Three obstacles to adding up

1. How to compare monetary expenditures on policy with non-

monetized damages to human health and the environment?
(valuation)
2. How to compare net benefits this year with net benefits 10 years
from now? (time)
3. How to incorporate uncertainty?
Group project
• Suppose we all believed
the following:
– There is a 90% chance the
consequences of our current
trajectory of CO2 emissions will, at most,
be moderate and manageable
– But there is a 10% chance the consequences
will be severe, including sea level rise and
extended draughts that lead to the forced
dislocation of 100’s of millions of people

How should policymakers think about this decision problem?

Risk
• Often for environmental problems, we don’t know
what is going to happen, even if we do have a
directional sense of what plausible outcomes may be
• In fact, we face the same kind of uncertainty when
making personal decisions such as where to go to
college, what to major in, who to marry, etc.
• For scientific problems, information is often available
in the form of probabilities
• (similar to the probabilities one might have when gambling
in Vegas)
Example
• Based on past experience, scientists predict that
if a certain chemical is released into the FC water
supply, the risks are described in the following
table:

Number of deaths Probability

0 0.8
1 0.14
2 0.05

• How should we use this type of information?

3 0.01

4 0
Expected value probability/possibility math problem
number x probability = ...
• Gives a kind of average
• It is the quantity that a “risk-neutral” decision
maker would typically accept
• Like a casino or an insurance company

Number of deaths Probability Expected Value of Deaths

0 0.8 0 x .80 = 0

1 0.14 1 x .14 = .14

2 0.05 2 x .05 = .1

3 0.01 3 x .01 = .03

4 0 4x0=0

Expected value = 0.27

Are we risk neutral?
i.e., do we or should behave like
casinos?
• Consider a “fair” bet: I’ll flip a coin and if it is heads,
you give me $20, if tails I give you $20
• It is “fair” because the expected value is zero
• Will you take the bet?
• Typically individuals are risk averse
• When the environment is involved, people may be
very risk averse
• This means we would value a risk more than the
expected value
Three obstacles to adding up

1. How to compare monetary expenditures on policy with non-

monetized damages to human health and the environment?
(valuation)
2. How to compare net benefits this year with net benefits 10 years
from now? (time)
3. How to incorporate uncertainty?
Valuation: the information we want is
captured in a demand curve
• Why force the things we value in the environment
into “consumer theory” (the branch of economics
that models the demand side of an economy)?
• A demand curve captures trade-offs from a consumer’s
point of view: it tells us how much money the individual
would devote to the good, out of many possible choices
• This idea of sacrifice gets at the heart of what policymakers
need to know in thinking about environmental policy
decisions: how much resources would society be willing to
commit to protect the environment?
Classifying reasons to value environmental
services
• USE VALUE
• Direct impacts on people
• Health effects (e.g., asthma attack caused by high PM concentration)
• Nonhealth effects (e.g., odor or color in air)
• Increase in production costs (e.g., cost of cleaning air to manufacturer semiconductors)
• Materials damage (e.g., cost of deterioration in outdoor art or buildings)
• Damage to ecosystems
• Forests and fisheries with economic value
• Agriculture
• Recreational value
 Classifying reasons to value
environmental services
• NON-USE VALUE
1. Existence value
• I derive benefit from knowing that blue whales exist.
2. Altruistic value
• I derive benefits from knowing that other people can derive benefit from experiencing
the Arctic National Wilderness Refuge in Alaska (even though I never expect to visit it
personally)
3. Bequest value
• I derive benefit knowing that my grandchildren’s generation will inherit a preserved
Rocky Mountain National Park
Benefits estimation is hard
• As one environmental economist put it:
“Benefit estimation often involves a kind of
detective work for piecing together the clues about
the values individuals place on [environmental
services] as they respond to other economic
signals.”
Methods
1. Revealed Preference
• Use information from actual market transactions to try to
tease out peoples underlying willingness to pay for
environmental amenities
• Based on actual choices that people make, infer how they
must be making trade-offs between market goods and
environmental goods
Methods
2. Stated Preference
• Contingent valuation is the main stated preference
approach: contingent on a market existing
(hypothetically), how much would you be willing to pay?
• Survey people and ask how much they would be
willing to pay for some policy outcome
Revealed preferences approaches
• First type:
• Direct damage approach: add up expenditures by recreational visitors or
averting costs due to possible damages
Example: noise pollution
1. Homeowner expenditures to reduce noise levels inside homes
• For example, insulation, windows, fences, bushes, etc.
• We call these averting costs
• What might these expenditures tell us about people’s willingness to pay to
avoid harm?
• What are advantages and disadvantages of this approach?
Revealed preferences approaches
• Second type:
• Hedonic price approaches: statistical inference based on observed market
transactions
Example: noise pollution
2. Housing values: How might we use this
information?
• If two houses are identical except for noise, then
difference in housing values should reflect willingness to
pay by these individuals to avoid noise
• Would you expect this approach to capture everyone’s
willingness to pay? And in particular, what problems
might there be in using this approach to value the
environment?
• How do we control for other factors?

