By Eric Richert, Queen’s University

On November 16-17th the second workshop on the economics of strategic communication and persuasion was hosted by CIRANO, CIREQ, the John Deutsch Institute and the Dean of Arts Development Fund at CIRANO in Montreal. The keynote speaker was Joel Sobel who presented on “Sequential versus simultaneous disclosure”. Six other papers were presented in this workshop.

These papers are broadly concerned with a situation in which the person who has information is not the one with the decision making power. This creates distortions in the information that will be provided when preferences of the informed individual do not match those of the decision maker. A simple example used in Haifeng Xu’s presentation is of a teacher sending students for internships. Suppose the teacher has 10 students. The teacher knows the quality of the students and can send a signal about this quality to an employer by sending reference letters. The teacher wants all of his students to be hired while the employer only wants good students. The employer knows there are 7 students who are low quality and 3 high but not which is which. Suppose the employer is willing to hire a student only if there is at least a 50 percent chance they are good. The employer will not be willing to hire any students without some information from the teacher. Therefore, if the teacher sends all good reference letters none of his students will be hired. If he sends only good reference letters with his good students, all 3 of the good students get hired and all the bad are unemployed. But the teacher can get more of them hired, by sending all the good with a good reference letter and also sending three of the bad students with good letters. Then when an employer sees the good signal they are still willing to hire the student, and 6 of his students end up employed.

The example highlights how having communication between people with different preferences leads to distortions that would not be present if these were the same individual. In the example, the teacher prefers not give a fully informative signal to the policy maker and by doing so he can get jobs for six of his students as opposed to only 3 if the firm learned about the student quality directly. This suggests a potential role for policy, which could alter this friction.

Another easy example to imagine is a pharmaceutical company performing drug trials for drugs they would like to sell. The company has knowledge on the effectiveness of the drug before the trial. With this prior knowledge, it decides on the design of the trial. This design can affect the variance of the results but not the probability of the effectiveness of the drug. A trial design that adds noise to the experiment makes it more likely that a low quality drug would be shown to be effective. Then,he company needs to balance the incentive to increase noise with keeping the study sufficiently informative that a high quality realization changes physicians prescription behaviour.

These examples are not unique. The same basic information frictions have also been examined with criminal trials, volatility in financial markets and runs on banks, in advertising, drug trials, lobbyists and probably many more. The costs of misinforming physicians on drugs may be false prescriptions and health complications while for lobbyists it may be the passage of a policy that is only good for some private interests. Fortunately, there seems to be a place for policy interventions that add additional rules with disclosure requirements to help reduce the distortion of information away from the truth.

 

References

[1] Beauchene D., Li, J. and Li M. (2018). “Ambiguous Persuasion”. Journal of Economic Theory (Forthcoming).

[2] Boleslavsky, R., Hwang, I. and Kim K. (2018). “Competitive Advertising and Pricing”. Presentation, Workshop on the Economics of Strategic Communication and Persuasion: Application to Evidence-based Public Policy, November 17, 2018.

[3] Dughmi, S. and Xu, Haifeng  (2018). “Algorithmic Persuasion with No Externalities”. Presentation, Workshop on the Economics of Strategic Communication and Persuasion: Application to Evidence-based Public Policy, November 16, 2018.

[4] Fahn, M. and Klein, N. (2017). “Relational contracts with private information: The upside of implicit downsizing costs”. CESifo Working Paper Series 6590, CESifo Group Munich.

[5]  Hu, K. and Sobel, J. “Sequential versus simultaneous disclosure.” Presentation, Workshop on the Economics of Strategic Communication and Persuasion: Application to Evidence-based Public Policy, November 16, 2018.

[6] Markis Miltos and Renou, Ludovic (2018). “Information design in multi-stage games”. Presentation, Workshop on the Economics of Strategic Communication and Persuasion: Application to Evidence-based Public Policy, November 17, 2018.

[7] Pei, H. and Strulovici, B. “Strategic Abuse and Accuser Credibility”. Presentation, Workshop on the Economics of Strategic Communication and Persuasion: Application to Evidence-based Public Policy, November 16, 2018.