Transcript Public Perspectives - National Academy of Engineering

UNEP Paint for the Planet: Charlotte Sullivan, 13 Years, England
Public Perspectives:
Trust, Engagement, and Engineering
Expertise
Workshop on Climate, Society, and Technology
Beckman Center of the National Academies
June 7, 2011
Ann Bostrom
University of Washington
Outline
• Expert (engineering) risk assessments
• Lay risk perceptions
• Two steps to better decisions:
1. Engage stakeholders
• Values and uncertainty
• Technical and translational challenges
2. Emphasize effective engineering
• What to do
• How it cuts losses
Occupational Outlook Handbook, 2010-11 Edition:
Estimated temperature increase in °C
IPCC μ=0.73
Class μ=1.1
σ=0.68
σ=1.8
Median=1.8
IPCC μ=0.19
σ=0.04
Class μ= 2.2
σ=1.7
Median=1.1
2009 μ=2.6
σ=5.0
Median=1.7
2009 μ=3.9
σ=7.0
Median=1.5
1992 μ=2.7
σ=3.3
Median=2
1992 μ=3.2
σ=4.3
Median=2
To date
10 years
IPCC μ=1.39
Class μ=5.5
σ=0.26
σ= 6.0
Median=3.3
2009μ=6.3
σ=13.3
Median=2.8
1992 μ=6.5
σ=9.1
Median=4
2050
How do people understand probabilities?
•
Which is larger?
1/100,000
1/10,000
• Peters et al (2007) found that ~10% of respondents had
trouble with this kind of comparison.
Affect: evaluability
Affect heuristic: risk-as-feelings
(heuristic = rule-of-thumb or mental shortcut)
• “affect” means the specific quality of
“goodness” or “badness”
I. experienced as a feeling state (with or without
consciousness) and
II. demarcating a positive or negative quality of a
stimulus.
• Affective responses occur rapidly and
automatically
• Reliance on such feelings = “the affect
heuristic”
Affect: Evaluability
Explanations:
1. Ease of mapping stimulus onto response scale (compatibility)
2. The more precise the affective impression, the greater the
influence on judgment and decision-making.
– “Affect bestows meaning on information”
– When consequences have strong affective meaning, people tend
to be relatively insensitive to the probabilities of those
consequences.
Trust, Confidence and Cooperation – T. Earle and M. Siegrist
Journal of Applied Social Psychology, 2006, 36, 2, pp. 383–416.
Cultural worldviews
“preferences for how to organize society”
Cultural worldviews
“preferences for how to organize society”
Risk perception
Polarization
Expert risk assessment
• Expertise is by definition narrow
– Experts solve problems differently than non-experts, in
their domain of expertise.
– Superior recognition gives experts an advantage
– 10,000 hours….
• BUT, expertise is only one piece…
– Trust AND confidence
– Public processes and values
– Members of the public may have local expertise
Risk perception
• “Lay risk assessment” without formal decomposition into
probability and harm
• Probability neglect; subjective probabilities tend to be
“compressed”
• Risk perception is affective, social, and cognitive
– Affect heuristic - common source used to evaluate both
benefits and risks
– Social and cultural differences - individualism, hierarchy
and egalitarianism
– Information processing - attention and memory, problem
solving as search
• Risk attitudes and behaviors depend on perceived threat and
efficacy - Will it harm me? Can I control it?
Risk perception
Mental models of hazardous processes
• Understanding of causal processes - exposure,
effects, mitigation
Method:
– Decision analysis
– Semi-structured interviews, surveys
– “open” comparison with formal decision model
• By analogy? e.g., lead is like mercury.
“I don’t know what happens. Is it like mercury,
where it is there forever?”
Tell me about the issue of climate change
• Oh! something I think about all the time. I pay attention to
the weather myself all of the time. I’m just sort of an
amateur weather forecaster, and I’ve been doing this since I
was a child. I know what clouds mean. I’ve definitely seen a
change in my lifetime. The mid-80’s I started paying
attention.
Can you tell me more about that?
• I feel like I’m seeing differences in weather patterns. I live in
Portland myself and we’ve had a much dryer spring so far.
Even though it’s not technically spring I’ve had to water my
garden a few times, already, which kind of scares me a little
bit.
- Seattle, 2008
Ozone?
[Tell me all about the
issue of climate change]
[If you were to explain climate change to
someone else, is there anything you would add
to what you have said?]
“I probably would go into a little more depth
about how the ozone works and how we’re
holding in—how the warming is actually
happening to the best of my knowledge.”
