Experimental Design Details
Our experiment proceeds in the following steps:
1. Randomization: For a given individual, we draw their information treatment’s dimension d∈{Political,Environmental,Diversity,Placebo}. We also draw a set of five firms F from our complete set of firms*.
2. Benchmarks: We tell individuals the average value among Fortune 500 firms for each CSR dimension, to use as a benchmark when making their subsequent predictions.
3. Prior Beliefs: For each CSR dimension, individuals simultaneously predict the value associated with each firm in F. Individuals also predict the share of Republicans and Democrats that report participating in a boycott.
4. Gift Card Valuations Practice: Respondents complete an example and comprehension checks for our gift card valuation elicitation used in (5) and (9).
5. Prior Gift Card Valuations and Favorability: For each firm f∈F, we then elicit the value $x such that the respondent is indifferent between $x in AmeriPoints** vs. a $50 e-gift card at firm f. Individuals also report their favorability toward each firm on a feeling thermometer.
6. Firm CSR Treatment: Randomly selected individuals receive the true values for all firms F along dimension d (some receive political information for all firms, some receive placebo information for all firms, etc.).
7. Strategic Incentive Treatment: Randomly selected individuals receive information on the share of Republicans and Democrats who have reported boycotting any firm before.
8. Posterior Beliefs: Repeat (3), eliciting beliefs about a firm’s CSR and the share of Republicans and Democrats that report participating in a boycott.
9. Posterior Gift Card Valuations and Favorability: Repeat (5), eliciting for each firm a respondent’s valuation (in AmeriPoints) of a $50 firm gift card.
a. Facebook Sharing Treatment: Prior to giving their posterior gift card valuations, randomly selected individuals are notified about the option to later share their values on Facebook in exchange for a higher probability of winning a lottery for AmeriPoints.
10. Firm Gift Card Ranking and Familiarity: Individuals rank $50 in AmeriPoints and $50 gift cards at each of the firms f∈F. They also indicate their familiarity with each of these firms.
11. Ideal Points: We ask individuals for their most preferred point on each CSR dimension (e.g., the share of firm donations which should go to Republicans).
12. Network Ideology: We ask about the ideology of a respondent’s Facebook friends.
Accuracy in all belief predictions is incentivized by a lottery for AmeriPoints, with answers closer to the true value increasing a respondent’s probability of winning this lottery. Our gift card value elicitations are also incentivized by a separate lottery. For at least 50 individuals randomly selected as lottery winners (2% of respondents), we randomly draw a firm f, a number x∈[0,100], and a gift card elicitation. The lottery winners are then given their choice between a $50 e-gift card at firm f and $x in AmeriPoints, as determined by the preference they expressed in the chosen gift card elicitation.
*The complete set of firms is a subset of consumer-oriented Fortune 500 firms for which political, environmental, and gender diversity information is available from the specified sources, and which offer $50 e-gift cards.
**AmeriPoints are the typical cash-equivalent incentive scheme used by NORC’s AmeriSpeak® Panel, through which TESS fields successful proposals. AmeriPoints can be redeemed for Mastercard® Rewards Cards, Amazon.com gift cards, movie tickets, restaurant gift cards, and other rewards. For detail, see https://www.amerispeak.org/faq.