Will agricultural intensification save tropical forests? Arild Angelsen School of Economics and Business, Norwegian University of Life Sciences (UMB), Ås , Norway & CIFOR , Bogor, Indonesia [email protected] Warzaw 12.11.2013
Nov 22, 2014
Will agricultural intensification save tropical forests?
Arild Angelsen School of Economics and Business,
Norwegian University of Life Sciences (UMB), Ås , Norway & CIFOR , Bogor, Indonesia
Warzaw 12.11.2013
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Three major roles of agriculture in climate change mitigation
1. Agric land encroaching forests (defor & A/R)
2. Fluxes on (existing) agric land
3. Substitution effects from changes in agric production (e.g.
biofuel replacing fossil fue: l)
Focus only on 1., and ask one main question
Can/will agric intensification save forests?
Land consuming vs. land sparing (Jevons vs. Borlaug)
(A related: land sharing vs. land sparing)
Agric intensification (increase output/ha = yield):
– Technological progress (more outputs with same inputs)
– Factor substitution (more inputs per ha)
2
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
The Borlaug world (the full belly or subsistence model)
Food = Food
Food/pop * pop = Food/ag land * ag land
Subs req. * pop = yield * ag land
Ag land = (subs req * pop) / yield
Land = ag land + forests
A simple theory of deforestation
Ag intensification (higher yield) reduce need for ag land
=> less encroachment into natural forests
Can be apllied at various scales (e.g. global food
equation)
3
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Example: cereals
1961-63 2006-08 Pct . increase
Demand
Pop (bn) 3.13 6.62 111.6
Consumption (kg per capita)
294.3 358.3 21.8
Supply
Area harvested (mill ha)
653.7 697.2 6.7
Yield 1.41 3.40 141.5
4
Source: Stevenson et al. (2011),
based on: http://data.un.org/
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
The Jevon world: (A market/partial equilibrium/von Thünen model)
Define ag rent per ha as
profit = gross sales – costs
r = py – wl – qk –vd
p(rice), y(ield), w(age), l(abour) per ha, q(cost of k),
k(capital), v(distance costs per km & ha), d(istance)
5
Deforestation (d)
Ag rent
Sch
ool o
f Eco
nom
ics and B
usin
ess
Policy: reduce ag rent:
Lower yield will save
the forest!
$
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Can the two worlds be reconciled?
Two very different logics
– Subsistence model and global food equation:
higher yield saves forests
– von Thünen model:
higher yield gives encroachments into forests
How can they be reconciled?
1. Extending global food equation
2. Market demand conditions
6
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
National deforestation equation (NDE) (Angelsen, 2010)
Pop * (Food cons/Pop) = (Food cons/Food prod) * (Food
prod/Ag prod) * (Ag prod/Ag land) * (Ag land/Forest) *
Forest
deforestation ≈ pop growth + ∆ food cons per capita - ∆
self-sufficiency ratio (inverse) - ∆food share - ∆yield -
∆ag/forest ratio
One among several factors
Yield change can affect other factors:
Self sufficiency (more competitive)
Food share, e.g. biofuel
Be careful with identities:
they are always correct (a warning sign!)
cannot assume ceteris paribus (other factors will change)
7
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
It’s the demand elasticity, stupid!
Demand elasticity: how sensitive is demand to price
changes (1% ∆ price => x% ∆ quantity)
What is the impact of technological change (supply shift)?
Inelastic (quantity given – 1. Borlaug world: B) vs. Elastic (price given – 2. Jevons/von Thünen world: C)
8
Perfectly
inelastic
demand
Perfectly
elastic
demand
Supply
Supply after
tech progress
A=B C
Price
Quantity
B
A=C
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
What is the demand elasticity?
Depends on:
How widespread is the tech change; market share
Scale of analysis:
– The higher scale, the more inelastic demand
Type of commodity:
– Inelastic: food
– Elastic: non-food with substitution (e.g. rubber, biofuel)
9
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Empirical studies – macro level
Area & yield links at national level over time (by crop or
total)
Ewers et al. (2009):
– 23 staple crops, 1979-1999, 124 countries
– The yield-area elasticity
– Borlaug hypotheses: -1
– Developing countries: -0.152 (t=-1.78)
– Developed countries: -0.089 (t=-0.57)
– Weaker and non-significant for total cropland
– Weak tendency in developing countries for the per
capita area to decline as cropland increase
10
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
… empirical studies
11
filled: developing; countries
open: developed countries
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Case studies summary
12
Reduced (win-win)
Impact on deforestation
Increased (win-lose)
Intensive (high) L & K intensity Saving (low)
Constrained Farmer characteristics
Well-off
Local Output market Global
Yield increasing Technology Cost-saving
Local, segmented
Labour market Mobile (migration)
Intensive (lowland)
Sector experiencing tech. change
Frontiers (upland)
Global Scale of adaptation Local
Short term Time horizon Long term
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Win-win outcomes
Agricultural technologies suited specifically for forest-
poor areas
Labour-intensive technologies where labour is scarce
and migration limited
Promote intensive systems where farmers are also
involved in low-yielding extensive farming practices
Agricultural technologies that substantially raise the
aggregate supply of products with inelastic demand
BUT, some of win-win technologies are least likely to be
adopted by farmers
– Produce commodities for local markets where prices
quickly drop
– Choose technologies that use the most scarce resources
intensively 13
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Win-lose outcomes
Agricultural technologies that encourage production
systems that require little labour and/or displace labour
New agricultural products for sale in large markets in
labour-abundant contexts
Eradication of diseases that limit agricultural expansion
Technological changes in forest margin areas with
rapidly growing labour forces
14
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Some trends
1. Globalization – increased
market integration; more likely
to be price takers
2. Deforestation driven by
commercial actors
3. Separation of forest and agric
land
15
Sch
ool o
f Eco
nom
ics and B
usin
ess
land sparing
less likely
agric int.
less important
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
So what?
Mistake 1: assume that technological change & agric
intensification will save forests
Mistake 2: be against new technologies & intensification
because it may put pressure on forests
Agric intensification needed for a number of reasons, but
forest conservation is not on top of that list
BUT, will enable and make other forest conservation
measures more effective and politically feasible
It’s not the solution, but part of the package
16
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Deforestation policies – what works
Selective agric technologies
Agric rent in frontier areas
– Roads
– Subsidies
Forest rent and its capture
– Community management
– PES schemes
Regulations
– Protected areas (enforcement)
– Land use planning
17
Sch
ool o
f Eco
nom
ics and B
usin
ess
NO
RW
EG
IAN
UN
IVERSIT
Y O
F L
IFE S
CIE
NCES
www.umb.no
Will agricultural intensification save tropical forests?
“… no one can guarantee that
economic development –
whether agriculturally driven
or not – will lead to a forest
transition and an end to
inappropriate deforestation.
Informed proactive policies will
have to do that.”
18
Sch
ool o
f Eco
nom
ics and B
usin
ess