Estimating Social Welfare Preferences

1
Estimating Social Welfare Preferences

• Helen Scarborough
• Deakin University

2
Distributional issues

• Increasing awareness of need to incorporate
  distribution in policy analysis
• Disparity in distributional effects of
  environmental policy both environmental quality
  and financial

3
Examples of distributional issues

• Benefits and costs of reduction of greenhouse
  emissions
• Disparity between private costs of revegetation
  and social benefits
• Location of environmental bads between areas of
  higher and lower income

4
CM to estimate distributional preferences

• Interested in development of CM as a method of
  estimating distributional preferences
• Emphasis on estimating social welfare function
  and social welfare preferences.
• Social Welfare Function (SWF) ranks social states
• Function of utility for individuals (or groups)
  in society
• Each SWF represents one persons view of
  allocation of utility across individuals, or
  groups, in society

5
Choice experiment

• Welfare maximisation
• Three hypothetical environmental policies
• Attributes in terms of utility to different
  generations
• Described in dollar terms
• Respondents reminded that dollars represent total
  well-being
• Described in terms of change in utility to person
  representing particular group

6
Attributes and levels

• Attributes
• Utility change Person Aged 50
• Utility change Person Aged 25
• Utility change Newborn
• Levels Five
• -1,000, -500, 500, 1,000 1500
• (focus groups suggested these amounts considered
  realistic and large enough to influence choice)

7
Research instrument

• Reference key for choice set

8
Example of choice set
9
Social Marginal Rates of Substitution
Aged25/ Aged50 Newborn/ Aged50 Newborn/ Aged25
MNL Model 1.63 (0.94, 3.76) 2.23 (1.25, 5.26) 1.37 (0.89, 2.14)
ML Model 1.75 (0.59, 4.71) 2.34 (1.00, 5.83) 1.36 (0.77, 2.67)
10
Five areas for discussion

• 1. Strategies for conveying to respondents the
  distinction between social welfare maximisation
  and utility maximisation
• 2. Determining attributes and levels to describe
  utility changes
• 3. Exploring sensitivity to choice of numéraire
  in estimating marginal utility
• 4. Analysing decision strategies and how these
  advance social welfare literature
• 5. Incorporating efficiency cost?
• 6. Need to split distributional weight?

11
1. Welfare max not utility max

• Can we expect respondents to adopt social welfare
  max as decision making criteria?
• Application of veil of ignorance?
• Cheap talk strategies?

12
2. Attributes and levels describing utility
changes

• Index of well-being?
• Using dollars to describe utility changes?
• Difficulties with common metric
• Over time
• Over geographical boundaries

13
3. Choice of numéraire

• Exploring difference between utility changes
  expressed in terms of dollars or for example,
  access to environmental quality
• Which marginal utility?

14
4. Analysing welfare max strategies

• Linking decision-making strategies to welfare
  literature
• Utilitarian?
• Rawlsian?
• Egalitarian?
• CM provides potential to analyse heuristics.

15
5. Need to split distributional weights

• Two components-
• Marginal utility of income and
• Marginal social welfare
• Is it possible to estimate each component?

16
Conclusion

• Distribution important aspect of policy analysis
• Significant unresolved issues.
• Many further research areas