Transcript Impacts_week9_post

Economics of Climate Change:
Impacts
Economics 331b
Spring 2010
1
Agenda
This week: Impacts
Week 10: Further impacts and discounting
Week 11: Mitigation
Week 12: Alternative policies
Week 13: International agreements
2
Price of carbon emissions
The basic analytical structure
Marginal Damages
Marginal Cost
Pcarbon*
0
Market!
Abatement*
Abatement
4
Where does the marginal damage function come from?
1. Consider output as function of climate:
Qt = f(Kt , Lt , At ; Tt )
This would be appropriate sum over sectors, individuals, countries,
and time periods.
2. Then get the damage function as function of T relative to base T:
Dt = f(Kt , Lt , At ; Tt ) - f(Kt , Lt , At ; T0 )
3. We then relate temperature to past emissions:
Tt = g( E0 , E1 , E2 , … , Et )
4. From which we get the marginal damage function. Here, the
marginal damage (sometimes called social cost of carbon) is:

SCC t = MD t = [PV (D t )] / E t = [

v=t
D ve
rv
] / E t
5
Impacts Analysis
Central task is to evaluation the impact of climate change on
society
Two major areas:
–
–
market economy (agriculture, manufacturing, housing, …)
non-market sectors
•human (health, recreation, …)
•non-human (ecosystems, species, fish, trees, …)
6
Basics of Impact Analysis
1. Start with a production function:
Q j,t = F(K j,t ; W j,t , T j,t), t in future
where Q j,t = output in sector j at time t; K j,t = capital and other
conventional inputs; W j,t = weather (realization); T j,t = climate
7
What is climate?
Consider the complex system as a stochastic process:
dx(t)/dt = h[x(t); α, ρ, …]
x(t) is temperature, precipitation, ocean currents, etc. α, ρ,
etc. are parameters.
Weather is the realization of this process.
Climate is the statistics of the process (mean, higher
moments, extremes). It is usually calculated as moving
averages (e.g., 30-year “normals”).
8
Basics of Impact Analysis
1. Start with a production function:
Q j,t = F(K j,t ; W j,t , T j,t), t in future
where Q j,t = output in sector j at time t; K j,t = capital and other
conventional inputs; W j,t = weather (realization); T j,t = climate
2. We often have data on the impact of weather changes , ∂Q j,t /∂Wj,t.
3. But, we need to understand climate impacts, ∂Q j,t /∂ T j,t
4. Moreover, we must consider a vector of important climatic variables
(temperature, precipitation, soil moisture, snow pack, …). We will
need to translate global mean temperature change into the relevant
climatic variables, which is non-trivial given regional resolution of
climate models.
5. This also requires estimating the production function in the distant
future, at which time the impacts will occur.
6. Finally, we need to discount back the impacts to the present.
9
Market sectors by vulnerability to climate change, US, 2007
Billions of dollars
Gross domestic product
13,808
Major Potential Impact
Farms
Forestry, fishing, and related activities
Nordhaus, based on BEA , industry accounts,
http://www.bea.gov/industry/index.htm#annual
168
1.2
598
4.3
13,041
94.4
11
281
159
121
27
Small to Negligible Impact
Mining
Construction
Manufacturing
Trade
Retail trade
Transportation and warehousing
Information
Finance, insurance, real estate, etc.
Services and government
100.0
137
31
Moderate Potential Impact
Water transportation
Utilities
Real estate
Coastal
Accommodation
Arts, entertainment, and recreation
Outdoor
Percent
of GDP
275
611
1,617
1,698
893
407
586
2,652
4,302
10
Example from Agriculture
Long history of agricultural production functions in which
weather is a variable. Remember:
Q j,t = F(Kj,t ; W j,t )
This produced first set of estimates of impact of global
warming; led to very large estimates of losses.
11
Example of impact of weather on yield
Why is this
completely
wrong for
understanding
the impact of
climate change
on agriculture?
12
Example from Agriculture
Long history of agricultural production functions in which
weather is a variable.
