project and examine the future, to using the modeling framework as a test-bed to evaluate the relative robustness, across a wide range of plausible model futures, of alternative strategies that regional actors in the model might adopt.

We populated the model's regions with simple decision agents and asked, which behavioral strategies are robust in the face of uncertain futures, which get us in trouble.

Thus, for example, it turns out that tracking and responding to atmospheric concentration is more likely to lead regional policy makers in the model to stable strategies than tracking and responding to emissions.



Our conclusionPrediction and policy optimization are pretty silly analytical objectives for much assessment and analysis related to the climate problem.

It makes much more sense to:

• Acknowledge that describing and bounding a range of futures may often be the best we can do.

• Recognize that climate is not the only thing that is changing, and address the problem in that context.

• Focus on developing adaptive strategies and evaluating their likely robustness in the face of a range of possible climate, social, economic and ecological futures.




• What is risk?








Strategies for managing risk

Exposure processes

Effects processes



benefits and costs

human activity

nature


Changes that occur

Perceived changes

Modify the human activity

Modify the effects process

Modify the exposure process

Modify human perceptions

Modify human valuations

Mitigate or compensate



Examples

Modify naturalor humanenvironment

Live close towork and walk.Build rapidtransit systems.

Eliminatepoverty,inequality,anger, mentalillness, and soon.

Risk

Risk of vehicleoccupantinjury in autoaccident.

Risk of gettingshot bysomeone withhandgun.

Avoid ormodifyexposureprocesses

Change speedlimits.Get tough withDUIs.Train people todrivedefensively.

Ban handguns.Impose harshpenalties for usein crimes.Stay out of highcrime areas.

Avoid ormodify effectsprocesses

Wear seat belts.Install pro-tective systems.

Wear protectivebulletproofclothing.Duck.

Mitigate orcompensatefor effects

Carry autoinsurance.Operate goodemergencymedical system.Sue otherdriver.

Carry healthinsurance.Operate goodemergencymedical system.Sue person whoshot you.

No strategy is best everywhere. Depending on context, different strategies do best in different settings.



Strategies for managing risk

• Tort and other common law (e.g., negligence, liability, nuisance, trespass).

• Insurance (private, public or hybrid).

• Voluntary standard setting organizations (UL, ASTM, etc.).

• Individual and collective corporate initiatives (CMA - Responsible Care).

• Information-based strategies (e.g, TRI, green labels).

• Mandatory government standards and other regulations (performance standards, design standards).

• Market-based solutions (emissions taxes, tradable permits).




• What is risk?




• Basic ideas in risk communication.M. Granger Morgan, Baruch Fischhoff, Ann Bostrom and Cynthia Atman, Risk Communication: A mental models approach, 351pp., Cambridge University Press, New York, 2002.




The traditional approach……to risk communication involves two steps:

1. Ask an expert what people should be told.

2. Get a "communications expert" to package it.

And, if you are being really fancy:

3. Run some tests to see how people like it.

However, if you think about it for a few minutes, this approach ignores two critical issues:

• What the people receiving the message already "know" about the topic.

• What they need to know to make the decisions they face.



Finding out what people know…

…is not simple! If I give people a questionnaire, I have to put information in my questions.

People are smart. They will start making inferences based on the information in my questions.

Pretty soon I won't know if their answers reflect what they already knew, or the new ideas they have come up with because of the information I have given them.

I need a better, less intrusive way to learn what they know. The method we have developed to do this is called the mental model interview.



Five step process:In work done over the past 15 years at Carnegie Mellon, we have developed a five-step approach to risk communication, based on people's mental models of risk processes:

1. Develop an "influence diagram" to structure expert knowledge.

Number of interviews

Cum

ula

tive

num

ber

of

conce

pts

0

40

80

120

160

1 10 20 30

A

B1

B2 B3S1

S2 S3

S4

2. Conduct an open-ended elicitation of people's beliefs about a hazard, allowing the expression of both accurate and inaccurate concepts. Map the results to the expert influence diagram.



