Jill McCluskey and Elise Golan

1
All Food is Not Created EqualPolicy for
Agricultural Product Differentiation

• Jill McCluskey and Elise Golan

2
Role of Government in Changing Food Markets

• Explosion of differentiated products
• new demands
• new technologies
• new industry structures
• New role for Government?

3
Justification for Government Intervention

• Farmers fear of exploitation by buyers
• onerous, egregious, and downright illegal
  requirements by buyers
• growing power of retailers - slotting fees,
  private label, etc
• Consumer concerns (credence attributes)
• is food safe? (bioterrorism)
• does advertised attribute exist?
• Unfair competition (fraud)
• Napa Valley wine from China

4
More Justification

• Facilitate trade by reducing search and
  transaction costs
• too much information (cell phones and nutrition)
• Social objectives
• fair trade, sustainable, free range, slow, etc.
• Too much differentiation (the anti-justification)
• spatial models suggest too much product
  differentiation
• utility is non-monotonic in choice

5
Policy Intervention Grades and Standards

• How are mandatory grades and standards set?
• objective safety considerations
• producer driven?
• political/consumer driven?
• Do they improve the market outcome?
• unsure because of multiple and confounding market
  failures
• not usually but more likely when quality
  differences are great and difficult to detect
• insane for government to help producers to
  differentiate

6
Policy Intervention Certification and
Accreditation

• Certifying private standards
• high quality justifies intervention?
• Certifying the certifiers
• Certifying testing methodologies

7
Policy Intervention Information

• Reveal safety inspection information
• Change consumer behavior? Maybe
• Change firm behavior? Probably
• Reveal nutrition information
• Change consumer behavior? Maybe
• Change firm behavior? Probably
• Information least distortionary intervention?
• Cost/benefit evidence?

8
Policy Intervention Help Consumers Make the
Right Choices

• Help consumers interpret information
• smarter consumers - education programs
• smarter information - labeling and adverts
• Restrict choice (choice is making us fat)
• Paternalism (food assistance programs)
• Libertarian paternalism (food defaults)
• Expand choice (tyranny of the majority)
• stimulate effective demand (WIC)
• subsidize farmers markets, etc.
• Construct choice?

9
Research Directions

• Policy
• Mandatory versus voluntary compliance with
  standards
• Who chooses standards, who monitors government
  (which agency?) or private group
• International trade barriers to trade vs.
  consumer sovereignty.
• Free-riding in state agricultural products
• Information overload and scarcity of label real
  estate
• How product differentiation at the retail level
  filters down through the food system, e.g. farmer
  to retailer obstacles.
• Cost-benefit analysis
• Valuation
• Revealed vs. stated preferences
• Choice experiments vs. contingent valuation
• Effects of information supplied to survey
  participant.
• Theory
• Asymmetric information (search, experience, and
  credence goods)
• Heterogeneity of not well modeled from a consumer
  point of view (e.g. representative consumer
  models)
• Psychological economics (Placebo effect in
  credence goods?)
• Behavioral economics

10
Valuation Issues Revealed vs Stated Preferences

• Policy makers often must make decisions based on
  non-market valuation estimates.
• RP techniques are based on actual behavior but
  are indirect and sensitive to model
  specification.
• SP techniques are direct but hypothetical.

11
Revealed vs. Stated preferencesConsistency
Across Approaches?

• Intuitive definition of consistency revealed
  preferences agree with stated preferences.
• Statistical definitions of consistency are
• Complete consistency equal parameters and equal
  variances.
• Partial consistency equal parameters, but
  differences in terms of variances

12
Example Survey vs. Experiment

• Numbered coupons were linked to numbered surveys.
• 2 Step model approach
• model stated preferences with a double-bounded
  model.
• model actual behavior as a function of stated
  preferences and other attributes and
  socio-economic characteristics.
• Test whether consumers acted consistently in the
  market experiment with their stated preferences.
• Respondents with higher stated WTP were more
  likely to actually purchase the product.

13
Valuation Effects of information supplied to
survey participants

• Studies show individuals place a greater weight
  on negative information than on positive
  information.
• However, benefit information has a statistically
  significant effect on WTP for GM foods.
• Information provision has greater effects for
  particular groups of consumers.
• Source information matters.

14
Additional Research Directions

• Free-rider effect
• branded vs. minimum quality standards
• whats the threshold size to stop imposing
  standards?
• Differentiation by package size
• may explain quantity surcharges
• Psychology and economics
• Placebo effect utility depends on beliefs