Embed Size (px)
Citation preview
Economics and Climate Change
Michael HanemannDept. Agricultural & Resource
EconomicsUniversity of California, Berkeley
2
How does economics come in?
• Economic cost of reducing GHG emissions – impact on economy.
• Economic value of damages from climate change – benefits from reducing GHG emissions– Market impacts: value placed on items that people
obtain by purchasing them (i.e., through a market)
– Non-market impacts: value placed on items that people care about but do not obtain through the market
3
Climate change is an externality
• Climate change involves what economists call an externality, wherein a person undertakes an action that imposes an adverse effect on some other people.
• In the presence of externality, competitive markets are unlikely to generate an outcome that is in the public interest.
• Reliance on a purely voluntary approach to deal with externality is unlikely to be satisfactory or useful.
4
Taxonomy of economic impactsTYPES OF ECONOMIC IMPACTS
MARKET IMPACTS
A) Climate is an "input" to production
AGRICULTUREFORESTRYFISHERIESWATER SUPPLYENERGY
B) Disruptions caused by climate
Disruption of employment/production due to extreme eventsDamage to property due to extreme eventsResponse/protetction costs to deal with extreme events
NON-MARKET IMPACTS
HUMAN HEALTH - HEAT WAVESHUMAN HEALTH - VECTOR BORNE DISEASESECOSYSTEM LOSSESLANDSCAPEAMENITY/LIFESTYLE
5
Agriculture & Water
• Agriculture, along with forestry, is the sector of the economy that is most directly affected by climate.
• Water supply is the single most important pathway by which most other sectors of the California economy will be affected.
.
6
The Climate Impact Studies in California
• Hayhoe et al., Proceedings of the National Academy of Sciences, August 24, 2004.
• Governor’s Report on climate change http://www.climatechange.ca.gov/climate_action_team/reports/index.html
7
Sectoral studies
• Agriculture
• Water
• Commercial forestry
• Fire
• Coastal resources
• Public health
8
Broad, varied approach
• Work from common set of scenarios• In each sector review what is known, what can
be said, qualitatively or quantitatively, about impact of scenarios, using case studies where appropriate.
• Identify, also, what is not known and what needs to be done to provide a more complete assessment in the future.
9
Approach continued
• Develop physical impacts• Then trace this down to social and
economic impacts• Besides impacts, identify opportunities for
adaptation policy• Quantify economic impacts where possible• Carry impacts through to rest of California
economy
10
Some general observations
• There is necessarily some interaction between the economic analysis we perform and considerations of policy analysis.– When you ask a question such as “How much fossil
fuel will be burned by the x industry” the answer depends partly on the constraints on firms in the industry, and partly on the incentives they face. Thus, the question cannot be answered completely without specifying incentives (i.e., policies),
– Therefore impact analysis and mitigation analysis must come together.
11
The new impact (“Scenario”) analyses
The new analyses contrast two global emission scenarios: (i) business as usual with continuing high rate of growth in greenhouse gas (GHG) emissions; and (ii) sustained global effort to reduce GHG emissions by 2100.
The analyses use statistical downscaling to translate the larger scale climate model results to California with finer resolution.
12
How are the new versions of these climate models different?
• The new versions of both models are somewhat more pessimistic with regard to precipitation than the previous two versions.
• The new versions are substantially more pessimistic with regard to the increase in summer-time temperature.
• The latter result may be due to improved modeling of the linkage between surface temperature and ambient air temperature.
13
IPCC Emissions ScenariosCO2 Concentrations
Higher (A1fi)
Lower (B1)
Previous(IS92a)
14
Rising TemperaturesCalifornia statewide
Projected average summer temperature changes
15
Time Lags
• The temperature trajectories for the two emission scenarios remain fairly intertwined until about 2045.
• They illustrate the fact that California’s future climate for next 40 years is already determined by past emissions.
• Emission reductions initiated now show a significant effect after about 2045, with their impact increasing over time.