Five steps…(Cont.)3. On the basis of results from the open-ended interviews,

develop and administer a closed-form questionnaire to a much larger group in order to determine the prevalence of these beliefs.

4. Develop a draft communication based on both a decision analytic assessment of what people need to know in order to make informed decisions and a psychological assessment of their current beliefs.

5. Iteratively test successive versions of those communications using open-ended, closed-form, and problem-solving instruments, administered before, during, and after the receipt of the message.

In 1994, we applied these methods to study public understanding of climate change. While the results are now getting old, I will show you a few since they are unlikely to have changed much.



Example of opening response in interview

Interviewer: "I'd like you to tell me all about the issue of climate change."

Subject: "Climate change. Do you mean global warming?"

Interviewer: "Climate change."

Subject: "OK. Let's see. What do I know. The earth is getting warmer because there are holes in the atmosphere and this is global warming and the greenhouse effect. Um... I really don't know very much about it, but it does seem to be true. The temperatures do seem to be kind of warm in the winters. They do seem to be warmer than in the past.. and.. hmm.. That's all I know about global warming.



Another example…

Interviewer: "Tell me all about the issue of climate change."

Subject: "I'm pretty interested in it... The ice caps are melting -- the hole in the ozone layer. They think pollution from cars and aerosol cans are the cause of all that. I think the space shuttle might have something to do with it too, because they always send that up through the earth, to get out in outer space. So I think that would have something to do with it, too."



Another example…

Interviewer: "Tell me all about the issue of climate change."

Subject: "Climate change? Like, what about it? Like, as far as the ozone layer and ice caps melting, water level raising, rainforest going down, oxygen going down because of that? All of that kind of stuff?"

Interviewer: "Anything else?"

Subject: "Well, erosion all over the place. Um, topsoils going down into everywhere. Fertilizer poisoning. "Interviewer: "Anything else that comes to mind related to climate change?Subject: "Climate change. Winter's ain't like they used to be. Nothing's as severe. Not as much snow. Nothing like that."



Schematic of the approachConduct open- ended interviews of representative lay people.

Develop/refine representations of lay mental models.

Validate representation(s) and prevalence of mental model(s) with closed-form questionnaires.

Structure model of lay values.

Structure decision problems lay people face.

Develop expert model of the risk problem.

Structure risk managment strategy.

Design risk communication (if required).

Evaluate risk communication with open-ended, closed-form and problem solving instruments.

Evaluate risk managment strategy (eg. changes in behavior, etc.).

Other social and economic considerations.

1

2

3

4

1

2

3

4

5



Based on what we have learned…

…in our studies of public understanding of various risks we have produced various public communication materials:

After the initial two-page overview, three two-page spreads concisely summarize the key ideas.

Light-weight "Details Books" in pockets at the back of each of the three sections provide a third level of details.

Part 1: What is climate change?

•Part 1•Part 1•Part 1•Part 1•Part 1•Part 1••Part 1•Part 1•Part 1•Part 1•Part 1•

An initial two-page spread hits the main points and lists common misconceptions

I have brought a few examples for people who are interested.



A few conclusions

There is no such thing as an "expert" in public communication who can simply tell you what to do.

An empirical approach is essential.

Before developing a communication one must learn what the public knows and thinks.

The mental models method offers a promising strategy for doing this.

Uncertainty must be quantified.




• What is risk?








In the Department of Engineering andPublic Policy at Carnegie Mellon…

…we've had an active program in risk-related research for 30 years. Since time is very short, I've included a few illustrative slides on four recent projects:

• Dealing with mixed and extreme uncertainty.Elizabeth A. Casman, M. Granger Morgan and Hadi Dowlatabadi, “Mixed Levels of Uncertainty in Complex Policy Models,” Risk Analysis, 19(1), 33-42, 1999.

