N Squared [Ormerod]
Azeem Azhar’s “The Exponential View” is one of my favorite weekend reads. A few weeks ago it contained a link to Amna Silim’s “What is New Economic Thinking?”. Buried in its reference section, I’ve found this gem by Paul Ormerod:
N squared stands for Nudging * Networks.
The first part focuses on “Nudging” — the kind of gentle behavior change interventions based on behavior science (rather than standard neo-classical economics) popularized in books such as Thaler and Sunstein’s Nudge.
The meat of the paper focuses on “Networks” and in particular, human networks. It explored the implications of a very interesting assumption:
“Many of the decisions we make are based not so much on the independent, rational calculation of the costs and benefits of different actions — the mode of behavior posited in economic theory — but on observing and copying others”
Ormerod provides a very easy-to-understand explanation of the behavior in a random structure network using the model developed by Duncan Watts.
The outcome of the model is best summarized by the following quote (and graph):
“Systems of interconnected agents whose behavior influences each other are both robust and fragile. These are key words. Most of the time, the system is robust to small disturbances and these do not spread very far. But occasionally, the system is fragile, vulnerable to exactly the same size of shock that usually it is able to contain…
… The common sense causal link between the size of an event and its eventual impact is broken. “
One of the immediate implications of that outcome is that:
“networked systems bring problems when it comes to measuring impact. What worked and what did not work? And why? A great deal of policy evaluation is carried out paying little or no attention to the potential impact of network effects. But if these effects are significant, studies that ignore them can generate misleading results. A successful outcome may arise not because of a nudge factor, but because of imitation across the network. The risk is that success can be mistakenly attributed and policymakers left puzzled when a similar policy leads to failure in a different context.”
As a final note, Ormerod introduces two additional network structure archetypes, “Scale-free” networks and “Small-world” networks, to show how an intervention strategy that works in one, may not work in another.
Good papers leave you with questions that you’ve never even considered before, that this was definitely the case for me here:
- What type of network archetype best describes companies?
- Does the archetype change as the company scales and the dynamics of the connection changes?
- What does 1+2 mean for corporate governance and more specifically, driving change inside companies?
- Given that many employees are tasked with enacting change in networks, how should the now common “accountability to impact/outcomes” mindset change?