Our society is driven by social influence. The salesman who wants to sell you a car, the politician who wants your vote and the Tiktok influencer who wants you to “like” their videos have one thing in common: they are vying for your attention.
For scientists – in psychology, politics and computer sciences – understanding quite how we attract people’s attention is a challenge. In a recent study, we found that social influence is best understood as a competition.
People often think of social influence as a one-to-one relationship between the influencer and their target. But every election has at least two candidates. Similarly, thousands of videos are uploaded on TikTok every day, each hoping to be the one that goes viral. And, every time a salesman sells a car, his competing colleagues lose a customer.
Influencing is a zero-sum game. More than how to influence people, the question is how to be more influential than others.
We designed a laboratory model of social influence in the form of a game to be played by three people: one client and two advisers. The client has to buy one of two lottery tickets but has no information about which is better. The advisers, who have private access to such information, and compete for being hired by the client.
Our model, much like social influence in real life, is a zero-sum game: one adviser’s success is the other’s failure. This allowed us to use game theory to find an optimal strategy for the adviser.
Our analysis of game theory showed that a clear strategy can be formulated: if you already have influence (if you are hired), be vague and stay close to the truth. If, conversely, you are ignored, be loud, exaggerate and, if necessary, just lie to stand out.
We conducted seven experiments with more than 800 participants who played the role of the client. We found that strategic distortion of the truth outperformed honest advising in winning over and retaining individual clients in up to 80% of the time. When advisers were strategically dishonest, they also succeeded in swaying groups of clients who elected their adviser democratically in each round.
This strategy, of course, is familiar to anyone who lived through the Brexit campaign, as former UK prime minister David Cameron clearly describes in his book, For the Record. According to Cameron, Boris Johnson played precisely the card we would expect the disadvantaged candidate (the one challenging the incumbent) to play. Cameron advocated remaining in the EU, so Johnson embraced the leave campaign.
Cameron writes that Johnson was making a strategic choice to differentiate himself from the incumbents. Johnson, he says, “risked an outcome he didn’t believe in because it would help his political career”. And, he adds, because Johnson was certain the leave side would lose, backing it brought little risk of breaking up the government he wanted to lead one day. “It would be a risk-free bet on himself,” Cameron writes.
Central to this model were the three hallmarks of competition for social influence: information asymmetry, delegation of future decisions and intractable uncertainty.
Information asymmetry occurs when influence seekers (politicians or advisers) know more about an issue than the people they seek to influence (voters or clients). In the political arena, the issues at stake are often multidimensional and too complex for people to be fully informed about. In the Brexit vote, for example, the regions most strongly favouring Leave were also —- to the surprise of many voters —- the most dependent on European Union markets for their local development.
Competition for social influence also often involves a delegation of power: voters or clients granting politicians or fund managers the power to make future decisions on their behalf.
Finally, predicting the future is hard. Political science writer Philip Tetlock, in his 2017 book, Expert Political Judgement shows how pundits who are regularly tasked to predict uncertain future events in finance, politics, or sports often turn out to be wrong. Competition for social influence thus tends to take place under high outcome uncertainty. Evaluating advice accuracy is difficult under high uncertainty. This creates opportunities for competing advisers to seek influence strategically because few would remember the failure of their radical but dishonest predictions.
Our findings suggest that the success of dishonesty is due to our willingness to jump to conclusions in hindsight. This chimes with what research shows on how we assess the choices we have made.
If an adviser was the only one to predict a bad outcome before it happened, we tend to think that they must have known something that others did not. While this may sometimes be true, often it is just pure luck. A strategic adviser takes advantage of this willingness we have to trust our hindsight to inflate their confidence or even, dishonestly advise against the available evidence simply to stand out.
An honest adviser, when ignored, is less effective (than their dishonest rival) in persuading the client to shift: commitment to honesty stops them from positioning themselves as a radical alternative if there is no evidence to justify it.
These kinds of strategies are repeatedly and ruthlessly employed by attention-hungry influencers because they work. Our analysis helps explain why politicians who are repeatedly found out to have lied could continue to enjoy public support. We hope that our work will generate awareness in the public and help us all to see through such manipulative and dishonest strategies and protect the citizens against them.
This research was done in collaboration with Ralf Kurvers, Jurgis Karpus, Uri Hertz, Marta Bolade, Bertrand Jayles and Ken Binmore. We acknowledge financial support by the Max Planck Institute for Human Development to R.K. Support from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (819040; acronym: rid-O) to B.B. and J.K. is acknowledged. BB was also supported by the NOMIS foundation and the Humboldt Foundation . J.K. was supported by LMUexcellent, funded by the Federal Ministry of Education and Research (BMBF) and the Free State of Bavaria under the Excellence Strategy of the Federal Government and the Länder. U.H. was supported by the National Institute of Psychobiology in Israel ( 211-19-20 ) and the Israel Science Foundation ( 1532/20 ).