• The use of bounding analysis.Minh Ha-Duong, Elizabeth A. Casman, and M. Granger Morgan, “Bounding Poorly Characterized Risks: A Lung Cancer Example,” Risk Analysis, in press.

• Risk ranking.H. Keith Florig et al. "A Deliberative Method for Ranking Risks (I): Overview and test bed development," Risk Analysis, 21(5), 913-921, 2001 and Kara M. Morgan et al. "A Deliberative Method for Ranking Risks (II): Evaluation of validity and agreement among risk managers," Risk Analysis, 21(5), 923-937, 2001.

• Public perceptions of deep geological disposal of CO2.Claire R. Palmgren, M. Granger Morgan, Wändi Bruine de Bruin, and David W. Keith "Initial Public Perceptions of Deep Geological and Oceanic Disposal of Carbon Dioxide," in review at ES&T.



AcknowledgmentsOver the years my work on risk has been greatly assisted by collaborations with many colleagues including:

Elizabeth CasmanMichael DeKayHadi DowlatabadiPaul FischbeckBaruch FischhoffKeith FlorigMax HenrionKaren JenniDavid KeithLester LaveKara MorganLouis PitelkaElena ShevliakovaPatti SteranchakHenry Willis



Limited domain of

model validity

Year2075 21002050202520001975

0

1

2

3

4

5

6

Me

an

ch

an

ge

in

g

lob

al t

em

pe

ratu

re,°

C

0

1

2

3

4

5

6

Me

an

ch

an

ge

in

g

lob

al t

em

pe

ratu

re,°

C

0

1

2

3

4

5

6

Me

an

ch

an

ge

in

g

lob

al t

em

pe

ratu

re,°

C

0.00

0.10

Pro

ba

bili

ty o

f a c

ha

ng

e

in

the

clim

ate

sta

te

0.20

0.30

0.00

0.10

Pro

ba

bili

ty o

f a c

ha

ng

e

in

the

clim

ate

sta

te

0.20

0.30

0.00

0.10

Pro

ba

bili

ty o

f a c

ha

ng

e

in

the

clim

ate

sta

te

0.20

0.30

Year2075 21002050202520001975

Year2075 21002050202520001975

Expert 1

Expert 8

Expert 15

Examples of warming estimated via the ICAM model (dark curves) and probability that the associated climate forcing will induce a state change in the climate system (light curves) using the probabilistic judgments of three different climate experts.




Modelswitching

Schematic illustration of the strategy of switching to progressively simpler models as one moves into less well understood regions of the problem phase space, in this case, over time.

Time

period during which one thinks output from detailed model is meaningful

period during which one thinks output from a simple order-of- magnitude model is meaningful

period during which only bounding analysis is meaningful

complete ignorance

0

1




Illustrationof modelswitching

Results of applying the model switch-over strategy to the ICAM demographic model (until about 2050) and an estimate of the upper-bound estimate of global population carrying capacity based on J. S. Cohen.

0.0

0.2

0.4

0.6

0.8

1.0Weight for ICAM model

Weight for Cohen model

2000 2050 2100 2150 2200 2250 2300

20

40

60

0

80




Bounding

When the science is poorly understood, probabilistic risk analysis is routinely used to obtain estimates of health impacts, with results typically reported in the form of a very broad subjective probability density function.

For impacts with multiple causes, such estimates are usually made separately, by different investigators, for each cause of interest.

However, if those separate probabilistic estimates were brought together and summed, the results could sometimes substantially exceed the numbers of cases actually observed.




Bounding analysisIn a commentary in ES&T in April of 2001, I argued that methods of bounding analysis could be used in environmental risk analysis to avoid such problems. For health endpoints with multiple external causes, the available knowledge can be used to constrain estimates of the magnitude of the poorly characterized risks.

If most risks were known with precision, this would be a simple subtraction problem. But health risks from environmental causes often involve high uncertainty. However, in many cases, there is agreement on the general magnitude of the impacts of the best-studied causes. The idea is to use this knowledge to bound the sum of the other less well known risks.

