AGEC 608 Lecture 18, p. 1

1
AGEC 608 Lecture 18

• Objective Consider the rationale for, and
  methods of distributional weighting of benefits
  and costs among groups in CBA.
• Readings
• Boardman, Chapter 18
• Homework 5 Chapter 13, problem 3 Chapter
  14, problem 3 Chapter 17, problem 1
• due today

2
Distributional Weighting
Policies and programs affect individuals
differently. The Kaldor-Hicks criterion
emphasizes that benefits and costs should be
summed across all groups with standing,
regardless of income or other characteristics. Bu
t the distribution of benefits and costs is
typically of interest to decision
makers.Distributional weighting is a way of
modifying benefits and costs according to the
individuals or groups in the analysis.
3
Transfers deadweight loss
Transfers of income lead to deadweight losses.
Even a transfer from high-income to low-income
individuals cannot pass the Kaldor-Hicks
test. See Figure 18.1 tax on luxury good
4
Arguments for income-based weighting
1. Diminishing marginal utility of income 2.
Income distribution should be more equal 3. One
person, one vote should apply
5
1. Diminishing marginal utility of income
U Utility Y income ?Ulow/?Ylow gt
?Uhigh/?Yhigh One received by a low-income
person has a greater impact on utility than one
received by a high-income person.
6
2. Income distribution should be more equal
Argument is premised on the idea that social
welfare would be higher if income were more equal
across individuals. Reasons 1. Inequality leads
to social ills 2. There should be an income
floor 3. Utility interdependence
7
3. One person, one vote
In most cases beneficial policies raise the
incomes of high income people more than those of
low-income people. Argument is that everyone
should have an equal influence over public
allocations. See Figure 18.2
8
Distributional weights
M number of groups r social discount
rate bjt benefits to group j at time t cjt
costs to group j at time t Wj weight on group j
9
How to derive implement?
Requires 1. Average income level of each
group 2. Estimate of income elasticity of demand
for each good affected by the policy or
program 3. Estimate of market demand curve for
each good Then compute separate CSs and weight
according to values. Note there is no economic
argument for determining the value of
weightsjudgment call.
10
Politically-determined weights
Several approaches 1. Use taxes (e.g. marginal
tax rates) 2. Use observed public expenditures
3. Use experiments to test social preferences
11
A pragmatic approach
Several approaches 1. Display weighted and
unweighted BCA results 2. Conduct sensitivity
tests 3. Compute internal weights best with
only two groups if there is an efficiency-equity
tradeoff set weight of advantaged group
1 set weight of disadvantated group to a
value that makes the NPV for everyone else 0
12
Upper bound estimates
Use a transfer program as a standard to compare
other programs with distributional effects Note
that this still requires internal
weights. Previous estimates suggest weights for
disadvantaged groups between 1.4 and 2.0
13
When to use distributional weights

• Two conditions should be satisfied
• The weighting should be targeted as a truly
  disadvantaged group, and
• either
• 2. The weighting should reduce overall social
  efficiency but make low-income groups better off
  and high-income groups worse off
• or
• 3.Weighting should offset effects of programs
  that increase efficiency but make low-income
  groups worse off.