Ex Ante Impact Assessment and Seasonal Climate Forecasts

Ex ante impact assessment
and seasonal climate forecasts:
status and issues
Philip Thornton
International Livestock Research Institute, Nairobi, Kenya
Institute of Atmospheric and Environmental Sciences,
University of Edinburgh, Scotland
International Workshop on Climate Prediction
and Agriculture: Advances and Challenges
WMO, Geneva, 11-13 May 2005
Outline
• Ex ante impact assessment
• Some methods and tools
• Impact assessment and climate
forecasting
• Moving the agenda forward
2
Presentation focus
•
Ex ante methods of assessing impacts at an
aggregated level
•
•
Change as a result of
•
•
•
3
potential societal impacts of change in
agricultural systems
indigenous innovation
research (technology, policy)
drivers such as population growth
Presentation focus
4
•
A large and growing literature on ex ante
assessment of climate forecast use at the
household and individual level
•
Much less seems to have been done at
aggregated levels
A traditional view of impact assessment
Research
Project
The Adoption Period
A%
Impacts
Year
0
Year
X
Research
Output
Year
Y
Year
Z
When adoption
reaches highest
level
0
Development
Output
Adoption
on-farm
Research costs
Extension costs
Adoption costs on-farm
5
Adapted from Randolph et al. (2001)
A traditional view of impact assessment
• A vast literature exists based on this model
• The effectiveness of this type of ex ante IA is
dependent on monitoring and evaluation
• In practice, if things cannot be valued relatively
easily, they tend to be ignored
• Sees the innovation process as being highly linear
and one-way
6
Sayer and Campbell (2003)
7
Another view of impact assessment
Adaptation, uptake, dis-adoption
%
Impacts on:
Year
Y
Year
0
Adaptation
Updating
Subsystem
Identification
8
Action
Successive INRM
learning cycles
Reflection
Implementation costs of doing INRM
• Production
• Income
• Food security
• Vulnerability
• Adaptive capacity
• ...
Questions of ex ante impact assessment
However the innovation process is seen, it involves
some sequence of change  uptake  impact, and
there are common questions to be answered:
• Who are the clients?
• Impact where?
• Impact on whom?
• Which impacts?
• How to value the impacts?
9
Who are the clients for ex ante impact
assessment?
• Policy makers at national, regional, local level (decisions to
be made in pursuit of policy objectives)
• Donors (priority setting, targeting)
• Researchers (priority setting, targeting)
• Private sector (investment decisions)
• General public (direct impacts of the use of public
resources)
10
Impact where, and on whom?
• Physical location – “recommendation domains”,
targeting
• Characteristics of target populations in these
areas
11
Site selection, Sub-Saharan Africa Challenge Programme
12
Spatial data
Non-spatial data
• Administrative boundaries
• Climatological data
• Farming systems
• Length of growing period
• Livestock populations
• Market access
• Human population
• Soils and erosion risk
• Vegetation cover
• Protected areas
• Watersheds, lakes, rivers
• Institutional environment
• Policy environment
• Local livelihood options
• Critical health issues
• Broad poverty trends
• Social capital
• Commercial sector linkages
• Added value
• Representative-ness
• Potential for impact
SSA-CP site selection
Site characteristic
LGP (months)
Annual rainfall (mm)
Relief
Lake Kivu
ZimbabweMozambiqueMalawi corridor
2.5 - 6
>9
>5 - 10
500-1100
1,500-2,000
700 - 800
Mostly flat
Mostly
intersected with mountainous
inland valleys
From
mountainous to
flat plains
towards coast
Policy environment
Medium
Weaker
Weaker
Market environment
Medium
Weaker
Medium
Institutional
environment
Stronger
Weaker
Stronger
Principal NRM issue
Site area (km2)
13
Kano, Katsina,
Maradi
Soil nutrients
83,900
Vulnerability
19,500
Soil fertility
management
274,000
SSA-CP extrapolation
domain for Lake Kivu
Elevation > 1500 m
Rainfall > 800 mm
Pop density > 50 / km2
Access indicator < 90
Area 19,500  361,700 km2
Population (2000) 15  69 million
Population (2030) 29  131 million
14
Notenbaert (2004)
Which impacts, and how to value them?
Which impacts will depend on the situation:
• Production, productivity
• Poverty alleviation
• Food security
• Environment
• Capacity building
• Commodity prices for consumers
• Others ...