16
Patterns of Temperature Change
2070-2099 relative to 1961-1990Summer
HadCM3 higherHadCM3 lower PCM higherPCM lower
LOWER EMISSIONS HIGHER EMISSIONS
14.4
18.0 °F
10.8
7.2
3.6
0.0
-3.6
17
Diminishing Sierra Snowpack% Remaining, Relative to 1961-1990
18
Impacts assessed for California
• Impacts on flooding• Impacts on agricultural & urban water use• Impacts on agriculture• Impacts on forestry and on fire• Impact on human health• Impacts on vegetation• Impacts on energy demand and hydropower
supply• Impact of sea-level rise
19
Assessment approach
• Based on spatially disaggregated downscaling of climate change scenarios (monthly temperature and precipitation) to 2100.
• Uses case studies and bottom-up analyses.• Focuses on what is known now; identifies gaps
which will require future research. • Tries to push through to some of the economic
implications of physical changes.• A work in progress. Will be special issue of
Climatic Change. But, much more needs to be done.
20
• Key innovation in Scenario Project was detailed spatial downscaling.
• Key Lesson from this: BEWARE AVERAGING• Nonlinearities & thresholds occur at quite fine
scale of spatial and temporal resolution. Using broad spatial and temporal averages is highly misleading.
• This matters greatly both for measuring economic impacts and for designing adaptation policy.
21
An example: what is the increase in temperature?
HOW TO CHARACTERIZE THE CHANGE IN TEMPERATURE, 2070-2099, USING HADCM3
EMISSION SCENARIO**A1fi B1
Change in global average annual temperature 4.1 2
Change in statewide average annual temperature in California* 5.8 3.3Change in statewide average winter temperature in California* 4 2.3Change in statewide average summer temperature in California* 8.3 4.6Change in LA/Sacramento average summer temperature ~10 ~5Extra # of extreme heat days in Los Angeles/Sacramento 100 30
*Change relative to 1990-1999. Units are ˚C
22
Why variation matters
• The nonlinearity of the damage function means that the aggregate damage is larger than if one simply assesses the damage corresponding to the average temperature change.
• Fenchel’s inequality: if D(x) is concave,
E{ D(x) } > D( E{x} )
23
Schlenker & Roberts (2006) Relation of Temperature and Crop Yield
• Relationship is not symmetrical; it is distinctly asymmetric, fairly flat at first and then sharply declining beyond an upper threshold.
24
Non-linear relationship between temperature and yields and Fenchel’s inequality
0.04
0.02Log Yield
0 a
-0.02
-0.04
-0.0610 15 25 30 35 40
Temperature Celsius
25
Climate, water & agriculture
• The standard view in the literature on the economic impact of climate change on US agriculture is that (1) precipitation is the key variable to focus on, rather than temperature, and (2) “wetter is better.”
• Disagree on both counts. Spatial and temporal considerations lead to a different perspective.
26
Importance of spatial and temporal details for water supply
• 2/3 of precipitation occurs north of Sacramento.
• About 2/3 of all water use occurs south of Sacramento.
• 80% of precipitation occurs October-March.
• 75% of all water use occurs April – September.
• Snow pack holds the equivalent of ~1/3 of our major storage capacity
27
• In California, changes in winter precipitation are
far less significant economically than changes in temperature.– Water is not a scarce resource in the winter.– To make winter precipitation an economically valuable
asset requires an investment in some form of storage. This becomes cost of climate change.
– Unlike precipitation, changes in winter temperature directly affect spring and summer water supply.
• Economically, it is the change in temperature that is especially significant for California.
28
Institutional variation
• Water supply in CA is managed by ~300 individual agricultural and urban water agencies.
• They are quite distinctive. They have different sources of supply, different water rights, different economic needs, and they are independently managed.
• Given this heterogeneity, a top-down (averaged) approach to impact analysis produces different implications than a bottom-up (disaggregated) approach.
29
Research approach
• Water district is key unit of observation
• Focus on reliability & economic value of changes in reliability
• Combines hydrology and economics
• Measure existing (baseline) reliability
• Model change in deliveries & reliability
• Assess loss of income & consumers surplus due to change.
