The role of innovation in climate
change mitigation: new
perspectives using the WITCH
model
V.Bosetti*, C.Carraro*, R.Duval**, M.Tavoni*
* FEEM, ** OECD Economics Department
1
Motivation for the paper
• Issue of role of innovation support in international climate
policy package has gained prominence:
- Energy R&D spending is low and has declined
- Innovation failures (R&D and learning spillovers) and
environmental externalities call for joint policy intervention
in both areas
- International cooperation may be easier to achieve on R&D
than on mitigation (recent game-theoretic literature)
- And indeed technology (although not so much innovation)
issue at the core on ongoing climate negotiations
2
Purpose of the paper
• Little analysis of environmental and economic gains from
innovation in GE framework so far
• Paper contributes to this literature by providing
quantitative assessment of extent to which:
1. Stand-alone
innovation
policies
emissions/stabilise concentration
could
reduce
2. Innovation policies, when used as a complement to a
carbon price, could lower cost of climate stabilisation
3
The WITCH model
• Hybrid IAM developed by climate change group at FEEM:
- Forward-looking optimal (Ramsey-type) growth model
- Climate module and feedback through damage function
- Specific representation of energy sector: 6 fuel types, 7
technologies for electricity (+1 advanced tech, see below)
• Game-theoretic framework:
- 12 regions set optimal paths of investments (goods, energy,
R&D) to maximise PV of (log) consumption
- 2 solutions concepts: fully cooperative solution (world
social
planner),
equilibrium)
non-cooperative
solution
(Nash
4
Endogenous technological change in WITCH
• 3 main channels / 3 main innovation policy areas:
- Energy efficiency (E.E) improvements through R&D
investments (LbR with high but decreasing social returns)
- LbR and Learning-by-doing (LbD) in wind, solar and CCS
(W+S & CCS) via 2-factor learning curves
- LbR and LbD in 2 unspecified “advanced” techs
(substitutes for nuclear in EL and oil in NEL): calibrated
on past experience with other techs, arbitrarily high initial
price, 10-year commercialisation lag
• International knowledge and technology spillovers:
- R&D and/or technology deployment in 1 country increase
5
R&D and/or deployment returns elsewhere
1. Innovation as a stand-alone policy
• We consider a global R&D fund amounting to 0.08% of
world GDP (~ early 1980s level > 4 times current levels)
• 3 possible allocations : E.E, W&S + CCS, advanced techs
• Main finding: no such Fund can reduce emissions from
current levels  continuous rise in concentrations ~ BAU
• Advanced techs policy misses medium-term abatement
opportunities but is still >> E.E and W&S+CCS:
- E.E: decreasing returns to R&D, no decarbonisation
- W&S + CCS: decreasing returns to R&D, all the more so
with limits to price declines and storage constraints
6
1. Innovation as a stand-alone policy
(continued)
• Sensitivity analysis:
- Varying the size of the fund (up to 2% of world GDP), no
fund of any size succeeds in reducing emissions from
current levels
This reflects decreasing marginal returns to R&D in long
run, and positive impact of R&D on GDP & emissions
- Varying the allocation of the fund across 3 types of R&D,
no significant improvement is achieved relative to the
advanced techs case
- Overall, maximum achievable emission reduction relative
to BAU ~ 15% at 2100 horizon
7
2. Carbon pricing and endogenous
technological change
• Innovation policies are environmentally ineffective but
may still bring eco efficiency gains  assess these gains
• 2 scenarios : BAU (non-cooperative solution) and 450 ppm
CO2 (non-cooperative solution under world emission
constraint)
• Carbon pricing alone has large effects on energy R&D and
technology deployment…
• …which in turn has a sizeable impact on mitigation costs,
with option to invest in advanced techs being crucial
(nearly halves discounted costs)
• E.E and W&S+CCS matter much less than advanced techs,
consistent with analysis of stand-alone policies
8
2. Carbon pricing and endogenous
technological change (continued)
• The NEL advanced tech is more important than the NEL
one for lowering mitigation costs:
- Higher abatement costs in NEL  larger gains from R&D
- The EL advanced tech would matter more if existing
options (W&S, nuclear, CCS) were constrained
• Advanced techs do not come for free:
- Strong carbon price signal needed in the first decades to
provide incentives to invest in advanced techs
- Global mitigation costs actually rise in the first decades,
