Improving policy analysis for global change issues: a synthesis from the CISHDGC values, learning and decision processes projects

1
Improving policy analysis for global change
issues a synthesis from the CISHDGC values,
learning and decision processes projects

• By
• Tim McDaniels
• University of BC, CISHDGC

2
Outline

• Introduction
• Kinds of and structures of decisions?
• Decision processes?
• Approaches to values for these decisions?
• Role of learning?

3
1.) Introduction

• This talk has three points of departure
• Morgan, et al editorial why standard approaches
  to PA are insufficient for climate change
• Hard versus soft systems analysis
• Kates et al, 2000 Sustainability science

4
Assumptions, Inadequacies of Standard PA (Morgan,
et al)

• Single DM, single problem, agreement on a single
  societal view
• Non-marginal, manageable change
• Values known, static Utility max is goal
• Time preference discounting sensible
• Uncertainties modest, manageable
• Systems linear

5
Hard vs Soft systems analysis (Checkland et al)

• Hard clear decision, well bounded, values
  defined, single DM, uncertainties definable,
  analytical/optimization tools (OR) applicable
• Soft problems ill defined, vaguely bounded,
  values unclear, no obvious DM,
• Their view on path forward soft/hard
  complementarity e.g., use soft systems ideas to
  structure analysis for hard systems application

6
Sustainability Science (Kates, et al, 2000)

• Recognizes limits of reductionist science for
  global change problems, sustainability
• Argues for a new, integrative, multiple scale,
  north/south sensitive, human/natural systems
  approach to science
• Their concern is only a new perspective on
  science
• Says nothing about how to conduct analysis to
  analyze, make choices, implement, informed by
  that science

7
CISHDGC research contributions

• Many incremental steps toward addressing and
  improving on the standard tools of policy
  analysis for global change
• Imagine matrix of inadequacies (listed earlier)
  and contributions from investigators to address
  them
• Need a summary effort to recognize, assemble our
  findings and insights put forward perspectives,
  approaches to PA for global change
• Here focus on those associated with problem
  choice, decision process, values, learning

8
2.) What kinds of decisions?

• Conventional PA for GC often addresses the
  global decision (CC emission controls) or single
  level, constrained decisions (e.g. resource
  management, env health, disaster preparedness)
  given status quo
• We have addressed the links among decisions
  across scales, time, groups
• Reasoning is that such decisions are so path
  dependant (time) or scale dependant (levels of
  governance) we need to address such links to
  make sensible progress

9
Multiple scales and levels

• Match scales of assessment, and levels of
  regulation and administration to the problem
• McDaniels and Dowlatabadi (2004)
• Effective regulation across levels requires both
  appropriate knowledge and effective jurisdiction
• Appropriate Knowledge nature and dynamics of
  impacts, values, alternatives
• Effective Jurisdiction compatible with scales,
  competent, legitimate

10
Implications of gaps, mismatches

• Example of salmon aquaculture
• A site-by-site regulatory regime, but cumulative
  impacts are the predominant concern
• Site-based regulation not really feasible to
  judge sustainability of industry
• Local governments have attempted to regulate
  land use to respond to regional, provincial
  concerns

11
Short term objectives for longer term ends

• Framework of multiple objectives for
  characterizing what is desired in CC policies
  over the next 20 years
• Emphasis on learning as an explicit objective
• Intent is create the capability, opportunities
  for better decisions 20 years into future
• Keeney and McDaniels 2002

12
Short term decisions to get to long term desired
ends

• Georgia Basin Project, Quest
• Visualization tool, gaming, value formation
• Given an expressed desired future for a region
  (where we want to go) , what are the important
  near-term decisions that must be made within the
  regions governance structure to get onto the
  desired paths (how to get there) what are
  tradeoffs for other related decision contexts
• Charlie Wilsons current work (2004)

13
3. Approaches to decision processes

• Most policy analysis guides recognize benefits of
  iteration, recognition of different elements of
  good decision making
• Yet how to make this work on extremely complex
  problems, with stakeholders, addressing GC
  problems, isnt well understood
• How to proceed through decisions, involving whom?