• Respondent 102:
• “Well, I don’t think it’s in
[You mentioned ozone. Can you explain what
dispute anymore, um I
ozone has to do with it?]
think we are pumping
“The ozone is kind of the protective layer around
out too much waste into
the earth and that it allows sunlight in and it
our environment. Gasses, allows sun heat out, but at the rate that we are
emissions from cars, coal
producing things like uh, carbon dioxide and
plants, etc. And that’s
other gasses that our ozone is holding too much
warming. And I think that
in, which means that it’s heating the earth
everybody thinks we
essentially. Um, that’s kind of my basic
should do something
understanding of how it works.”
about it but I think it’s
probably too late.”
Climate change expert model:
a hierarchical influence diagram
Tr ans por t at io n
Sola r r adia t io n
aut om obile s
Coas t al Zones
eus t at ic
s ea le v el
bus & r ail
la ndus e
s t or m dam age
r is e
t r ans por t at io n
er os io n
air c r af t
geophy s ic al
pr oc es s es
def or es t at io n
Alb edo
s alt wat er
in t r us io n
in aquif er s
Build in gs
Clim at e ( <T>,
weat her , et c . )
wat er
heat
dis t r ib ut io n of
s t or m f r equenc y
& s ev er it y
lig ht s
CO 2
em is s io ns
build in gs
c hr onic f lo odin g
ec os y s t em
s er v ic es
r ec r eat io n
r adia t iv e bala nc e
ec os y s t em
im pac t s
( i. e. bio div er s it y )
la ndus e
popula t io n
m ig r at io n
dis t r ib ut io n
of r ain f all
f r es h wat er
s upplie s
Aer os ols
applia nc es
f os s il
f uel us age
N2O
em si s io ns
G HG s in k s
Pla nt s
agr ic ult ur al
pr ac t ic es
s pac e c oolin g
Tem per at ur e
s pac e heat
s oil m ois t ur e
dis t r ib ut io n
CH4
em is s io ns
Land and s ea ic e
agr ic ult ur al
wat er
c onc ent r at io n
of gr eenhous e
gas s es
agr ic ult ur al
y ie ld s
Agriculture
Hum id it y
dis t r ib ut io n
f ood
s upply
r edis t r ib ut io n
of wealt h, lo s s of
r es our c es ,
in c r eas ed m or t alit y
c lim at e s y s t em
CFC
pr ac t ic es
Volc anic ac t iv it y
build in g
ener gy us e
em is s io ns
build in g
t ot al wat er
v apour in
at m os pher e
I ndus t r y / Ut ilt ie s
s oc io - ec onm ic im pa c t s
heat in g/ c oolin g
build in gs
ele c t r ic it y f r om
f os s el f uel
G r eenhous e gas es
Dom ain of geoengin eer in g s t r at gie s
in dus t r ia l
ener gy
us age
in dus t r ia l
s olv ent s &
lu br ic ant s
c ons um er
dem and f or
goods & s er v ic es
Dom ain of Adapt at io n St r at gie s
nuc le ar ener gy
Renewable
ener gy s our c es
Dom ain of Av oid anc e St r at gie s
Deforest
ation
Buildings
Water
heat
Lights
Buildings
Appliances
Fossil
fuel
usage
Space
heat
Space
cooling
CO2
GHG
sinks
N2O
Concentration of
Greenhouse gasses
Agriculture
CH4
CFC
Industry/Utilities
Water
vapor
Electric
Greenhouse gases
Industry
energy
use
Industrial solvent
and lubricants
Consumers
Nuclear
energy
Renewable
energy sources
Domain of Avoidance Strategies
= Domain of Expert Model of Climate Change
Global Climate Change: Expert Model
= Component of Expert Model of Climate Change
Transportation
Automobiles
Coastal Zones
Bus & rail
Land use
Eustatic sea level
rise
Solar radiation
Storm damage
Transportation
Aircraft
Erosion
Salt water
intrusion into
aquifers
Ecosystem
services
Chronic flooding
Recreation
Climate System
Deforestation
Buildings
Geophysical
processes
Water heat
lights
Radiative
balance
Albedo
Climate (<T>,
weather, etc.)
Distribution of
storm frequency
and severity
Ecosystem
impacts (e.g.
