NCSE Conference

1
Some Curves and a Bird
NCSE Conference February 4, 2005
David Rejeski Director, Foresight and Governance
Project Woodrow Wilson International Center for
Scholars
2
Forecasts and Policy Change
Things work fairly well if - Change is
predictable, i.e., linear with respect to
cause and effect - Change is moderate, not too
slow or too fast - The system is bounded -
Unintended consequences are minimal (or such
consequences are controllable) - System
feedback is low or negative (damping
effects) - The rate of change in the
environment does not exceed the rate at which
governing institutions can change and adapt
MODERATE CHANGE
3
Emergent Threats and Challenges
4
Slow Threats
Electricity Prices inCalifornia (20-30)
X
Cost of Prescription Drugs (17)
X
Percent per year change
7
6
X
Adult Obesity
X
College Tuition
5
People Needing Humanitarian Assistance
X
Refugee Population
X
4
Threshold of political concern
U.S. CO2 Emissions
X
3
New Tuberculosis Cases (Globally)
X
X
Inflation
2
U.S Vehicles Miles Traveled
X
X
U.S Population Growth
X
X
1
Global Deforestation
U.S. Family Income
5
Slow Threats Meet Political Realities
Tell Me Again Why I Should Do Something About This
What the Scientist Sees
CO2
End of an evolutionary phase
One Million Years
6
Technology-Related Slow Change
Last new products and industries. Total market
saturation
Full constellation of industries and
infrastructure
Technological maturity and market saturation
Full expansion of innovations and market potential
Early, new industries. Explosive growth.
About 50 years
Adapted from Perez, C. (2002) Technological
Revolutions and Financial Capital.
7
Technology-Driven Exponential Change
Displays Moores Law Storage 1.5Xs
Moores Law Bandwidth 2Xs Moores Law GPUs
2-3Xs Moores Law
8
Step Change/Convergence
9
You will be Surprised, but How?
Out-of-the-Blue We never saw anything like this
before.
Direction Anticipated north, went west.
Anticipated plus, but minus occurred.
Amplitude Anticipated 1X change, but 3X happened.
Speed Figured it would take 2 years, not 6
months.
Frequency Usually happens every 50 years, not 10.
10
The Ultimate Surprise The Black Swan
A real outlier beyond normal human expectations
Cygnus atratus

• However, black swans are not unthinkable

See www.fooledbyrandomness.org
11
Take Away Lessons for Forecasting

• Know your modeler, scenario builder, etc.

• Address bias, especially cognitive bias

• Know what is left out of the analysis (as
  important
• as what is included)

• View uncertainty as valuable information

• Seek multiple viewpoints/perspectives(requisite
  variety)

• Developed a nuanced view of surprises

• Combine techniques include novel approaches

• Remain humble

12
Failure Modes in the Policy World
Failure of Imagination Occurs when all the
available facts and forecasts are marshaled
correctly -- but when the really vital facts are
still undiscovered or the possibility of their
existence not even admitted.
Heuristic Bias
Failure of Nerve Occurs when given all the
relevant facts, the policy maker cannot see that
they point to an inescapable conclusion.
Pattern Bias
From Clarke, Arthur (1962) Failures of Nerve
and Imagination, Profiles of the Future.
13
Failure Modes in the Policy World (2)
1. Failure of Imagination We fail to anticipate
a problem, think around limitations, or develop
innovative solutions.
2. Failure of Perception Once the problem
arrives, we fail to perceive it.
3. Failure of Nerve After we perceive the
problem, we may fail to address it.
4. Management Failure We may address it, but
use the wrong techniques, and fail to
solve it.
From Jared Diamond, Why Some Societies Make
Disastrous Decisions, April 28, 2003, Edge
(www.edge.org)
14
Addressing Management Failure
Simulation Gaming Instrumentation
Instrumentation
15
Final Thoughts
I never predict the future, I just look out the
window and see what is visible, but not yet
seen. Peter Drucker
Perspective is worth 80 I.Q. points. Alan Kay
16
For Further Information
www.foresightandgovernance.org www.environmentalf
utures.org www.wilsoncenter.org/foresight rejes
kidw_at_wwic.si.edu