Lecture 10 Case Study The Exxon Valdez Contingent Valuation

1
Lecture 10Case Study The Exxon Valdez
Contingent Valuation

• AEDE/NR 531
• Spring Quarter, 2006

2
(No Transcript)
3
Photos of the spill
4
The Spill

• March 24, 1989 Exxon Valdez oil tanker ran
  aground as it left shipping lanes to avoid ice
  bergs.
• The crew failed to realize how far out of the
  shipping lane it was.
• Released 11 million gallons of crude oil into
  Prince William Sound.
• State of Alaska decided to prosecute Exxon on
  lost passive-use value, also called non-use
  value.

5
Non-use Value in Legislation

• 1989 Ohio vs. The U.S. Department of the
  Interior
• Court determines that non-use values are
  compensable under CERCLA (Superfund).
• 1992 Utah vs. Kennecott Corporation
• Rejected proposed consent decree for failing to
  account for lost non-use values associated with
  groundwater contamination.
• Oil Pollution Act of 1990.
• NOAA believes that non-use values should be
  included in the Natural Resource Damage
  Assessment.
• NOAA believes reliable estimates of lost non-use
  values can be estimated using CV.

6
Survey Design

• Survey focuses on asking respondents how much
  they would be willing to pay to prevent another
  disaster similar to the Valdez spill.
• Five objectives of the survey
• Valuation of only the injuries defined in the
  survey.
• Consistency with economic theory
• Scenario comprehensibility
• Scenario plausibility
• An overall perception of neutrality by the
  respondents.

7
Survey Design

• Scenario Plausibility
• Used a referendum style survey asking Would you
  vote for a referendum that if it passed, would
  cost your household a certain specified amount.
• Focus groups used to refine publics knowledge,
  beliefs, and attitudes about the subject matter.
• Draft of the questionnaires was given in trial
  interviews.
• Draft survey sites where done in three regions
  with distinct socio-economic settings
  (California, Ohio, Georgia).

8
Key Design Issues

• Choice of elicitation method
• Nature of the payment vehicle
• Higher oil prices
• Higher taxes
• Higher prices on a spectrum of other goods
• In preliminary studies, WTP was higher for oil
  price than for tax.
• Chose tax vehicle though because at the time oil
  prices had become unstable due to Iraq invasion
  of Kuwait.
• Number of years over which payments are collected
• Chose a one time lump sum payment
• Whether the good is valued in a sequence of other
  goods

9
The Survey

• First set of question tests opinions about how
  much money should be spent to solve six social
  problems unrelated to oil spills.
• Interviewees are then shown a series of pictures
  of the oil tanker and the surrounding area,
  including wildlife.
• Information is then given about impact of the oil
  spill on wildlife.
• 100 of the areas 5,000 bald eagles were found
  dead.

10
The Survey

• A protection plan is proposed where lead boats
  would be used for tankers to follow to prevent
  any future spills.
• Asked what willingness to pay for implementing
  this lead ship program would be if it reduced
  risk of future spill by certain percentage.
• If the lead boat program is approved, all oil
  companies and all households in the U.S. will
  incur a one time tax to pay for it
• The money will pay for the program for ten years
  at which time double hulled tankers will be used.

11
Survey Results

• Four different versions of the survey were used
  with different WTP amounts.
• Used a two stage design, if respond yes to first
  amount, a second higher amount is then proposed.
  If no, then a lower amount.
• Approximately 1,500 households were interviewed.

12
Survey Results
13
Survey Results

• 30 of all people were not willing to pay 10.
• Expect Yes-Yes to decrease from A to D, and No-No
  to increase from A to D.
• A respondents WTP can then be determined to be
  between 0 and X, X and Y, or Y and infinity.

14
Survey Results

• 30 of respondents are in the 0 to 5 range.
• 11 of respondents are in the 250 and greater
  range.
• The median was in the 30 to 60 range.
• A statistical model is used to relate a persons
  WTP with their personal characteristics.
• It is then possible to make general statements
  about how different characteristics increase or
  decrease WTP.

15
Survey Results

• Theoretically, willingness to accept is the more
  appropriate measure but it is subject to greater
  biases so WTP was used.
• WTP, as opposed to WTA, then serves as a lower
  bound.
• Using the estimated median WTP to prevent another
  spill, the estimated damage from lost non-use
  value is 2.81 billion.
• Median WTP 48.97
• U.S. Household count 90,838,000
• 48.9790,838,000 2.81 Billion
• Using the mean WTP instead of the median WTP
  gives a damage estimate of 5 Billion.