14
Whats good decision process?

• Hammond, Keeney and Raiffa, 1999
• Smart Choices
• Keeney, 1992
• Value-focused thinking
• Neither cast in terms of, or really address
  environmental choices, particularly with
  stakeholder group, which have their own kinds of
  complexities

15
Efforts regarding decision processes for GC
decisions

• Conceptual elements of process and application
  for decision aiding in any context
• (McDaniels and Gregory 1996-2004)
• Experimental good group process can lead to
  improved valuation decisions (measured via
  whether standard biases can be avoided)
• (McDaniels, Gregory, Arvai, Chunpagdee, 2003)

16
Applied Findings

• Water use planning efforts in BC
• Good process, attention to values, AM all
  contributed to achieving great outcomes,
  consensus decisions for electric utility planning
  with stakeholders to balance power, fisheries,
  flood control, recreation
• McDaniels, Gregory, Fields, 1999 Gregory,
  McDaniels and Fields, 2002 others in process

17
Participatory Action processes

• Will Trousdales work in Philippines and Jamil
  Bundallis work in Kenya shows that VFT is a good
  way to structure, conduct participatory action
  research (Tourism, rural energy planning)
• McDaniels and Trousdale (1996)
• Jamils MSc project draft underway (2004)

18
4.) Values and Valuation

• Multiple objective approaches seem only sensible
  basis for valuation of global change issues
• fewer strict assumptions about decision rules,
  how values are measured, role of income (How much
  should society pay, not how much are you WTP)
• McDaniels, 2001, Encyclopedia of GC
• Judgment tasks must be kept manageable, help
  construct values in context
• Voting, referendum notions for preference
  elicitation
• McDaniels, 1996, McDaniels and Thomas, 2000

19
Values for Non-marginal change

• Work on compensation for Metis lands losses has
  helped us understand both what is needed to value
  non-marginal change, and applied approaches for
  work with indigenous people on valuing their
  intangible, spiritual, cultural losses from loss
  of resources over 50 years
• McDaniels and Trousdale, 2004

20
What is needed for valuing, compensating for NMC?

• A legal property rights regime establishing
  compensation for losses (all kinds)
• envisioning what the future(s) could be with and
  without the change
• Understanding what other changes will go on
  around, outside of this particular change (the
  context)
• judging what values matter in comparing with and
  without cases
• Judging the relative importance of the range of
  values affected by the change, from the
  viewpoint of the affected parties
• Use of a metric, method for aggregation, across
  the different kinds of values, w and w/o the
  change (indirect processing out if one impact is
  in dollars)
• Some means of addressing the impact changes,
  aggregation over time
• Using the results to argue for compensation, or
  better still, the kinds of actions to mitigate
  losses

21
Valuation within context of alternatives

• Issues of evaluability (Hsee)
• Experimental work on values within a structured
  process, with alts, and a context, compared to
  standard practice
• Use of structured decision process reduces
  standard heuristics, biases (embedding)

22
5.) Learning (AM) as a strategy

• Three projects concerned with learning
• Objectives that comprise a framework for for
  short term CC policy decisions that emphasize
  learning
• Keeney and McDaniels 2001
• Treating learning as an objective in working with
  SH groups, the value of learning
• McDaniels and Gregory 2004

23
Newest AM project learning plan for salmon
aquaculture

• Developing a learning planfor salmon
  aquaculture
• First step what are the questions people want to
  answer (SH workshop, some problems)
• Future steps refining questions, developing
  strategies to answer them, setting priorities
• Other parts a FN plan, Issues of time

24
6. Conclusions, implications

• The effort to build new approaches to policy
  analysis that can adequately address the kinds of
  issues called for in Sustainability Science could
  take as much work as, is linked to, the science
  itself
• It is at least as important as the science in
  making better decisions

25
Where we have come to

• In total, the UBC work has made some progress on
  parts of the agenda for building the new PA
• We have highlighted only some of it (nothing here
  on risk perception, risk communication, other
  work on values and decision making we have
  pursued)
• Our emphasis has been on the problem-structuring
  and values-based side
• Far more needed, huge opportunities
• Making better decisions is a big concern for
  making new science relevant and helpful