Prosocial bonuses lead to higher performance: a case of circumvision beating supervision

I have examined the issues surrounding work disengagement in numerous posts in recent months (see Engagement requires happinessRemote workers are more engaged?There’s something in the airDig your own hole, sharpen your own shovel, and The way to engage employees is to disengage management), and looked into the idea of basing companies around the psychology and techniques of high performers, who are generally very engaged in their work (see What top performers do, and how to do it, and What do Amazon and Netflix have in common?). What I haven’t written about much is the role that bonuses play — or don’t — in engagement.

Recent research by Lalin Anik (who is a post doc at the Center for Advanced Hindsight) and a group of colleagues examines the impact of prosocial bonuses, where individuals are given bonus money to give to others, and contrasts that with more conventional team and individual bonuses.

The TL;DR version: social bonuses can be an effective stimulus to better performance.

As the researchers noted, when asked why they work most people say they work for money. While pay bonuses has been correlated with better performance, large bonuses are divisive and can decrease group productivity as a result of decreases in trust, cooperation, and other negative impacts. One counter is to institute group or team bonuses, which intuitive would lead to greater social cohesion and cooperative behaviors. However, group bonuses have their own downsides, as people can become disenchanted with those they consider ‘freeloaders’, or because of perceived inequities. This means that simply treating a collection of people as a team doesn’t lead to high performance. (Note that this corroborates the ideas explored in Voluntarily formed teams perform better than alternatives, where individuals forced to perform as teams do less well than teams the form voluntarily).

Is there a third way to reward performance — or good behavior — through bonuses?

The researchers ran two studies:

  1. The researchers looked at the impact of awarding bonuses to employees of an Australian bank, but which the employees had to reallocate to charities, and retaining none of the money personally. Groups had either a $100 or $50 charity voucher to distribute, or no bonus. Job satisfaction was measured immediately after the award period.
  2. In a second study the researchers distributed bonus money across sales and sports teams, and directing the participants to use the money to reward others for good behavior, buying presents for them in effect, and contrasted that with direct bonuses to participants to spend on themselves. Team performance was measured at the start and a month after the prosocial bonuses were introduced.

The findings?

  1. The bank employees that got the $100 charity voucher showed a significantly higher level of happiness and job satisfaction than others.
  2. There was a significant effect from prosocial bonuses in the sports and sales teams, and were especially effective for sales teams, which showed large and significant increase in performance during the testing period, but less so in the sports situation. A $10 bonus, when passed along to other, led to an average return of increased sales of $52.

The second study is very interesting to me, because it support the argument I have made for the transition from ‘supervising’ to ‘circumvising’: instead of a one-to-one relationship with a manager, who decides your bonus from a pool they control, instead there is a transfer of that role to the network of people that an individual works with.

As I wrote earlier this year in Why are disengaged employees disengaged?,

My bet is that as companies become more fast and loose with the transition to social business models of operations, employee engagement may decrease in criticality. As employees build their own support and trust networks, and have more say in what they do and how they do it, the role of management will shift. In a sense, the relationship between the employee and their ‘supervisor’ will become less central to the perception of the company and the job, because ‘supervising’ will gradually be replaced with ‘circumvising’. Instead of a manager you report up to and who directs the work of those below, the social context — social practices, social tools, and the immediate social network — will constrain and support the worker from all around.

The best answer to disengagement is not trying to train first line managers to be more caring. That’s unlikely to work, or to scale. We should instead move steadily into building circumvision as a characteristic of social contexts, where engagement emerges from strong caring relationships with many colleagues, instead centralizing that into one supervisor.

The success that Anik Lalin and her colleagues had in exploring the benefits of prosocial bonuses demonstrated the untapped power of circumvision.

The takeaway is to think through the implications: the role of the leader is increasingly not to decide who should be rewarded for their individual contributions, or to force people into potentially inequitable and unwanted group rewards. Both have been shown to ‘work’ but the negative impacts can lead to decreases in engagement and social cohesion. The third way is prosocial, bottom-up, distributed rewards distribution: where social relation and cooperation becomes the context and the participants act as circumvisors, rewarding the best behaviors and squeezing out the causes of disengagement.