This would be an example of a hedonic price approach

Why aren’t revealed preference methods
enough?
• Requires tricky statistical methods and lots of data
• Often not possible to use this approach to measure demand curve for
specific nonmarket goods
• Also, even when it is possible, often gives lower bound on value but
does not capture all reasons to value
Measuring demand: methods
2. Stated Preference
• Contingent valuation is the main stated preference
approach: contingent on a market existing
(hypothetically), how much would you be willing to pay?
Example: noise pollution
3. Stated preference: contingent valuation
• Conduct survey to ask homeowners how much they
would be willing to pay for reductions in noise levels
around and inside their homes
• Why do you think this type of survey would be called
contingent valuation?
Video on CVM
Contingent Valuation Study: Halong Bay
Contingent Valuation Study: Halong Bay
• An environmental treasure with
tremendous aesthetic (and recreational)
value
• Source of pride for many Vietnemese
• Major draw for international tourists
• Named one of the “seven new wonders of
the world”
Environmental pressures in Halong Bay
• In 1999, there were very few
boats and no overnight cruise
ships for foreign tourists
• Between 2005 and 2011,
number of boats roughly
doubled (from 250 to 500)
• Water pollution including plastic
trash is a major problem
Contingent Valuation Study
• Consider the following three plans
• Plan A: Remove 90% of existing physical trash
(in water and shore) and keep trash at this
level for ongoing future
• Plan B: Remove 50% of existing physical trash
and keep trash at this level for ongoing future
• Plan C: Keep trash at current level for ongoing
future
Contingent Valuation Study
• What is the maximum amount (VND) you
would be willing to pay for each policy?
Suppose this is the amount you would
need to pay once to guarantee policy
achieved

• Plan A:

• Plan B:

• Plan C:
Group exercise
• Discuss the in-class survey
• How useful is the data?
• What are possible concerns?
• How might the survey be
improved?
Major criticisms of CVM
• Strategic responses: If won’t actually have to pay
anything, may overstate (or understate) in attempt to
influence policy choice
• Studies testing for this find that strategic responses are
actually not that common (and there are ways to identify
these and rule them out)
• Cognitive difficulties or lack of knowledge: Often
people have never thought about their WTP for a
change in environmental quality; consequently, there
is no obvious reason they should know their WTP
• This seems to be the biggest difficulty in real applications
of contingent valuation
Cost effectiveness
• Given a policy target that policy-makers want to pursue—for example,
to reduce the concentration of particulate matter in Hanoi by 30%--
the criterion of cost effectiveness identifies the cheapest option for
achieving this goal
• How does it compare to CBA? Which is more ambitious?
• Is a socially optimal policy cost effective? Is a cost effective policy
necessarily socially efficient?
Group exercise
• Which of the following do you think would be a more cost-effective
way to reduce air pollution in Hanoi?
• Expand public transportation
• Tax on motorbikes
• Another idea you come up with
Aim to get “the greatest bang for your buck” --
commit resources to the highest return projects
from a specified list of projects
Relationship between efficiency
and cost effectiveness?
• Which is stronger?
• Does one imply the other?
• Vice versa?