CostBenefit Analysis: The Details

1
Cost-Benefit Analysis The Details

• How can economics help determine the optimal size
  of a project or extent of a regulation?

2
A few examples

• What should be the CO concentration standard in
  tailpipe emissions?
• How large should the Channel Islands marine
  reserve be?
• Can we measure loss to recreationists of the
  Forest Adventure Pass?
• Add another lane to Hwy 101?
• Close Mission Canyon to cyclists?
• What habitat to buy to protect endangered species
  (e.g., Least Bells Vireo bird)

3
CBA main principle

• Quantify all costs and benefits in a common
  measure (usually )
• Note that we have ways of quantifying non-market,
  even non-use values.
• Common metric need not be -- e.g., health-
  health analysis, with health as metric
• E.g., consider a regulation that saves lives
• Benefits directly measured in terms of lives
  saved
• Costs indirect costs increase deaths since lower
  incomes lead to higher mortality (13 million in
  extra costs results in 1 statistical death)
• E.g., Compare risks in lower and higher income
  countries
• Compare projects based on net effects on health.

4
When to do CBA?

• Need to compare regs to achieve externally
  defined social goals CEA
• Need to set a level of a reg, balancing pluses
  and minues CBA
• Need to decide on a project with multiple goals
  that cannot be reduced to a single metric
  multi-goal analysis

5
Example issue is how big

• Project size, Q.
• Benefits B(Q)
• Costs C(Q)
• Maximize Net Benefits, B(Q)-C(Q)
• Ignore issue for time being of how to quantify
  B(Q) and C(Q)

6

One view of this problem Maximize TB-TC
TB(Q)
TC(Q)
Q
Q

TB(Q)-TC(Q)
Q
7

Q is where TB-TC is maximized. Also where
MBMC.
TB(Q)
TC(Q)
Q
Q

MC(Q)
MB(Q)
xC
Q
8
Discrete sized projects

• If deciding between projects A, B, C
• Pick one with highest net benefits (TB-TC),
  provided net benefits gt 0.
• May have values that are difficult to quantify.
• Quantify values you can, then compare projects
  along as few dimensions as possible
  (multi-criteria analysis) examine tradeoffs
  between alternatives.

9
Big Questions

• How to measure/conceptualize benefits?
• How to measure/conceptualize costs?
• How to treat costs and benefits that occur at
  different points in time?
• How to deal with distributional issues?
• How to deal with risk?

10
Question 1 Conceptualizing Benefits

• Suppose a river is dead, what is the value of
  reducing pollution by one unit?
• What is the value of increasing visual range by a
  unit in the West?
• How do we value the extra units of electricity?

11
Basic measure of value is willingness-to-pay

• Demand curve is marginal willingness to pay

MWTP
First units very valuable
Last units less valuable
Quantity of water
12
Consumers Surplus (CS)

CS(q)
p
D(x)
x
q
13
How are benefits calculated?

• Demand, D(x), measures MB.
• Consumers Surplus is the total benefit to
  consumers minus their cost.

14
Example gross value of water from new dam
(excluding costs)
Price
New dam
. . . . . . . . . . . . . . .. . . . . . . . .
Demand for water
Addl Value
Acre-feet of water
15
What about nonmarket goods?
Suppose there were a market
Price
Demand for air quality
Air quality
BUT, NO MARKET price similar to MWTP
16
Environmental goods

• Demand for env goods just as real as demand for
  market goods just harder to measure
• Demand is a measure of intensity of preferences

17
Costs are simpler

• Some units are cheap to produce
• Marginal units are most expensive
• Costs consist of
• Fixed costs
• Marginal costs

18
Marginal costs plus fixed costs add up to total
costs
Last units pricey
MC
First units cheapest
Quantity
19
Costs come in different flavorsPrivate,
external and social

• In principle, need to capture all costs and
  benefits.
• Social costs may exceed private costs.
• Difference is the external cost the monetized
  cost of the externality.

/gal
MPC
P0
Q0
Gallons Of Gasoline
20
Social vs. Private Costs

• In principle, need to capture all costs and
  benefits.
• Social costs may exceed private costs.
• Difference is the external cost the monetized
  cost of the externality.

/gal
MPC
P0
MEC
Q0
Gallons Of Gasoline
21
Social vs. Private Costs

• In principle, need to capture all costs and
  benefits.
• Social costs may exceed private costs.
• Difference is the external cost the monetized
  cost of the externality.

/gal
MSC
P
MPC
P0
MEC
Q0
Q
Gallons Of Gasoline
22
Producer Surplus (PS)

MC(x)
PS(q)
p
x
q
23
How are costs calculated?

• Supply, S(x), is same thing as MC.
• Producer Surplus is the total revenue to
  producers minus their cost.

24
Put it together
P
Total Surplus
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
Q
25
Where is CSPS maximized?

Tension Too little produced At too high price.
CS low, PS high
CS
p
Supply, S(x)
PS
Demand, D(x)
x
q1
q
26
Suppose goods supplied in fixed amount
Price
Supply of land
Consumer surplus
Producer surplus (goes to land owners)
Market Price
Demand for land
Land
27
If captured all costs benefits

• Then we want to maximize CS PS which would
  occur where Supply Demand.
• Challenge is to capture all costs and benefits to
  accurately measure MC MB.
• Note
• Surplus Max equals CSPS maximum

28
First Theorem of Welfare Economics

• In a competitive market
• Surplus is maximized at a market equilibrium
• Implications
• Can rely on market if we are sure of competitive
  market

29
Example Add a dam
Supply of water increases price falls. What
happens to PS? CS?
Price
PS Before and X After X and CS ??

XXXXXXXXXXXX
Q Water
30
Implicit Assumptions Distribution

• Distributional consequences ignored
• Put in dam some win, some lose
• Benefits cancel losses
• OK if compensation occurs
• Compensation Principle
• Dam is good idea if winners can compensate losers
• If there is enough surplus, dam is a good idea
• BUT, compensation need not occur

31
Implicit Assumptions Income
Price
Y50,000 per year
Y30,000 per year
Restaurant meals
Demand and thus surplus depend on income
distribution Therefore Change in income
distribution will change results of CBA TO USE
CBA, MUST BELIEVE INCOME DISTRIBUTION IS OK
32
Implicit Assumptions Completeness

• What happens with difficult to monetize benefits?
• E.g., clear view of Santa Cruz Islands
• Difficult to monetize benefits often omitted
• Results in bias against environmental benefits

33
Implicit Assumptions Other

• Moral and political dimensions omitted
• Should we do a cost-benefit analysis on executing
  someone who has committed a crime?
• Are there other issues when lives are at stake?
• Are intergenerational issues adequately treated
  by CBA?

34
Ten Steps to doing and using a CBA

• 1. Decide whose benefits and costs count
• 2. Select the portfolio of alternative projects
• 3. Catalog potential physical impacts and
  determine how they are measured
• 4. Predict quantitative physical impacts over
  life of project
• 5. Monetize (or other metric) all impacts
• 6. Discount for time to find present values
• 7. Sum add up all benefits and costs
• 8. Perform sensitivity analysis
• 9. Choose alternative with largest social
  benefits
• 10. Make policy recommendation, using CBA only
  as part of guidance
• 11. Identify limitations specific to your
  analysis