M. Granger Morgan, "The Neglected Art of Bounding Analysis," Viewpoint, Environmental Science & Technology, 35, 162A-164A, April 1, 2001.




To illustrate the method……we have developed an example that harmonizes informed beliefs about the relative contribution made by different causal factors using total current U.S. lung cancer mortality.

The goal of the analysis is to generate an upper bound on the mortality attributed to the group of poorly-characterized factors, derived from information on the group of well-characterized risk factors.

To perform this bounding analysis it is first necessary to apportion lung cancer mortalities among the various known causes, or groups of causes, according to current scientific knowledge and opinion.




Illustration

All lung cancers:

Lung cancers attributable to:

tobacco:

radon:

asbestos:

all others:




In the example…

…I am presenting today, we have developed the needed bounds by reviewing the relevant literatures and making our own judgments.

We are currently running an expert elicitation to seek the carefully informed judgment of a number of lung cancer experts.




We partition…

…the lung cancer deaths into non-intersecting categories of single causative factors and causal factor groupings and constrain the sum of all mortality across those partitions to equal the observed mortality.

In addition to causal categories, we define a "background" category for the lung cancer deaths that would have occurred in the absence of exposure to all carcinogens.

While there is no way to measure this quantity, clearly it exists in principle, and it can be bounded through consideration of lung cancer deaths in groups with low exposures to known carcinogens.




In our illustration…

… the set of possible causes of lung cancer () consists of active, former, and passive smoking (C); domestic exposure to radon (R); occupational exposure to inhaled asbestos (A); and the group of all other environmental risk factors (X).

Note that our analysis deals with annual lung cancer mortality in the U. S. population as a whole and not in subpopulations that experience different exposure histories, display different susceptibilities and have different access to health care.

Since genetic factors play a role in all cancers, we consider them to be a priori, non-manipulable attributes of the overall U.S. population.




Graphically

The basic statistic n, showing only risk factors C, R, and X, for simplicity. N is the totalnumber of lung cancer fatalities. n is the number of fatalities apportioned to each riskfactor or combination of factors. n() is the background number of lung cancer deathsthat would occur absent all the various risk factors. n() = n(CRX) and represents thenumber of cases for which no risk factor can be excluded.

N

n() n(C) n(R) n(CR) n(X) n(XC) n(XR) n()




Smoking

Smoking-Attributable Mortality (SAM), Cancers of the Trachea, Lung and Bronchus,1999 (Source: CDC)

Mortality fromNeoplasms of

Trachea, Lung,and Bronchus

Lung CancerDeaths fromSecondhandSmoking(28)

Active & PassiveSmoking

AttributableFraction

Total deaths 89,337SAM 78,459 1,110Males

percent 87.8% 89.1%Total deaths 62,613

SAM 44,727 1,890Femalespercent 71.4% 74.4%

Total deaths 151,950SAM 123,186 3,000 83.0%

MalesandFemales percent 81.1% 2%




In summary

Based on our review of the literature, we have constructed a set of judgments apportioning lung cancer deaths among these causes which imply the following constraints:

f u (C) = 0.70, f l(C) = 0.95 f u (R) = 0.21, f l(R) = 0.02 f u (A) = 0.05, f l (A) = 0.01 where f u denotes an upper bound on the fraction of lung cancer deaths due to a particular cause and f l the lower bound.




From the bounds……on the proportions of annual lung cancer mortality due to the major lung cancer causes (C, R, and A), information about the magnitude of risk from all other lung cancer causes (X) is

then inferred, using a consistency constraint on the total number of deaths.

