Using the workplace as an example, consider someone called Jane who interacts with different people over time on collaborative projects. For instance, this week she is working on a project with Sam, next week she is going to work on a project with David, the week after a project with Sam, and so on. The question of interest is whether her experience, say, working with Sam influences how she behaves when working with David? This, in turn, gives us some idea of how norms can emerge and evolve within a particular workplace. Can, for example, one slacker ruin productivity across a whole firm?
To make things more concrete suppose that the basic choice Jane has to make is how much effort to exert on a project. She can cooperate or slack. We can then think of a project as either a public good game or minimum effort game. In both games the best outcome for the group is mutual cooperation. The differences lie in individual incentives. In a public good game Jane maximizes her material payoff by slacking and so working hard basically requires her to be 'nice' (or forward looking to potential future gains). In the minimum effort game it is in Jane's interests to work hard if others on the project work hard and so the issue is more one of trust in others. We can now re-frame our question: Does bad experience on a project lower willingness to be nice and trust others on unrelated projects?
For insight on this we can look to a 2013 paper by Armin Falk, Urs Fischbacher and Simon Gachter published in Economic Inquiry on 'Living in two neighborhoods: Social interaction effects in the laboratory'. They randomly put subjects into a matching group of nine people. You can think of this as a workplace. A subject then interacted over time with two distinct groups of people within the matching group, as illustrated below. Here Jane is involved on a project with David and Robert and a distinct project with Sam and Freya. This set up allows us to nicely see if Jane's experience with, say, Sam influences how she behaves with David. The study considered both a public good game and minimum effort game.
The main finding of the study was that subjects behavior within groups converged but not across groups. So, for example, Jane might converge on an equilibrium of high effort with Sam and Freya and one of slacking with David and Robert. This convergence within group suggests that people can freely adapt their behavior to the environment - cooperating with and trusting some colleagues and not others. This, in turn, suggests that contagion across the workplace may not be as pronounced as some would expect. If, for instance, David is a slacker the consequences can be contained to the projects he is involved with.
In a recent study with Thomas Singh published in the Journal of Behavioral and Experimental Economics entitled 'Observation and contagion effects in cooperation' we find similar results. In our study subjects were in a matching group of 4 and paired off to play with each other. Again, we found that people were happy to adapt their behaviour to the person they were paired with - cooperating with one person and slacking with another. This, again, meant that behavior within group converged but not across groups.
We, though, added an extra factor into the mix by having treatments where subjects could see what others were doing. So, for instance, Jane can see what is happening on the projects Graham is involved with even though she does not directly interact with him. In the public good game this had no effect at all. In the minimum effort game, by contrast, it significantly increased rates of cooperation. Which is a good thing. Our design did not allow us to precisely disentangle why we get such an effect but we put it down to the cooperation of one group providing a focal point for others. Essentially, if Sam and Jane see Graham and David succeeding then they can more easily coordinate themselves.
So, what are the lessons we can take from this? The results of these two studies and others suggest behaviour need not be contagious across groups. So, one slacker is not going to infect the whole workplace. But, equally, one hard worker is not going to dramatically boost the workplace. Observing others may, though, have a positive benefit. This is not so much conformity or 'blind' learning but more that success provides a focal point that can help others coordinate around. Let me finish by noting that a lot of the studies on networks (both theoretical and empirical) constrain a person to choose the same action in all interactions. So, Jane has to decide either to cooperate on all projects or slack on all projects. You can see that this is quite a strong (and potentially unrealistic) assumption.
For insight on this we can look to a 2013 paper by Armin Falk, Urs Fischbacher and Simon Gachter published in Economic Inquiry on 'Living in two neighborhoods: Social interaction effects in the laboratory'. They randomly put subjects into a matching group of nine people. You can think of this as a workplace. A subject then interacted over time with two distinct groups of people within the matching group, as illustrated below. Here Jane is involved on a project with David and Robert and a distinct project with Sam and Freya. This set up allows us to nicely see if Jane's experience with, say, Sam influences how she behaves with David. The study considered both a public good game and minimum effort game.
The main finding of the study was that subjects behavior within groups converged but not across groups. So, for example, Jane might converge on an equilibrium of high effort with Sam and Freya and one of slacking with David and Robert. This convergence within group suggests that people can freely adapt their behavior to the environment - cooperating with and trusting some colleagues and not others. This, in turn, suggests that contagion across the workplace may not be as pronounced as some would expect. If, for instance, David is a slacker the consequences can be contained to the projects he is involved with.
In a recent study with Thomas Singh published in the Journal of Behavioral and Experimental Economics entitled 'Observation and contagion effects in cooperation' we find similar results. In our study subjects were in a matching group of 4 and paired off to play with each other. Again, we found that people were happy to adapt their behaviour to the person they were paired with - cooperating with one person and slacking with another. This, again, meant that behavior within group converged but not across groups.
So, what are the lessons we can take from this? The results of these two studies and others suggest behaviour need not be contagious across groups. So, one slacker is not going to infect the whole workplace. But, equally, one hard worker is not going to dramatically boost the workplace. Observing others may, though, have a positive benefit. This is not so much conformity or 'blind' learning but more that success provides a focal point that can help others coordinate around. Let me finish by noting that a lot of the studies on networks (both theoretical and empirical) constrain a person to choose the same action in all interactions. So, Jane has to decide either to cooperate on all projects or slack on all projects. You can see that this is quite a strong (and potentially unrealistic) assumption.
Comments
Post a Comment