biodiversity)
Land use
Population
migration
Buildings
CO2 emissions
appliances
GHG sinks
Fossil fuel usage
Distribution of
rainfall
Agricultural
practices
Fresh water
supplies
Space heat
Space cooling
Plants
Aerosols
N2O emissions
Concentration of
Greenhouse gasses
CH4 emissions
Temperature
distribution
Food supply
Land and
sea ice
Soil moisture
Agricultural
yields
Volcanic activity
Redistribution of
wealth; loss of
resources;
increased mortality
Water
Agricultural
practices
Humidity
distribution
Agriculture
CFC emissions
Total water
vapor in
atmosphere
Industry/Utilities
Electricity from
fossil fuel
Building energy
use
Domain of Geoengineering Strategies
Building heating
& cooling
Industrial
energy use
Industrial solvents
and lubricants
Consumer
demand for
goods and
services
Buildings/Infrastructure
Socioeconomic Impacts
Greenhouse gasses
Domain of Adaptation Strategies
Nuclear energy
Renewable
energy sources
Domain of Avoidance Strategies
= in expert model and cited
by 1 or more respondents
Global Climate Change: “Best” Models
= in expert model but not cited by
more than 1 respondent
= item not in expert model but
cited by more than 1 respondent
Transportation
Sun-earth
geometry
Automobiles
Coastal Zones
Bus & rail
Land use
Eustatic sea level
rise
Solar radiation
Storm damage
Transportation
Aircraft
Erosion
Fire
Salt water
intrusion into
aquifers
Ecosystem
services
Chronic flooding
Recreation
Climate System
Deforestation
Buildings
Geophysical
processes
Water heat
lights
Radiative
balance
Albedo
Climate (<T>,
weather, etc.)
Distribution of
storm frequency
and severity
Ecosystem
impacts (e.g.
biodiversity)
Land use
Population
migration
Buildings
CO2 emissions
appliances
GHG sinks
Fossil fuel usage
Distribution of
rainfall
Agricultural
practices
Fresh water
supplies
Space heat
Space cooling
Plants
Aerosols
N2O emissions
Concentration of
Greenhouse gasses
CH4 emissions
Temperature
distribution
Food supply
Land and
sea ice
Redistribution of
wealth; loss of
resources;
increased mortality
Soil moisture
Agricultural
yields
Volcanic activity
Water
Agricultural
practices
Humidity
distribution
Agriculture
CFC emissions
Total water
vapor in
atmosphere
Industry/Utilities
Electricity from
fossil fuel
Domain of Geoengineering Strategies
Building energy
use
Local weather
change
Building heating
& cooling
Industrial
energy use
End of
civilization
Industrial solvents
and lubricants
Consumer
demand for
goods and
services
Season Change
Renewable
energy sources
Buildings/Infrastructure
Socioeconomic Impacts
Greenhouse gasses
Domain of Adaptation Strategies
Recycling
Nuclear energy
Human health
effects
Pollution
Nonspecific
human/ animal
Domain of Avoidance Strategies
= in expert model and cited
by 1 or more respondents
Global Climate Change: “Worst” Models
= in expert model but not cited by
more than 1 respondent
= item not in expert model but
cited by more than 1 respondent
Transportation
automobiles
Coastal Zones
Bus & rail
Land use
Eustatic sea level
rise
Solar radiation
Storm damage
transportation
aircraft
erosion
Salt water
intrusion into
aquifers
Ecosystem
services
Chronic flooding
recreation
Climate System
deforestation
Buildings
Geophysical
processes
Water heat
lights
Radiative
balance
albedo
Climate (<T>,
weather, etc.)
Distribution of
storm frequency
and severity
Ecosystem
impacts (e.g.
biodiversity)
Land use
Population
migration
buildings
CO2 emissions
appliances
GHG sinks
Fossil fuel usage
Distribution of
rainfall
Agricultural
practices
Fresh water
supplies
Space heat
Space cooling
plants
aerosols
N2O emissions
Concentration of
Greenhouse gasses
CH4 emissions
Temperature
distribution
Food supply
Land and
sea ice
Redistribution of
wealth; loss of
resources;
increased mortality
Soil moisture
Agricultural
yields
Volcanic activity
Water
Agricultural
practices
Humidity
distribution
Agriculture
CFC emissions
Total water
vapor in
atmosphere
Industry/Utilities
Electricity from
fossil fuel
Domain of Geoengineering Strategies
Building energy
use
Local weather
change
Building heating
& cooling
Aerosol cans
Industrial
energy use
Human health
effects
End of
civilization
Ozone
depletion
Industrial solvents
and lubricants
Consumer
demand for
goods and
services
Season Change
Buildings/Infrastructure
Socioeconomic Impacts
Greenhouse gasses
Domain of Adaptation Strategies
Nuclear energy
Renewable
energy sources
Pollution
Domain of Avoidance Strategies
Global Climate Change: Average Model
= mentioned by
50% or more
respondents
= mentioned
by 25%-50%
of
respondents
= item not in expert model but cited by
more than 10% of respondents
Transportation
= mentioned
by 10%-25%
of
respondents
= item in expert model but not cited by
more than 10% of respondents
Sun-earth
geometry
automobiles
Coastal Zones
Bus & rail
Land use
Eustatic sea level
rise
Solar radiation
Storm damage
transportation
aircraft
erosion
Fire
Salt water
intrusion into
aquifers
Ecosystem
services
Chronic flooding
recreation
Climate System
deforestation
Buildings
Geophysical
processes
Water heat
Lack of oxygen
lights
Radiative
balance
albedo
Climate (<T>,
weather, etc.)
