Experimental Design
The intervention consists of two experiments.
The first experiment was conducted in Pakistan in July 2013 with 1,152 individuals, Pakistani men aged between 18 and 35. They were asked to complete a survey and then offered a bonus which required checking a box expressing gratitude or checking a box declining the payment. The experiment randomly varied three separate components of the form, at the individual level, in a 2x2x2 design:
- The identity of the funding agency: the U.S. government or the Lahore University of management Sciences (a Pakistani university).
- The expectation of privacy: subjects were led to believe that their bonus payment acceptance decision would be observed by other experimental participants, or would be completely private.
- The amount of money offered: 100 Pakistani Rupees or 500 Pakistani Rupees.
The results from this experiment indicate that, when individuals act privately, approximately 25.2% of participants forgo 100 Rs because checking the box to thank the U.S. government would undermine their self-image, compared to rejection rate of 8.4% from LUMS in the private condition. When subjects anticipate that their behavior will be public, significantly fewer individuals rejected the bonus, suggesting that, in this context, a desire to conform to the majority behavior dominates any anticipated pressure from anti-American individuals. Finally, only around 10% of subjects forgo a 500 Pakistani Rupees from the U.S. government, suggesting that, even among individuals with deeply-help political identity, there is a "downward-sloping demand curve" for expression.
The second experiment was conducted in September and October 2015. In a first step of this experiment, 1,991 individuals in Lahore, Pakistan, participated in a household survey. In this first step, individuals were asked about their political affiliations. In the second step of this experiment, 1,212 of these individuals completed a 10-question survey on the Android tablet. Then, approximately half of them were offered a 100 Rs. bonus payment from the U.S. government, while the other half of them were asked direct questions about their views on the U.S. government and on the U.S. government aid to Pakistan.
Results indicate a rejection rate of 34%, similar to the findings in the first experiment. There is also an economically and statistically significant association between membership in Pakistan Tehreek-e-Insaf (PTI) political party (Pakistan's primary anti-American party) and rejection of the bonus payment offer. In contrast, there is no correlation between stated anti-Americanism and PTI membership.