By Greg Fisher
In a blog on Keynes and Hayek I mentioned that I viewed recessions as collective action problems. In this blog I want to expand on what I mean by this because it makes all the difference for economic policy. It also contextualises our conventional demand management approaches, namely fiscal and monetary policy.
To build up the hypothesis, it is necessary to dip our toes in to a number of fields of study. But let us start with some empirical evidence and build the conceptual framework from there.
In his study of decision-making in the financial system, set out in his book Minding the Markets, Prof. David Tuckett from UCL interviewed 50 investment managers and analysed how they made decisions under conditions of uncertainty. Here I will take Prof. Tuckett’s conclusions and apply them to the phenomenon of recessions.
Prof. Tuckett explained how investment managers used narratives to help them make sense of the past, present and future. The amount and complexity of the information and the high degree of uncertainty they have to deal with can be staggering. So they develop narratives – mostly sub-consciously – to form what seems to be a cohesive pattern of understanding. This is also true of people in everyday life.
This is very different to conventional economics where “rational expectations” dominate. In this conventional approach, people are typically fully informed, have a particular (unchanging) model of how the economy works, and use these to determine their expectation of some variable in some fully predictable future.
By contrast, in reality people use narratives – again, subconsciously most of the time – in complex environments and these narratives can include some expectation of the future. People also typically have a sub-set of the information required to make decisions in their daily lives and, in complex social systems, the future is inherently uncertain.
Narratives form an important part of our framing of some situation or object but that is not to say they displace conscious analyses of the economy. The two are not mutually exclusive: narrative formation can involve a complex interplay of the conscious and sub-conscious aspects of the brain.
Narratives are also formed in part through social interaction. We listen to what others have to say: they influence us and we influence them, both at the same time. This important point means we are compelled to think of social systems as “Complex” (meant in the formal, Complexity theory sense), with people continuously interacting and co-evolving. The concepts of emergence and global cascades are useful for understanding how narratives are “exchanged”, how they spread, and how a particular narrative might become a “consensus” view among a group of people.
With these micro and system-wide points in mind, my argument is that recessions come about when a dominant narrative emerges within society, leading to behaviour that is consistent with a recession subsequently arising. If people believe a recession is likely, or imminent, they tend to behave “prudently” with respect to both consumption and investment. An important point is that the narrative of a recession and how people respond to that narrative make a recession both inevitable and self-fulfilling. In more technical language, there is “time-consistency” between the narrative-expectation and people’s behaviour.
An important difference between a recession narrative, which leads to a recession, and traditional theory is that in the latter, an economy is typically expected to gravitate toward some future full-employment equilibrium. For example, the models used by central banks (the terribly-named Dynamic Stochastic General Equilibrium models) normally show economies necessarily recovering following a recession. But if the consensus narrative is one of recession, and remains there, it is plausible that an economy can remain in that state, i.e. a “depression” narrative could emerge, reinforcing the slump. Alternatively put, a depression could be viewed as a sub-optimal equilibrium in a complex system.
I refer to recessions and depressions as collective action problems because they arise out of the collective action of agents in the system, operating in a way that is consistent with self-interest but which result in outcomes that are detrimental to all. An analogy is the Prisoner’s Dilemma game in game theory. As is famously known, the outcome in the Prisoner’s Dilemma is sub-optimal for both prisoners. If they could collude (a form of collective action), they could achieve a different outcome that is preferable for both. But this outcome is contingent upon the collusion being viewed as credible by both prisoners: some mechanism is required.
Now let’s turn to the policy implications of treating a recession as a collective action problem. In traditional macroeconomics, the mechanisms used to mitigate recessions (fiscal and monetary policy) are viewed as managing the overall level of demand in the economy. But if the narrative approach above is accurate, are these traditional tools of demand management sufficient for bringing about a recovery? Not necessarily.
What is crucial is the impact these policies will have on people’s narratives about the economy, not only their impact on “aggregate demand”. How these changed narratives influence individuals’ behaviour is another important question. A key point is that the overall impact of people’s changing narratives concerning an economy can dwarf demand management policies. A second key point is that narratives seem to be formed through an emergent process, which means prediction and control are highly problematic.
It is for this reason that I am sceptical of orthodox Keynesian approaches that imply a mechanistic view of the economy, whereby people adjust – deterministically – to macro management policies. Narrative formation is much more complex than that. Indeed, my colleague Paul Ormerod noted in November that the US appeared to be enjoying an expansionary fiscal contraction. This is possible if narratives change in favour of a recovery, leading to recovery-consistent behaviour by individuals, despite a fiscal contraction.
However – and this is a big however – demand management policies can and do influence narratives. But they should be viewed as one of a complex set of influences on how people view the economy, including its future.
Before concluding, it is worth noting that these points relate to the fiscal policy debates of 2008-2010. During those debates the conventional Keynesian and Conservative perspectives were wheeled out. Keynesians said the fiscal deficit had to be expanded to counter a contraction in private demand; and Conservatives emphasised prudence to inspire “confidence”. A narrative-based framing shows that the Conservative perspective was not as unreasonable as Keynesians argued: we can re-state “confidence” through a narrative framing in stating that a prudent approach might inspire a shift away from a recession narrative to something more optimistic. But there is also a reasonable argument that a fiscal expansion might have inspired a recovery narrative. We will never know. What is clear is that we need to understand better how narratives emerge and perpetuate and how they can be influenced, including (if at all) by governments.
To conclude, I suspect that to those people not trained in conventional economics, this all sounds blindingly obvious. I would like to think that is because I started with a look at research based on empirical evidence, albeit based on the world of finance; and because I mixed this evidence with appropriate and cutting edge concepts from the new field of Complexity theory. But a lot more work needs to be done to deepen this framework, including by the academic community.