Using the last trials, the distribution of the bids in treatment 1 was compared with the distribution of
the bids in treatment 2. We could not find an ambiguity effect either between the treatments (p=0.292). In
summary, no ambiguity effects were generally found in our experiments either from the within-sample
treatment (treatment 3) or from the between-sample treatments (treatments 1 and 2). The only exception
is when comparing the last trial before the provision of the unambiguous incidence rate information and
the first trial after the provision of the unambiguous incidence rate information in treatment 3. It is not
clear why ambiguity in incidence rate information does not generally have a significant effect on WTP.
However, it is possible that parents do not care about the clearness or ambiguity of incidence rate because
this is in regard to the health of their child. This is consistent with Kahneman and Tversky’s (1979)
finding that people tend to be risk-averse when they are faced with a small chance of losing a large
amount. This behaviour is generally referred to as the “overweighting of small probabilities”[18].
4.3 Testing for the effect of safe-handling information
We hypothesized that the safe-handling information we provided decreases the WTP because
parents can then self-control the health risk. To investigate this, we tested the null hypothesis that the
population distributions within each treatment are identical using Wilcoxon’s signed-rank test. We used
the last trial of the different treatments for the test. In all the three treatments, the effect of the information
was found to be statistically significant (see Table 4). In treatment 1, the WTP reduction is statistically
significant at the 5% level while in treatments 2 and 3, the decrease in WTP is statistically significant at
the 1% level. Hence, information on the preparation techniques significantly reduced participants’
valuation of the label (also see Figure 2).
In summary, the key finding that emerges from our experiment is the significant effect of the
provision of safe-handling information. Providing information to the participants on the preparation
techniques that helps to decrease the health risk influences the WTP and leads to a significant decrease in
WTP, as expected. Interestingly, the WTP did not decline to zero with the provision of the safe-handling
information. It is, however noteworthy that we had five zero bids in treatment 2 and seven zero bids in
treatment 3 when the information on the preparation techniques was provided. These results may suggest
that the information on the unambiguous incidence rate made it easier for the respondents to calculate the
risk and to value the information on the preparation techniques. In treatment 1 (“Unclear”), we just had 2
zero bids.
5 Conclusion
This study assessed parents’ WTP for quality assurance labelled powdered infant milk formula.
Using experimental auctions, our results indicate that the mean price premiums parents were willing to
pay ranges from 61 to 133 Eurocents, given a basic price level of 115 Eurocents per 100 grams of
powdered infant milk formula. This result means that parents are willing to pay price premiums from 53%
to 116% of the base price per 100 grams. Our experiments also examined if ambiguous information about
health risk (i.e., incidence rate) as well as information about safe-preparation techniques affect WTP. Our
results generally suggest no significant ambiguity effects but substantial safe-handling information effects
on WTP. The WTP declined by 39 to 69 Eurocents after the provision of information about the
preparation techniques. This finding suggests that our subjects attached a lot of importance to safe food
handling techniques that could reduce the health risk.
Our findings imply that parents indeed significantly value a quality assurance label with or without
clear incidence rate information. Parents’ valuation of this label, however, is reduced with the provision
of safe-handling information. It may then be prudent for the infant milk formula industry to provide both
a quality assurance label and the information on safe-handling preparation techniques. Future studies
should, however, replicate our study to assess the robustness of our findings in other countries. Analyzing
the welfare effects of our findings is also warranted given data availability.