CPIC and A11 Meeting April 2, 2003 Benjamin Simon Office of Policy Analysis

1
CPIC and A-11 MeetingApril 2, 2003Benjamin
SimonOffice of Policy Analysis
2
Why do benefit-cost analysis?

• To determine the net change (i.e., net benefit)
  in social welfare as a result of a policy change.
• To determine if a given program or activity is
  worth it.
• Need to have strong, defensible form 300s.

3
Structuring the analysis

• Analysis overkill is not helpful.
• The magnitude, level of detail, and resources
  invested in the analysis should be appropriate
  for the given activity being analyzed.
• The analysis required by form 300 is similar to a
  financial analysis for most (but not all) capital
  investments.
• Get help.

4
Structuring the analysis

• Who is the client? Analysis requires a
  perspective.
• What are the goals of the analysis? This
  question is one of defining the problem.
• What is the objective? The answer to this lays
  out how the problem is to be treated as a
  practical matter.

5
Structuring the analysis

• What are the alternatives and what are the
  consequences of the alternatives?
  Answering this involves choices of techniques
  for predicting consequences.
• How are the effects to be valued?
  A technique must be chosen for valuing outcomes.

6
Structuring the analysis

• How are certain are the outcomes? Outcomes are
  always uncertain and this should be addressed
  explicitly.
• What is the choice? This step is for the analyst
  who recommends a choice to the decision maker.
• What is the decision? This step is for the
  decision maker

7
Structuring the analysis

• Identify the baseline
• Determine the period of analysis
• Identify alternatives
• Evaluate impacts quantitatively (if possible)
• Select a discount rate
• Sensitivity testing

8
Defining the baseline

• Clearly state the question the analysis is to
  address. The type of question may affect the
  baseline.
• Identify all assumptions made in specifying the
  baseline.
• Identify the nature and magnitude of the baseline
  conditions that are subject to uncertainty.

9
Defining the baseline

• Use a consistent baseline throughout the
  analysis.
• Determine the appropriate level of effort for
  baseline specification. Concentrate on those
  components of the baseline that are most
  important to the analysis.
• Clearly state the starting point of baseline
  and policy scenarios.

10
Defining the baseline

• Let the duration of important effects of a policy
  dictate the structure of the analysis and
  baseline.
• Benefits and costs do not necessarily occur at
  the same time. Forecasts of economic,
  demographic, and technological trends necessary
  for baseline specification must span the entire
  period of analysis.
• Identify and quantify to the extent possible
  risks associated with the baseline.

11
Alternatives

• One alternative should be the status quo - do
  nothing.
• The status quo DOES NOT imply nothing changes
  over time.
• Alternatives should be evaluated relative to the
  defined baseline.

12
Alternatives

• Benefits are evaluated relative to the baseline.
• Identify potentially affected benefit categories.
• Quantify the significant physical effects.
• Estimate the values associated with the
  effects.
• Evaluate sources of uncertainty.