Graphically we perform:

R = N - n()i.e. R = 100%

R = 0%

C = 0%

C = N - n()i.e. C = 100%

X = N - n()i.e. X = 100%

X = 0% Pn




Example results

Results of Optimization: Upper and Lower Bounds on Attributable Fractions andRelative Risks of Cancer Risk Factors

C R A XUpper bound on af 95% 21% 5.0% 3.2%Lower bound on af 70% 2% 1.0% 0.0%

Exposure probability 45%(a) 50%*(b) 5%(c) 5%**Upper bound on rr 43.2 1.53 2.05 1.66Lower bound on rr 6.19 1.04 1.20 1.00a. National Center for Health Statistics, 2001.b. NRC, 1999.c. Dept. of Labor, 2002.* The fraction of U. S. homes with radon concentrations at or above 25 Bqm-1.** Our estimate.




Example results…(Cont.)This example calculation assigns between 0% and 3.2% of lung cancer deaths to X, the group of unspecified occupational and environmental pollutants.

Thus, for the group of known and suspected lung carcinogens other than C, A, and R, if one is confident in the bounds assigned to the well understood risk factors, the sum of the effects of the poorly understood factors collectively should account for no more than 3.2% of total lung cancer mortality.

This provides a constraint on estimates of those risks produced by more conventional risk analysis.




Checking diesel estimates

Two national associations of air quality control offices published a projection of the number of cancers due to exposure to diesel exhaust.

They estimated that diesel would be responsible for 125,110 cancers for all metropolitan and non-metropolitan areas of the U.S. (over 70-year lifetimes) for an annual rate of 1,787 cancers.

This figure is below 3% (4,716 deaths) of the projected 2003 lung cancer mortality rate of 157,200, even without adjusting for the non-lung cancer mortality inherent in their risk estimate, and therefore, to first order, would pass our plausibility test.




Our objective......in the work we've done on risk ranking has been to develop and demonstrate a method which:

• Uses experts to analyze and characterize the risks (because they have the necessary knowledge).

• Uses modern risk communication methods to describe the risks in multi-attribute terms (so that they will be understandable to educated members of the general public).

• Uses representative groups of laypeople to perform the actual ranking (because ranking requires the application of social values).

• Produces a fairly "thick" description of the deliberations, including ranks which are robust, and which are useful as an input in risk management decision making.




Steps in the risk-ranking method


Step ADefine andcategorize therisks to beranked.

Step CDescribe therisks in terms ofthe attributes inrisk summarysheets.Step B

Identify the riskattributes thatshould beconsidered.

Step DSelectparticipantsand performthe riskrankings.

Step EDescribe theissuesidentified andthe resultingrankings.



First experimental test-bed:risks in schools

We chose risks in a hypothetical middle school as our test-bed because:

• most people know and care about the topic -- major efforts are not required to get lay subjects briefed before participating in studies;

• risks in schools are not the responsibility of any single existing U.S. Government risk management agency;

• the topic offers opportunities to address a wide range of physical and chemical risks as well as important social issues;

• there are a number of recent studies on which we can build.




Example of a risk summary

sheet

Risk name

Brief summary description of the risk.

Tabular summary, evaluating the risk in terms of the standard attributes and description of the level of uncertainty.

General description of the risk, discussion of the risk in thespecific context of the Centerville Middle School, and discussion of what the school has done to deal with the risk.




Town of Centerville

Centervil le Middle School

Alvarez ExpresswayC&LL Rail Road

230kV transmiss ion l ine

C&LL Rail Road

water tower

Schuman Hil l regional park

fire s tation

police

c ity hall

air field

c liff

cliff

to Centerv il le Heights

Centervil le High School

Wau

sau

Riv

er

Centervil le Elementary School

City of Centerville

0 0.5 1.0

Scale in milesSewage treatment plant

Centervil le Landing

Water treatment plant

Wa

usau R

iver




CentervilleMiddle School

Room 107 girl's

washroom

Room 108 classroom

Room 109 classroom

Room 110 classroom

Room 104 classroom

Room 131 classroom

Room 119 boy's

washroom

Room 130 auditorium

stage Room 140 utility

Room 135 library

Room 120 cafeteria

Room 121 kitchen

Centerville Middle School First Floor

0 10 20 30 40 50

Scale in feet

Rooms 100 to 103 administrative offices

serving area

elevator

utility tunnel

Room 141 janitorial

retaining wall

loading dock

Room 105 classroom

Room 106 classroom

Room 132 classroom

Room 133 classroom

Room 134 classroom

Room 136 classroom

Room 137 classroom

Room 138 classroom

Room 139 classroom




Second test-bed

The middle school test-bed involved only health and safety risks.