reflecting diversion of resources towards R&D
9
2. Efficiency gains from hybrid carbon
pricing/innovation policies
• International R&D spillovers suggest potential gains
(international innovation policies yield an economic gain)
• We consider again a global R&D policy amounting to
0.08% of world GDP with the 3 possible allocations
• Results: discounted cost savings relative to carbon pricing
alone = between 1% (E.E) and 8% (advanced techs) under
450 ppm CO2 scenario
• Cost savings are larger under less stringent targets, e.g.
15% for advanced techs policy under 550 ppm CO2
scenario
10
Conclusions
• Innovation policies cannot be effective if used as standalone policies (lags, decreasing returns to R&D)…
• …but can lower the cost of a climate policy package,
especially if focused on NEL advanced technologies
• The magnitude of this economic efficiency gain ultimately
depends on the magnitude of R&D spillovers not already
internalised by carbon price
• Magnitude of international R&D spillovers is subject to
debate + not all such R&D spillovers are captured in
WITCH (e.g. apropriability problems)
 room for future work
11
Thank you!
12
Figures
13
Public energy R&D expenditures in OECD countries
Energy R&D
(Percentages)
0.14
14.0
0.12
12.0
0.10
10.0
0.08
8.0
0.06
6.0
As a % of GDP (Left scale)
0.04
4.0
As a % of total R&D (Right scale)
14
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1976
1975
2.0
1974
0.02
Carbon emission paths under alternative allocations of the global R&D fund
25
Gton C
20
15
BAU
E.E.
W+S & CCS
10
Advanced Techs
5
450
0
15
Carbon emission paths under alternative sizes of the global R&D fund
(allocated to all 3 types of R&D simultaneously)
25
Gton C
20
15
BAU
0.2% GDP fund
10
2% GDP fund
5
0
16
Projected energy R&D investments under 450ppm CO2 concentration stabilisation
scenario, with and without backstop technologies
% of GDP
0.14
BAU
450
0.12
450 with backstops
0.10
0.08
0.06
0.04
0.02
0.00
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
17
% change in GDP with respect to baseline
Projected world GDP costs under 450ppm CO2 concentration stabilisation scenario,
with and without backstop technologies
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
450 Backstop
450
-4.5
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
18
% change in GDP with respect to baseline
Projected world GDP costs under 450ppm CO2 concentration stabilisation scenarios,
with and without induced technological change
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
450
-4.0
450 no ITC
-4.5
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
19
$/tCO2e
Projected CO2 price levels under 450ppm CO2 concentration stabilisation scenario,
with and without backstop technologies
400
Price of carbon permits
350
450 Backstop
300
450
250
200
150
100
50
0
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
20
% change in GDP with respect to baseline
Projected world GDP costs under 450ppm CO2 concentration stabilisation scenario,
with electricity backstop or non-electricity backstop only
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
450 Backstop
-4.0
450 Backstop, Electricity only
-4.5
450 Backstop, Non-electricity only
-5.0
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
21
% change in GDP with respect to baseline
Projected world GDP costs under 450ppm CO2 concentration stabilisation scenarios,
with and without constraint on nuclear energy and CCS
0.0
-1.0
-2.0
-3.0
-4.0
-5.0
-6.0
-7.0
450
450 with constraints
-8.0
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
22
Cost savings from hybrid innovation/carbon pricing policies
(% GDP gain relative to carbon pricing alone, 450 ppm CO2 scenario )
20%
15%
10%
5%
Advanced Techs
0%
-5%
2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
W+S & CCS R&D
E.E. R&D
-10%
-15%
-20%
-25%
23
The Energy Sector
OUTPUT
1/2
KL
1
ES
0.4
K
HE
L
EN
1/2
NEL

EL
TradBiom

ELHYDRO
OGB
1/2
COALnel
EL2
OILnel
2
Biofuel

ELFF
ELW&S
2
GASnel
Trad
Biofuel
ELNUKE
ELCOAL

ELPC
ELIGCC
ELOIL
ELGAS
24
Advanced
Biofuel
Projected energy technology mix in the electricity sector under baseline, 550ppm and 450ppm
CO2 concentration stabilisation scenarios
Terawatt hour
Baseline
45000
40000
35000
30000
W&S
25000
NUCLEAR
COAL with CCS
20000
COAL without CCS
15000
GAS
10000
OIL
5000
0
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
Terawatt hour
Scenario 450ppm
45000
40000
35000
ENERGY REDUCTION
30000
W&S
25000
NUCLEAR
20000
COAL with CCS
COAL without CCS
15000
GAS
10000
OIL
5000
25
0
2007
2012
2017
2022
2027
2032
2037
2042
2047
2052
26