- This produced first set of estimates of impact of global
warming; led to very large estimates of losses.
Problem: The temperature-output relationship does not
take into account adaptation of farmers to climate.
This is the “dumb farmer” v. “smart farmer” controversy.
Ricardian methods are attempt to look at equilibrium effect
of climate by looking at cross-sectional impact of climate
on farm values (Mendelsohn key figure here)
- This produced much smaller estimates because of farmer
adaptation.
13
Short-run
v.
long-run productivity
14
Estimates of Impacts on Agriculture late in the 21st C
Impacts on net value of agriculture as percent of national or
global income:
Mendelsohn
Cline
North American
+ 0.4 %
+ 0.5 %
Africa
- 5.0 %
- 4.0 %
Global average
- 0.2 %
- 0.1 to -.05%
Estimated effect of ag on output is small because (1)
agriculture is small, (2) farmers can adapt, (3) CO2 is a
fertilizer.
Query: Assume this is for 2075 for constant share of ag. What if
share declines by half?
Source: Mendelsohn et al.; Cline
15
The tricky issue of declining share of agriculture
50
45
High income
East Asia
Latin America
South Asia
Sub-Saharan Africa
Share of agriculture in GDP (%)
40
35
30
25
20
15
10
5
0
1965
1970
1975
1980
1985
1990
1995
2000
2005
16
Global Warming and Sea Level Rise (SLR)
Major variations in geological history (-150 to +40 meters)
Sources in future:
- Thermal expansion (up to 2 meters in next 500 years)
- Small glaciers (0.5 meters)
- Greenland (up to 6 meters)
- Antarctic (56 meters), but major unstable is West
Antarctic Ice Sheet (7 meters)
- Arctic Sheet (very likely to disappear, 0 meters)
Major issues are stability and irreversibility
17
Observed Sea Level Rise
18 cm rise since 1900
Current rate:
3.3 cm per decade
Rahmstorf, Cazenave, Church, Hansen, Keeling, Parker and Somerville (Science 2007)
18
WBGU
Data:
Church and White (2006)
Scenarios 2100:
50 – 140 cm (Rahmstorf 2007)
55 – 110 cm (“high end”, Delta Committee 2008)
Scenarios 2200:
150 – 350 cm (“high end”, Delta Committee 2008)
Scenarios 2300:
250 – 510 cm (German Advisory Council on
Global Change, WBGU, 2006)
Delta Comm.
Recent Global Sea Level Rise Estimates
19
Rahmstorf at http://www.ozean-klima.de/
How to calculate SLR from thermal expansion?
1. Generate temperature profile over time and
ocean latitude, longitude, and depth
2. Look at physical relationship between
temperature and volume
3. Grind out the numbers
20
Density (inverse volume) as function of
temperature and salinity
21
Temperature/depth profile
22
Estimates of Thermal Expansion for IPCC AR4
Source: AR4, Chapter 10, p. 829.
23
Estimate of distribution of output and population by elevation
Cumulative distribution of global
population and output by year
.05
Output, 1990
Output, 1995
Output, 2000
Population, 1990
Population, 1995
Population, 2000
.04
.03
.02
.01
.00
0
2
4
6
8
10
Gecon.yale.edu
Elevation (meters)
24
Impact of Sea-Level Rise
On Value of Real Estate in Cape Cod
Caroleen Verly, “Sea-Level Rise on Cape Cod: Predicting the Cost of Land and Structure
Losses”, Yale University, 2008
25
Land value and altitude
Caroleen Verly, “Sea-Level Rise on Cape Cod: Predicting the Cost of Land and Structure
Losses”, Yale University, 2008
26
Land value
($/acre)
Land value as area
under value curve
Distance from year
2000 coastline
27
Loss with zero
adaptation or foresight
Land value
($/acre)
Retreat
Distance from year
2000 coastline
28
Land value
($/acre)
Loss with adaptation
and foresight
Retreat
Distance from year
2000 coastline
29
Impact of SLR on Land and Structures
Caroleen Verly, “Sea-Level Rise on Cape Cod: Predicting the Cost of Land and Structure
Losses”, Yale University, 2008
30
Central message
For adaptive systems, need to consider adaptations to
climate change (farmers, skiers, swimmers, trees, coral
reefs, …)
But must consider the costs of adaptation as one of the costs
of climate change (costs of moving, snowmaking
machinery, new crops, …)
31