15
Production objectives of livestock keepers in
Vryberg District, Northwest Province, RSA
Commercial
• Raise calves for market
(reproductive capacity of the
herd is key)
16
Communal
• Maintain cattle as a capital and
social asset
• Age-sex composition of the
herd is carefully controlled
• Maintain as large a herd as
possible, sell animals only in
extremis
• Want quick turn-over in calf
production
• Practise goat production as a
hedge against drought
• Cull unproductive animals
• Do not under-utilise pasture
Hudson (2002)
Mixed crop-livestock systems in Kenya and N Tanzania
after Seré and Steinfeld (1996)
17
Characteristics of four maize-based mixed systems
identified in the Eastern and Southern Africa region
Small intensive
Population
Functions
of livestock
(persons km-2)
Dairy,
SSImanure
>250
Dairy, MSI
manure, draft 100-250
meat, manure 30-100
Medium semi-intensive Draft,MSSI
Draft,
Medium extensive
MSE meat
<30
Medium intensive
% of land cultivated
>20
10 - 20
1.5 - 10
0 - 1.5
Source: Thorne et al. (2002)
18
Evaluating the impacts (a subsample)
Method
19
Description
Pros, Cons
Suitability for Assessing
Change
Uptake
Impacts
Low
Ad hoc
Informal assessment
involving little
analysis
Cheap and quick;
sometimes not
very good
Low
Low
Scoring
methods
Measurable
indicators and
weights assigned to
a set of criteria and
the results ranked
Intuitively
appealing, hard to
scale indicators to
match policy
objectives
Medium
Medium
Economic
surplus
Estimate how
change will improve
on-farm productivity
and reduce costs of
production and
consumer prices
Comprehensive,
data demanding
and needs
analytical skill
High
Medium
Medium
“Harder”
simulation
models
Assess biophysical
impacts at a range
of scales using
quantitative models
Data intensive,
time consuming,
difficult to calibrate
and test
Low
Low
High
Medium
Information needed for an ex ante assessment
Stage 1. Change (e.g. research)
How to obtain
Level of
uncertainty
Resources required
Time
Partnerships and skills
Intermediate and final outputs
Probability of success
20
Mod
Peer review
Scoring methods
Econometric methods
Mod
Low
Mod
Mod-High
Information needed for an ex ante assessment
Stage 2. Uptake
How to obtain
Level of
uncertainty
21
Who, characteristics
GIS, surveys
Mod
Where, characteristics
GIS, surveys
Mod
Infrastructure needed
GIS
High
Policies needed
Surveys
High
Adoption rate, ceiling
Scoring methods
High
Costs involved
Scoring methods
Mod-High
Information needed for an ex ante assessment
Stage 3. Impact quantification
How to obtain
Level of
uncertainty
22
Production
Biophysical models
Mod
Income
Household models
Mod
Environment
Models, scoring
High
Capacity building
Scoring methods
High
Costs, prices, elasticities
Lit review, surveys
Mod
Challenges in doing ex ante impact assessments
related to climate forecasts
1. The nature of climate forecasts
Which impacts to measure?
Seasonal climate forecasts may modify risk, and this has to be
taken into account
Impacts on whom?
People grow crops and keep livestock for various reasons, not
all to do with food production and cash generation
How to assess uptake?
Seasonal forecasts may be inaccurate
Their uptake will depend on credibility of the source and
forecast skill
23
Challenges in doing ex ante impact assessments
related to climate forecasts
2. The need to assess impacts across time and space
Which impacts to measure?
Aggregate impacts of seasonal climate forecast use may
substantially modify local prices
Impacts of modified management may be felt over entire
production cycles, or even multiple production cycles
24
Challenges in doing ex ante impact assessments
related to climate forecasts
3. Assessing what is required of the institutional and policy
environments
How to assess uptake?
What support is likely to be necessary, and how much may it
cost to set in place and maintain?
25
Information needed for an ex ante assessment
related to seasonal climate forecasts
Stage 1. Change (e.g. implementation)
How to obtain
Level of
uncertainty
Mod
Resources required
Time
Partnerships and skills
Probability of different levels of
success
26
Scoring methods
Peer review
Mod
High
High
Information needed for an ex ante assessment
related to seasonal climate forecasts
Stage 2. Uptake
How to obtain
Level of
uncertainty
27
Who, characteristics
GIS, surveys
Mod
Where, characteristics
GIS, surveys
Mod
Infrastructure needed
GIS
High
Policies needed
Surveys
High
Adoption rate, ceiling
Scoring methods
High
Costs involved
Scoring methods
High
Information needed for an ex ante assessment
related to seasonal climate forecasts
Stage 3. Impact quantification
How to obtain
Level of
uncertainty
28
Production
Biophysical models
Mod
Income, risk and food security
Household models
Mod-High
Changes in vulnerability
Models, scoring?
High
Changes in adaptive capacity
Models, scoring?
High
Capacity building
Scoring methods
High
Costs, prices, elasticities
Lit review, surveys
Mod-High
Future developments to help overcome the challenges
1 Understanding better who the potential clients are,
and what characterises them
• Partly a question of spatial info (poverty maps, new
continental/global data layers, etc)
• But also a question of information on non-spatial
determinants of poverty and vulnerability, how decision
makers actually make decisions, information flows and
power structures in communities, etc
29
Future developments to help overcome the challenges
2 Developing tools that are better able to cope with
the demands of climate forecast assessment
May need new or adapted behavioural frameworks, beyond
profit or utility maximisation, to take account of impacts on
• food security
• reduction of household vulnerability
• increases in household adaptive capacity
Different types of models may help: agent based, systems
dynamics
30
Future developments to help overcome the challenges
3 Developing approaches that combine quantitative
and qualitative elements
Linked also to provision of baseline data, for monitoring and
evaluation
That can then be linked to ex post impact assessments, so that
the lessons learned from this whole process can be applied
elsewhere in the pursuit of poverty alleviation goals
31
Future developments to help overcome the challenges
4 Making the process of impact assessment
participatory
The process is often as important as (if not more important than)
the results of the analysis
Getting all stakeholders involved in thinking broadly about the
problems involved and the potential impacts
32
Thank you
[email protected]