After developing and refining the method in that context, we then developed a second test-bed (based on a hypothetical county) in order to extend the work to include ecological and environmental risks.

A major part of the work in creating this second test-bed has involved developing an appropriate set of relevant attributes. That's a fairly long story that I will not go into, but Ph.D. student Henry Willis has done a great job with this problem.




DePaul County

Centerville

I-82

C&LL RR

DePaul County

Au Claire River

Wausau River

Dubuque River

1

2

1. Fishkill Power Plant 2. Site of former Johnson's Wood Products Plant 3. DeKay County Sanatary Land Fill 4. Cebulka's Premium Chickens 5. Ganley Estates 6. Feldman's Feeds 7. Site of proposed Twin Pines Mall

3

State Route 246

4

2 40 6 8

C&LL RR

Scale in miles

5

6

Harris State Park

Crystal Lake

7

I-82

Forest

Forest

Forest




Risks used in the two test-bedsThe county-level test-bed• Agricultural runoff• Air pollution from electric power generation• Food poisoning• Genetically-modified corn• Invasive species• Land filling municipal solid waste• Motor vehicle accidents• Recreational motor boating• Road salt and road salt runoff• Transporting hazardous materials by truck

The middle school test-bed• Accidental injuries• Airplane Crash• Allergens• Asbestos• Bites and Stings• Building Collapse• Commuting to school on foot, bike or

by car• Common infectious diseases• Drowning• Electrical Power• Electromagnetic fields (EMF)• Fire and Explosion• Food Poisoning• Hazardous material transport• Intentional Injury• Lead poisoning• Less common infectious diseases• Lightning• Radon gas• School bus accidents• Self-Inflicted Injury• Team Sports




Typical ranking procedure

Individualholistic ranking

Start groupranking

IndividualMAranking

Revisegroup ranking

FinalindividualH and MArankings

Studyall materials




Summary of findingswith school test-bed:

Consistency between the rankings that have resulted from the holistic and multiattribute procedures has been good for both individuals and for groups, suggesting that these procedures capture an underlying construct of riskiness.

Rankings of risks were similar across individuals and groups, even though individuals and groups did not always agree on the relative importance of risk attributes.

Lower consistency between the risk rankings from the holistic and multiattribute procedures and lower agreement among individuals and groups regarding these rankings were observed for a set of high-variance risks.




The story in numbers:

Initial IndividualHolistic

Risk Rankings

Final IndividualHolistic

Risk Rankings

GroupHolistic

Risk Rankings

Initial IndividualMulti-AttributeRisk Rankings

Final IndividualMulti-AttributeRisk Rankings

GroupMulti-AttributeRisk Rankings

(Step D1)

0.595

(Step D3)

0.829

(Step D5)

0.860

(from Step D2)

0.759

(from Step D4)

0.913

(from Step D6)

0.857

0.863

0.770 0.915

0.595 0.686 0.658

0.858 0.915

0.920

218 individuals (in 43 groups) have performed risk rankings for health and safety risks.




Summary...(Cont.)

Participants reported high levels of satisfaction with their groups’ decision-making processes and the resulting rankings, and these reports were corroborated by regression analyses.

Because of the generally high levels of consistency, satisfaction, and agreement we have observed we conclude that this deliberative method is capable of producing risk rankings that can serve as informative inputs to public risk-management decision making.




Findings fromthe environmental/ecological studies

• Participants are able to complete expanded ranking task.

• Holistic and multiattribute rankings are consistent with each other.

• Agreement shown between individuals and groups.

