In my last blog, I wrote about how economic theory needs to be ecologically valid – meaning that it needs to model agents who have the essential characteristics of real-world human actors and organizations, operating with their constraints in their environments.
Human agents are sentient and social and often have to deploy their cognitive skills in an environment that is radically uncertain.
This month I want to outline some features of a theory being developed at UCL which explores both how human actors take decisions when confronted with uncertainty and aims to examine the implications for how they coordinate to produce the macroeconomy.
“Human agents are sentient and social and often have to deploy their cognitive skills in an environment that is radically uncertain”
Uncertainty goes with different labels. Here, I simply have in mind those situations in which economic agents (individuals or organizations) have no means of knowing the outcome of their actions at the time they take them.
In standard macroeconomics, agents are modelled as able to calculate correct probabilities and expected utilities. They are, de facto, modelled as calculating machines operating with programmed utilities.
Humans do more than this. We are able to draw on a wider range of capacities than just the ability to calculate. We have bodies that feel and we can guess, imagine and hope. We copy and learn from those with whom we co-operate and those with whom we compete. It all provides many more capacities for coping with uncertainty than those available to machines.
Subjective perception and construction
In contrast with economics, a feature of much contemporary social and cognitive science is that the knowledge of the world and of the possible actions to take to influence it advantageously, are not given externally as objective fact. Perception and memory are complex processes so that we interpret or in a sense invent the world by creating narratives about it. And it all happens in a social context – so we can say usefully that it is constructed subjectively, not given.
A narrative is a powerful form of constructing the world recognised in social science, psychology and neuroscience as well as the humanities. We construct narratives because we have to organize and give meaning to the billions of stimuli we encounter in our daily lives. They help us decide what, among those, is important for our present purposes and also, by simulating how actions might have different imagined outcomes, to form judgements about the future. These judgements about future outcomes of an action are then used to support decision making as alternatives are compared against each other and the most preferable action is taken.
Because of their capacity to marry cognition with emotion, narratives are a special part of the cognitive armoury of human actors faced with uncertainty. Brains are not free-floating calculation machines. They are a set of evolved structures embedded in our bodies and acutely tuned for social co-operation and to maximise our chances of survival.
Emotions comprise vital and highly effective reward and punishment systems adapted to survival under uncertainty. The interaction of emotion and cognition in narratives allow us to “feel” confident about the narratives that support action.
In my group at UCL we refer to the narratives that underpin action in uncertainty as “conviction narratives”. They allow human actors to feel subjectively certain about the outcome of action to act. These narratives evoke emotions of approach (pleasure) and avoidance (displeasure) deep within our brain. It is as if, when we construct or hear narratives, we put our body out into the future, mentally time travelling to feel what it will really be like, if we take the action we are contemplating.
Shared Narratives and Economic Stability
Conviction narrative theory (CNT) provides an ecologically coherent set of microfoundations for how individual agents support action. How can an understanding of individual agents help us to understand macroeconomic outcomes?
The answer is that conviction narratives, although developed by individual agents, are constructed in a social context. This means that agents are influenced by what is influencing other agents and the stock of shared narratives circulating around them.
An economy enacted by economic actors using conviction narratives to cope with uncertainty is a very different one to the economy enacted by a representative agent.
For example, any equilibrium under CNT agents will be dynamic and fragile. CNT agents are social and create their own conviction narratives based around the conviction narratives around them – all assessed and used because they “feel” as well as think they are convincing. Because the underlying situation is uncertain, each individual agent’s conviction narrative could be highly sensitive to the conviction narratives of others, whether they are optimistic and excited or pessimistic and worried.
A good example of this is an ingenious bit of research published just after the Dotcom bubble. The share price of firms “which added ‘.com’ to their names in 1998 and 1999, regardless of other factors, exceeded those of a control group by 63% for the five days around the name-change announcement date, independent of a company’s actual level of involvement with the Internet.”
“Agents making decisions based on conviction narratives may revise expectation or actions very quickly”
So agents making decisions based on conviction narratives may revise expectation or actions very quickly – joining in optimistic risk-taking or rapidly responding to observed shifts in “animal spirits” such as losses of confidence created by events and/or new perceptions of old ones.
In my next blog, I will describe how these insights can be used to understand and further research the role narrative and emotion seem to play in the evolution of the economy and how we might forecast it.