Sandrine Blanchemanche Uncertainty, Decision Making & Field experiment Met@risk, Food Risk Analysis...

18
Sandrine Blanchemanche Uncertainty, Decision Making & Field experiment et@risk, Food Risk Analysis Methodologies Unit INRA, Paris & S. Marette, J. Roosen and P. Verger : Analyse des choix alimentaires et méthodes expérimentales, 3 dec 2008

Transcript of Sandrine Blanchemanche Uncertainty, Decision Making & Field experiment Met@risk, Food Risk Analysis...

Sandrine Blanchemanche

Uncertainty, Decision Making & Field experiment

Met@risk, Food Risk Analysis Methodologies UnitINRA, Paris

& S. Marette, J. Roosen and P. Verger

PAP: Analyse des choix alimentaires et méthodes expérimentales, 3 dec 2008

Context « Risk Analysis »

• Food safety regulation: reduce uncertainty to obtain a « scientifically-based » policy

• Framework of Risk Analysis: Risk Assessment, Risk Management and Risk Communication

• Increase of health and risk information used as a regulatory tool to manage some food risks (labelling, educational programs, hazard warnings, consumption advisory)

• Advisory– Not just an information given to public– Suggests a change in consumption behaviour

Problem

• Regulators give some information coming from risk assessment to public & try to change consumers’ behaviour

• This regulatory approach is rooted in normative theory in which optimality and rationality are central

• It assumes that individuals are unable to optimize their decisions because of imperfect knowledge about risk that the advisory is supposed to overcome by offer them proper information

• Can a regulatory tool based on rational choice theory be efficient in reducing risky behaviours?

Case: Fish Consumption Advisory

• Despite the initial scientific uncertainty about the risk it leads to a consumption advisory assuming a consumer who is a rational decision maker

Health problem: risk and benefit of fish consumption

• Methylmercury occurs in the environment and accumulates in fish (larger, longer-living and predatory species) This neurotoxin can cause damage to the central nervous system when young or unborn children are exposed to it The degree of exposure to methylmercury may vary depending on fish species and quantity consumed

• Fish is a healthy food: Omega 3 confers benefits to the foetus such as infant cognition and improvement of cardiovascular health for the whole population

• The regulation is tricky and will be based on the risk assessment

Uncertainty & Risk Assessment

• Several uncertainties to deal with:– Level of contamination of mercury in fish– Adverse health effects (controversy between

epidemiological studies)– Consumption of fish by population– Age cut-off for the main period of risk to neurodevelopment– Extrapolation from animals to humans

• Risk assessment defined a safe upper exposure limit to methylmercury to determine the appropriate level of protection for the population

• Risk management (most of countries): issued an advisory

Advisory

• It mentions the group at risk: Women who might become pregnant Pregnant women, nursing mothers Young children• It describes the health benefits (from Omega 3) and

the risks (from mercury) of fish consumption• It gives the following consumption advices:

- Limit to 2 meals per week fish and sea food- Within these 2 meals, limit to 1 meal per week: Canned tuna, or rock salmon, or grenadier, or ling (blue ling)- Do not eat: Fresh tuna, Shark, Swordfish, Marlin, Grouper

Field Experiment

• Objective: The experiment focuses on decision-making of household receiving the advisory

• 201 households in Nantes (803 individuals)– Eat fish at least twice a week– With a woman member with age between 24-38 years– With a child member under age of 15 years.

• Selection of households isrepresentative for age and socio-demographic charac.

• Field experiment :measurement of fish consumptionof each individual householdmember during 5 months

Nantes

Treatment group (H=99, N=400)

1st Visit

Introduction & Questionnaire

Control group (H=102, N=403)

2nd Visit

Consumption data May

3rd Visit

Consumption data June

4th Visit

Consumption data Sept.

Reminder by

phone

Advisory Questionnaire Questionnaire

May

May

June

June

September

September

July/August

July/August

Time

Time

Fish Consumption Frequencies per week

Treatment Control

May

June Sept.

May

June Sept.

Women

All fish 3.23 2.82 2.83 2.93 2.82 2.65

Fish non mentionned 2.61 2.29 2.39 2.33 2.21 2.15

Fish to limit to 1 meal per week 0.51 0.43 0.34 0.53 0.53 0.40

Fish to do not consume 0.10 0.10 0.10 0.08 0.09 0.09

Short term information effect: treatment effect (treatment/control) is effective (weak, but statistically significant) in June, but not in September:  Traffic Cop Effect

Compliance &

Rational Behaviour Advised• The rational behaviour advised is:

– Women and children should reduce their fish consumption and to limit it to 2 meals per week

– While men should not reduce their consumption

• Household do not comply with the advisory after few months

> Different advices for individuals in the household are not consistent with the usual habit of consumption

Discussion

• People do not know precisely how much fish they eat

> They do not see themselves as a « group at risk »• Ambiguity and Complexity in the Advisory

– Contradicting previous communication about fish consumption « fish is good for health »

– Too much information • At the end of the experiment, they remember

the main information (to limit fish) • But there is a low level of recall of species,

limits per week and people targeted

Uncertainty of risk

vs Certainty of the status quo

• Initial scientific uncertainty leads to a specific advisory where co-exist precaution and rationality

• Ambiguity, complexity, precaution, and risk which cannot be experiment directly leads to confusion and seems unlikely to provoke behaviour changes

• People feel uncertain about the risk and are not confident that they are really at risk so they prefer the certainty of the status quo

Conclusion

- Consumers change their consumption to a small extent but insufficiently to comply with the advisory.

- Finally, we conclude that the advisory doesn’t reduce uncertainty (as it is supposed to do) but transfers the initial uncertainty onto consumers turning it into ambiguity and complexity.

Thank you for your attention!

Rationale given to do not change behaviors

Do not want to change habit13%

Fish Species not consumed34%

No targeted28%

Cannot be careful with everything25%

75% of women did not change their behaviour