• Though less satisfied with attribute ranking process, participants revealed satisfaction with final group risk rankings.

• Results parallel previous findings from health and safety test-bed.




In order to study public perceptions……of CCD, we had to modify the basic mental model interview approach since the typical member of the general public knows nothing about this technology.

Part 1: Using language that we made as neutral as we could, we explained the motivation of wishing to reduce the accumulation of atmospheric CO2 and briefly outlined the basic design options of CCD technologies. We then asked a series of questions to elicit reactions.

Part 2: Briefly discussed three questions: 1) "Can the technology to separate and dispose of carbon dioxide be made practical, and cheap enough"; 2) "Once the CO2 is put down deep in rock formations or deep in the oceans, will it stay there?" and 3) "If the technology can be made cheap and reliable, will the energy industry adopt it and use it widely?" Again, we asked a series of questions to elicit reactions.

Part 3: Discussed "some of the concerns that critics might raise about these technologies." Topics covered including CO2 pipeline issues, slow leaks, fast releases, issues related to ocean ecology and issues related to hydrogen safety.We asked a series of specific evaluative questions.




Results from our pilot study suggested...

...that the standard concerns about siting would arise but may not be dramatically different than those associated with other large technologies.

Fears about rupture of high pressure CO2 pipelines did not appear to be any greater than those associated with natural gas pipelines.

Some concerns were expressed about large rapid releases from deep geological formations, but they looked manageable.

However, while deep geological injection appeared like it would prove to be publicly acceptable, the pilot results suggested that proposals for deep ocean injection were likely to lead to vigorous public opposition.




Ocean sequestration example commentsI think the concern that really strikes me the most would have to be pumping it under very high pressure into the deep ocean . . . I know the ocean is very big and is very deep but I’m wondering what kind of affect it would have on our oceans. [S1]

That, if this extra CO2 is absorbed into the ocean, would it disrupt whatever balance is in the ocean? That it might be harmful to things that live in the ocean. [S3]

Well, where are they going to build these? Do they have to be near the ocean? Or are they going to build big pipelines into the ocean to flush the stuff away? In the process of doing this, is there going to be pollution occurring, from this process? [S5]

So, I don’t necessarily like the fact that it’s being pumped down deep in the ocean, kind of like out of sight, out of mind. [S7]

So, if we were to put it, like, in the ocean, we could be messing with some form of life that’s on the bottom. I don’t think we have much knowledge of what’s down there. Because we really can’t explore that deep. So we’d be messing with something we have no knowledge of. [S8]




One of the things we noted……in our pilot studies was that subjects wanted to consider alternatives to CCD.

Accordingly, in designing the closed form survey, we included questions which allowed us to explore preferences across several alternative strategies for reducing CO2 emissions.

Survey outline:1. Background2a. Climate change 2b. General issues including climate change 3. Options for limiting CO2 4. Energy systems that dispose of CO2

5. Places to dispose of CO2

5.1 Deep rock formations5.2 Disposal of CO2 in the deep ocean6. Final evaluation7. Environmental issues (NEPS = New Ecological Paradigm Scale)

8. A few questions about yourself



Views on climate change

The continuing release of CO2 into the earth's atmosphere during this century may result in serious climate change.

1 4 7completely disagree

Government regulation should begin to significantly limit the amount of CO2 that is released into the earth's atmosphere.

Government regulation will begin to significantly limit the amount of CO2 that is released into the earth's atmosphere at some time in the next 20 years.

4.6

4.7

4.0




Not highon respondent's list of concerns

This is consistent with previous studies that we and others have done in the U.S. Results are quite different in Europe.

BUT, remember, despite this result, climate change clearly has political legs in the U.S. in many states.




OptionsOptions that will reduce CO2 emissions by 50%

from your energy consumptionRank

1 is what youwould pay the

most for.

Biomass Blend:50% from biomass, 50% from regular coalHalf of the energy comes from regular coal, and the other half comes from biomass. Biomassrefers to harvesting and burning special fast growing grasses and trees to make energy. Biomassreleases CO2 that plants have absorbed from the atmosphere so no new CO2 is produced.