Distribution of
storm frequency
and severity
Ecosystem
impacts (e.g.
biodiversity)
Land use
Population
migration
buildings
CO2 emissions
appliances
GHG sinks
Fossil fuel usage
Distribution of
rainfall
Agricultural
practices
Fresh water
supplies
Space heat
Space cooling
plants
aerosols
N2O emissions
Concentration of
Greenhouse gasses
CH4 emissions
Temperature
distribution
Nonspecific
water impacts
Land and
sea ice
Food supply
Redistribution of
wealth; loss of
resources;
increased mortality
Soil moisture
Agricultural
yields
Volcanic activity
Water
Agricultural
practices
Humidity
distribution
Agriculture
CFC emissions
Total water
vapor in
atmosphere
Industry/Utilities
Electricity from
fossil fuel
Domain of Geoengineering Strategies
Building energy
use
Local weather
change
Building heating
& cooling
Aerosol cans
Industrial
energy use
End of
civilization
Industrial solvents
and lubricants
Consumer
demand for
goods and
services
Season Change
Renewable
energy sources
Buildings/Infrastructure
Socioeconomic Impacts
Greenhouse gasses
Domain of Adaptation Strategies
Recycling
Nuclear energy
Human health
effects
Ozone
depletion
Pollution
Nonspecific
human/ animal
Domain of Avoidance Strategies
Public risk assessment, management and policy
• Problem solving as search – iterate between designing for safety and assessing risk
– broad engagement to avoid narrow bracketing
• Participative decision making can produce better
decisions (NRC, 2008).
– Decision aiding
– Mediated modeling
– Use (social) science and (new) technology to engage
people
Perceptions in risk assessments and decisions –
a two-step
1. Engage stakeholders to address:
•
•
Values and uncertainties
Technical and translation challenges
2. Emphasize effective actions!
•
•
What to do
How it cuts losses
Engage stakeholders: Values
• What endpoints are valued
(deaths, dollars, displacement?)
• How and by whom they are valued
• Not-so-obvious distributive implications, as with
airbags
• Methodological and ethical challenges (Fischhoff,
1991; Slovic, 1995)
Engage stakeholders: Uncertainty
• Both limits to what is known and
inherent variability
• How should uncertainty be treated and
interpreted?
• How uncertainty is analyzed can change
the outcome
• Representation of uncertainty can
influence interpretations and
preferences
Engage stakeholders:
Technical and
translational challenges
• Risk assessment is
analytically and dataintensive
• Even large-scale tests
may be insufficient to
discover devastating
rare adverse effects
• Results may be lost in
translation – 10% PE?
• Engaging engineers with
other parties in
facilitated discussions
can help.
USEFUL METRIC CONVERSIONS
1 million microphones = 1 megaphone
1 million bicycles = 2 megacycles
2000 mockingbirds = two kilomockingbirds
1 millionth of a fish = 1 microfiche
10 monologs = 5 dialogues
2 monographs = 1 diagram
8 nickels = 2 paradigms
2 baby sitters = 1 gamma gampa
- Anonymous
Effective actions!
• What to do
• How it cuts losses
– Protective actions
– Engineering tools and strategies
NUDGES: Choice Architecture
iNcentives: people have to feel they are getting something for their choice
Understand mappings: you have to understand how they see things
Defaults: make sure the ‘do nothing’ route is one of the best
Give feedback: investigate the rejected options, and experiment with them
Expect error: humans make mistakes, so well-designed systems allow for this
Structure complex choices: if it’s difficult, break it down into easier chunks
In sum:
• Expert (engineering) risk assessments can differ
from
• Lay risk perceptions
• But together they produce better decisions:
1. Better engineering by engaging stakeholders:
• Values and uncertainty
• Technical and translational challenges
2. Better actions by emphasizing systematic
evaluation of effectiveness:
• What to do
• How it cuts losses
Thanks to the U.S. National Science
Foundation (NSF9701785,
NSF0729302), Cynthia Atman,
Douglas Bostrom, Vanessa Collier,
Robert Dooley, Baruch Fischhoff,
Rebecca Hudson, Lester Lave, Granger
Morgan, Daniel Read, Travis Reynolds,
and many others.
UNEP Paint for the Planet: Charlotte Sullivan, 13 Years, England
Affect: evaluability
Kahneman & Tversky say
We evaluate prospects and tradeoffs from a starting
point; losses loom larger than gains… Prospect Theory