30
Tail probability events & the need to look at the full distribution
• Damages increase non-linearly with severity of supply reduction
• Adverse impacts of climate change are likely to be disproportionately larger in the worst years
• Hence, the worst years get worse.
31
CVP South of Delta Annual Deliveries under climate change scenarios PCM B1-A2 and GFDL B1-A2 for 2070-2099
0
500
1000
1500
2000
2500
3000
3500
0%10%20%30%40%50%60%70%80%90%100%
Probability of Exceedance (%)
Del
iver
y (T
AF
)
BASE GFDLA2 PCMA2 GFDLB1 PCMB1
75% Exc. (TAF)
BASE 2153GFDLA2 1414PCMA2 1858
GFDLB1 1723PCMB1 1997
10% Exc. (TAF)BASE 3332GFDLA2 2637PCMA2 3331GFDLB1 3297PCMB1 3332
25% Exc. (TAF)BASE 3257GFDLA2 2378PCMA2 3190GFDLB1 2845PCMB1 3281
50% Exc. (TAF)BASE 2833GFDLA2 1944PCMA2 2424GFDLB1 2261PCMB1 2618
32
TABLE 1. CHANGE IN SURFACE WATER AVAILABLE TO AGRICULTURALWATER USERS IN THE CENTRAL VALLEY, 2070-2099
REGIONSacramento San Joaquin Tulare LakeValley Basin Basin
APPLIED WATER USE IN 2000 (TAF) 7,735 7,358 10,879 % Surface water 64% 74% 60% % Groundwater 36% 26% 40%
CHANGE IN SURFACE WATER AVAILABILITY 2070-2099 (Surface water with climate change/surface water used in 2000)
YEAR CATEGORY
Upper 50% of years 98% 91% 90%Next 35% of years 90% 52% 51%Lowest 15% of years 47% 33% 30%
Average of all years 88% 69% 67%
33
Economic impacts under GFDL A2 – water supply
• By 2085, in an average year, 9% loss of net revenue in Central Valley agriculture; 26% reduction in lowest 15% of years.
• By 2085, urban shortages in Southern California occur twice as frequently and are about twice as severe; in about 35% of the years, rationing could cause loss averaging $5 + billion/yr for water users.
34
Common pattern
• There may not be a large change on average, but….
• Bad years occur more frequently
• Economic harm in a bad year becomes more severe.
• Raises issue of risk aversion & insurance
35
Insurance: an additional component of the economic cost
• With perfect foresight, the cost of adaptation can be minimized. E.g. purchase water from water markets only for months where there is a supply shortfall; purchase no more than the amount of the shortfall.
• With uncertainty (imperfect foresight), adaptation will be more costly.
• With risk aversion, adaptation will become even more costly adaptation. Water users will want to buy the equivalent of insurance.
• With insurance, costs are incurred in years where they turn out not to be needed.
36
Extreme events
• With extreme events (heat waves, floods, coastal storms) the consequences spill over to the larger economy, not just climate-sensitive sectors (agriculture, forestry, water, energy).
– There is property damage
– There is disruption of normal production
37
Implications for water policy
• A major reason impacts on water supply infrastructure is problematical is that modifying it or developing new infrastructure is tremendously expensive.
• Also, the cost structure of infrastructure is hard to deal with:– Immensely capital intensive– Very long-lived capital– Join products flow from it
38
The cost structure makes financing very difficult
• If the costs were mainly operating cost it would be very simple to rely on “the user pay principle” and pay-as-you-go financing.
• As it is, the capital intensity and capital longevity mean make it very hard to have “user pay.”
• One ends up relying heavily on transfers – between one group of users and on other, or between one generation and another.
39
CONCLUSIONS• Existing estimates are likely to significantly
understate economic cost of climate change impacts in California.
• They focus on equilibrium; they downplay costs of adjustment and adaptation.
• They ignore impacts on capital assets.
• They ignore costs associated with uncertainty and risk aversion.
• They ignore non-market impacts.
40
For more information
• http://calclimate.berkeley.edu
• http://www.climatechange.ca.gov