Coal Blend with deep geological disposal of CO2:50% from coal with deep geological disposal of CO2, 50% from regular coalHalf of the energy comes from regular coal. The other half also comes from coal, but won't let theCO2 enter the atmosphere. Instead they will separate the CO2 and dispose of it thousands of feetunderground.

Coal Blend with deep ocean disposal of CO2:50% from coal with deep ocean disposal of CO2, 50% from regular coalHalf of the energy comes from regular coal. The other half also comes from coal, but won't let theCO2 enter the atmosphere. Instead they will separate the CO2 and dispose of it deep in the ocean.

Energy Efficiency:100% regular coalAll of the energy comes from regular coal. This option will replace appliances and lights in yourhouse with much more efficient ones so that you use half as much electricity while enjoying usingyour lights and appliances just as much as you do now. Using less electricity will result in lessCO2 being produced.

Hydro-electric Blend:50% from hydro-electric, 50% from regular coalHalf of the energy comes from regular coal, and the other half comes from hydro-electric. Hydro-electric refers to making electricity with water power from dams. Hydro-electric produces noCO2.

Natural gas Blend:100% from natural gasThe energy comes from burning natural gas in highly efficient plants. Burning natural gasefficiently produces about half as much CO2 as burning coal.

Nuclear power Blend:50% from nuclear power, 50% from regular coalHalf of the energy comes from regular coal, and the other half comes nuclear power.Nuclear power produces no CO2.

Solar power Blend:50% from solar power, 50% from regular coalHalf of the energy comes from regular coal, and the other half comes from solar. Solar powerrefers to making electricity with energy from the sun. Solar power produces no CO2.

Wind power Blend:50% from wind power, 50% from regular coalHalf of the energy comes from regular coal, and the other half comes from wind. Wind powerrefers to making electricity with energy from the wind. Wind power produces no CO2.

"Whatever your own beliefs are about climate change, imagine that the U.S. government has decided that we must cut in half the amount of CO2 that is released by generating electric power.

Suppose that your electricity supplier has different methods of meeting the goal to reduce CO2 emissions by 50%. Some of these methods use a mixture of generation systems that produce little or no CO2, combined with regular coal-burning power plants. These methods do not all cost the same because some ways of making electricity with less CO2 are cheaper than others.

The supplier will be offering their customers a choice of how they would like them to meet this reduction. Some of the options to reduce CO2 emissions will be more expensive than the way we produce electricity today, so we would like you to tell us which method you would be willing to pay the most for."




Rank order of options

Options used different generation mixes (base of coal) in order to reduce CO2 emissions by 50%.




Evaluationof

specifics

The next three slides, which I will not show, give you the numbers on these and other results.




Summary evaluations

1 4 7completely oppose

completely favor

1 4 7completely oppose

completely favor

Geological disposal:

before details

after details

before details

after details

Ocean disposal:




CCD Conclusions

The results of this study suggest that, at best, the public is likely to view this technology with mixed feelings.

High levels of public acceptance will almost certainly require:

• broader public understanding of the need to limit carbon dioxide emissions and alternative options for carbon management;

• a much stronger scientific understanding and a larger empirical base for claims about the likely efficacy and safety of disposal; and

• an approach to public communication, regulation, monitoring, and emergency response which is open and respectful of public concerns.




CCD Conclusions…(Cont.)

An open and inclusive approach does not guarantee success.

However, an arrogant approach such as the one adopted in the past by the industries responsible for nuclear power and genetically modified crops, could create a level of public distrust that makes the wide-spread implementation of geological carbon disposal in the U.S. difficult, if not impossible.


Department of Engineering and Public Policy Carnegie Mellon University 1 Risk Analysis and...

Documents

Transcript of Department of Engineering and Public Policy Carnegie Mellon University 1 Risk Analysis and...