Fool me once

15/11/2018

Coordination is a critical component in much of our economic, social, and political lives.  We need to coordinate our actions in situations as trivial (dare I say) as a recreational football match, to critically important endeavors like spurring growth in developing economies.  Within any organization, many – perhaps most – employees work as members of a team and completion of projects requires the team to successfully coordinate behavior.  Coordination, then, captures moments when all members of a group have a shared goal and only their collective efforts will achieve that goal.  The problem is that coordination failure is widespread as seen in events like firm-level bankruptcy or countries failing to climb out of poverty traps, largely driven by the fact that individuals are uncertain that others will join them in their actions.

Research by economists and other social scientist points to some hope. In many situations, effective leadership can help to avoid coordination failure, but one important factor in being effective is that a leader must be credible. Previous studies demonstrate that leaders are effective through their access to and use of better information, which can be thought of as informational credibility. My co-authors (David Cooper, also at Florida State University, and Roberto Weber at the University of Zurich) and I wanted to investigate a possible added dimension of a leader’s credibility.

Consider the previously mentioned case of a nation trying to climb out of poverty.  A leader must attract business investment in the local economy, yet even if they report in good faith that investment will be profitable, a firm may rightly hesitate unless they know that their investment will be joined by many other firms as well.  Similarly, if a business is failing, then its employees will only stay and give their full effort if they know they will be joined by their co-workers.  Otherwise, they’d better spend that time updating their resumé.  This model has been used broadly to address bank runs and lending freezes, and to show that currency crises can be brought on by miscoordination among traders, leading to speculative attacks. Central banks will commonly issue public statements with the aim of preventing such behavior.

What do these situations have in common?  In each of them – and in fact many other settings analogous to these – the outcome relies on complementarities in actions between the involved parties.  That is, the returns to my actions depend on your chosen action as well as the choices of others. Furthermore, the leaders here – whether elected official, manager, or bank official – benefit from investment in some cases where the individuals may not.  For instance, if enough traders hold off on making pessimistic currency bets, the central bank may avert a crisis, but those traders who kept their holdings may see losses due to modest devaluation.

To overcome coordination failure, leaders need to use their superior information to generate “buy-in,” but must maintain credibility in the face of strong incentives to misrepresent their information. This means that they need to have full efforts from their group members, and critically, everyone needs to know that their fellow group members will also follow the leader’s advice.  We refer to this as a leader’s social credibility, and we show that it can be at least as important, if not more so, than a leader’s ability to accurately use their better information. 

We set out to study this idea of social credibility using controlled laboratory experiments.  Lab experiments have long been a staple of empirical social sciences. They allow us to recreate an environment from the ground up, stripping away confounding factors that can make data from naturally occurring environments problematic to interpret.  What we needed to do, then, was devise a game to re-create the central tension found in each of the examples above between the leaders and the group of individuals they advise.  

Our “investment game” involved six-person groups where each person had to decide how to deal with a coordination problem, needing to decide repeatedly whether or not to invest in a risky project.  The type of risk faced by the players was as follows:  If all investors invest and the project is high enough quality, everyone in the group benefits (in terms of payoffs).  For the worst quality project, nobody benefits from investment.  However, for one (out of six) possible project qualities, full investment benefits the advisor but harms the investors.  So in one out of six possible qualities, the advisor’s incentives are directly opposed to that of each investor.  To increase this tension, only the advisor has information on the likelihood of a high quality project (but they do not know the quality for certain), so investors receiving advice to invest must trust the advisor, but also must trust that their fellow investors will also trust the advisor.

A critical property of this game is that it includes multiple equilibria – sets of actions from which nobody would prefer to deviate on their own.  One equilibrium represents the advisor’s preferred outcome, in which they recommend investment whenever it benefits them directly.  Because there are four investment states that benefit everyone, and only one that harms investors, investors should be willing to go along with this advice.  A second equilibrium acknowledges that investors may “punish” aggressive advice by ignoring the advisors for a while.  If this behavior is expected, the advisor will only recommend investment when it benefits investors. 

The advisor, then, needs to make a choice.  If the advisor plays to maximize immediate earnings and pursue their preferred outcome, will they lose credibility with their group?  If a group follows [even accurate/good/other word] advice to invest but loses money, will they listen to the advisor in the future or will they trigger the punishment stage where they ignore advice?

What we find is interesting.  The most successful advisors in terms of maintaining credibility are those who take a cautious approach, using the investor-preferred equilibrium.  Advisors who take a more aggressive approach quickly lose their credibility and, critically, almost never manage to regain it once lost.  Because of this, the cautious advisors also make significantly higher payoffs than their aggressive counterparts, by regularly making recommendations that hurt their immediate earnings. 

So what can we learn from these results that can give insight into environments outside the lab?  It is common for informational credibility to take precedence in management, and a manager’s ability can often be interpreted as their strategic competence.  However, we find that in cases involving the need to coordinate behavior of followers, social credibility can be even more critical for success.  This is not simply a way to manage people who make mistakes in their job; social credibility is an equilibrium phenomenon

We are certainly not the first to show that managers must keep their workers happy, but this is often discussed as a compromise or deviation from what is best for the organization.  We show that such behavior is in fact directly in line with an organization’s best interest in a large number of cases. It is rational to prioritize a leader’s social credibility, and such abilities deserve equal recognition in evaluating a manager’s effectiveness.  A government leader’s knowledge of economic conditions will prove meaningless if they cannot convince independent firms to invest locally.  Similarly, an executive’s analytical brilliance or creative insight into consumer markets may not lead to business success unless they are able to maintain followership among employees in many areas of their firm.

 

reagissez !

Ajouter un commentaire