|
** Significant at the 5% level
* Significant at the 10% level
Appendix Table 8. Probit model: Vegetable Oil – Dependent variable = 1 if a consumer is are out of the market for vegetable oil (bid is 2/3’s of bid for plain-labeled chips)
(Standard errors in parentheses, N=172)
Regressors
|
|
|
|
| |
|
|
|
Intercept
|
-1.574 **
|
-1.339 **
|
-1.248 **
|
-1.496 **
|
-1.200 **
|
-1.392 **
|
-1.484 **
|
-1.378 **
|
|
(0.280)
|
(0.298)
|
(0.396)
|
(0.489)
|
(0.430)
|
(0.416)
|
(0.513)
|
(0.518)
|
Anti_info
|
0.950 **
|
0.958 **
|
0.778 **
|
0.829 **
|
0.772 **
|
0.791 **
|
0.822 **
|
0.864 **
|
|
(0.313)
|
(0.319)
|
(0.347)
|
(0.353)
|
(0.348)
|
(0.352)
|
(0.357)
|
(0.366)
| Pro_info |
|
|
-0.327
|
-0.302
|
-0.334
|
-0.368
|
-0.369
|
-0.359
|
|
|
|
(0.280)
|
(0.283)
|
(0.282)
|
(0.284)
|
(0.289)
|
(0.297)
|
Ver_info
|
|
-0.582 **
|
|
|
|
|
|
-0.691 **
|
|
|
(0.261)
|
|
|
|
|
|
(0.277)
|
Income
|
|
|
|
0.0034
|
|
|
0.0028
|
0.0040
|
|
|
|
|
(0.0038)
|
|
|
(0.0039)
|
(0.0040)
|
Labels
|
|
|
|
|
-0.071
|
|
-0.186
|
-0.143
|
|
|
|
|
|
(0.252)
|
|
(0.264)
|
(0.273)
|
Informed
|
|
|
|
|
|
0.357
|
0.378
|
0.493
|
|
|
|
|
|
|
(0.256)
|
(0.268)
|
(0.280)
|
|
|
|
|
|
|
|
|
| | |
|
|
|
|
|
|
|
|
** Significant at the 5% level
* Significant at the 10% level
Appendix Table 9. Probit model: All Products – Dependent variable = 1 if a consumer is are out of the market for All three products (bid is 2/3’s of bid for plain-labeled chips)
(Standard errors in parentheses, N=172)
Regressors
|
(1)
|
(2)
|
(5)
|
(4)
| (5) |
(6)
|
(7)
|
(8)
|
Intercept
|
-2.045 **
|
-1.744 **
|
-1.198 **
|
-1.223 **
|
-1.174 **
|
-1.364 **
|
-1.258 **
|
-1.108 **
|
|
(0.410)
|
(0.413)
|
(0.421)
|
(0.515)
|
(0.456)
|
(0.447)
|
(0.543)
|
(0.550)
|
Anti_info
|
1.204 **
|
1.156 **
|
0.718 *
|
0.724 *
|
0.714 *
|
0.751 *
|
0.735 *
|
0.786 *
|
|
(0.437)
|
(0.429)
|
(0.384)
|
(0.391)
|
(0.384)
|
(0.394)
|
(0.400)
|
(0.408)
| Pro_info |
|
|
-0.613 **
|
-0.611 **
|
-0.619 **
|
-0.667 **
|
-0.687 **
|
-0.721 **
|
|
|
|
(0.280)
|
(0.281)
|
(0.280)
|
(0.285)
|
(0.289)
|
(0.298)
|
Ver_info
|
|
-0.684 **
|
|
|
|
|
|
-0.666 **
|
|
|
(0.309)
|
|
|
|
|
|
(0.274)
|
Income
|
|
|
|
0.0034
|
|
|
-0.0034
|
0.0029
|
|
|
|
|
(0.0039)
|
|
|
(0.0040)
|
(0.0040)
|
Labels
|
|
|
|
|
-0.033
|
|
-0.147
|
-0.108
|
|
|
|
|
|
(0.253)
|
|
(0.267)
|
(0.275)
|
Informed
|
|
|
|
|
|
0.372
|
0.418
|
0.530 *
|
|
|
|
|
|
|
(0.259)
|
(0.273)
|
(0.285)
|
|
|
|
|
|
|
|
|
| | |
|
|
|
|
|
|
|
|
** Significant at the 5% level
* Significant at the 10% level
i
Endnotes
This is in contrast to the tradition in experimental economics of having an individual participate in multiple trials. See Shogren (2002).
ii When two adults in a household participated, the Iowa State Statistics Laboratory talked to both of them separately to obtain a commitment to participate and they were told that they would be assigned to different groups.
iii The complete set of information given to participants is available upon request from the author.
iv The nth-price auction has the advantage of having an endogenous clearing price, a main feature of the Vickrey auction, and a random clearing price, a feature of the BDM auction. Thus, in many ways, it combines the best features of these two demand revealing auctions. For a more detailed description of the benefits of the random nth price auction, see Shogren et al. (2001) or Rousu et al. (2002).
v If one assumes that there is little or no income effect from the deck of cards and box of pens, the two bids on the candy bar should be the same. The reason is that since the deck of cards and box of pens are neither complements nor substitutes to the candy bar, they should not impact the bids on the candy bar. A Wilcoxon signed-rank test confirmed that the bids for the candy bars are not significantly different in the two rounds, with a test statistic of 0.03. This result does not contradict the notion that the subjects’ bidding behavior was reasonable.
vi When a participant received both pro-biotechnology and anti-biotechnology information, the order was randomized, so that some participants received the pro-biotechnology information first, and others received the anti-biotechnology information first.
vii When third-party information was distributed, it always was distributed after the other information sources.
viii These experiments were set up to minimize endowment effects. i. e., participants were endowed at the beginning of the experiment with $40 but not GM-food. See Shogren (2002 for evidence on endowment effects.
ix Recall that our participants are only given money and no physical commodity, and this minimizes the endowment effects.
| 