The task we have set ourselves at Rebuilding Macroeconomics is to bring new ideas and research to macroeconomics and to see if they are any use.
My ideas come from a mix of old fashioned economics, sociology, psychoanalysis and two streams of empirical research – one concerned with whether and how patients and doctors understand each other and the other with how major asset managers make their decisions in financial markets.
Macroeconomists are trying to do two things. First, they want to have insights into how numerous economic actors combine together to produce the economic outcomes they do. Second, they want to offer solid advice to policy makers as to what will be the likely consequence of any interventions they are minded to try.
The general approach for the last fifty years has been (1) to treat an economy as a self-equilibrating system, and (2) to give agents the capacity for pretty much infinite foresight and insight into what they and everyone else are presently doing. Instability comes from outside the system; from what are called “shocks” – such as the invention of the Internet or a war interfering with oil supplies that nobody saw coming. As these shocks are fairly quickly understood in a shared and accurate way, the system soon returns to equilibrium with reasonably flexible markets.
Based on this sort of understanding the economy is believed mostly to operate in an optimal way and most of the ideas policy makers might get it into their heads to try to improve matters will just make things worse. An independent central bank is useful, however, insofar that it may adjust interest rates up or down and reduce cyclical volatility if markets are slow to respond and to anchor inflation expectations.
As many observers have suggested since 2008, these ideas look rather unconvincing. For a start, economic outcomes look far from satisfactory for most people and for human survival. If this is optimal, it is not being felt as acceptable. There is no clear interpretation of what is happening in the economy or what tools can influence its direction. In these circumstances there is a strong possibility that policy will be based on populism resulting from widespread dissatisfaction – but very likely to make matters worse.
From my perspective, modern macroeconomic models are problematic because they are unreal – not because they make necessary simplifications but because the features they abstract are grossly misleading.
First, models are made apparently blind to the reality of the decision-making context that faces economic agents. It is radically uncertain.
Second, they are made as if human agents are mostly efficient calculating machines, which is blind to (a) the fact that calculating machines can’t function if outcomes are radically uncertain, and (b) the weight of evidence about how agents actually make decisions in uncertain conditions, available from direct studies of economic and financial decision-making in organisations – for example Soros’ Alchemy of Finance, Bewley’s Why wages Don’t Fall in a Recession, and my own Minding the Markets.
That’s a sample of three but there have been many investigations and case studies of organisational functioning and fieldwork conducted in a wide variety of economic settings by sociologists and social anthropologists (see Guillen et al, 2003; Beckert and Aspers, 2011; Knorr Cetina and Preda, 2005) and by organizational theorists. Decisions are a result of complex innovative or organisational processes (for instance, Simon, 1946 (1997), Vuori and Huy, 2016; Sull and Eisenhardt, 2015: Lane and Maxfield, 2005). These studies tell the same sort of story. The conditions for action in the real world contradict those modelled in macroeconomics. No real-world study of this kind of economic decision making that I know concludes otherwise.
Radical uncertainty, in fact, is what most of today’s major companies and government agencies actually face. To take some topics recently presented to the UK Engineering and Physical Science CRUISSE network: how much money should a firm spend on cyber security? How should firms and governments mitigate the effects of changing weather and climate change? Which new technologies is it safe for the UK government to allow to be owned by foreign companies? What regulations should be applied to restrict future technology? What threats will exist and what armed services with what equipment are required ten years out? How do we design resilient infrastructure for the long term and how much to spend on it? And so on – without even beginning on BREXIT.
Outside the world of abstract academic modelling, no study of the activity of real world economic agents faced with the typical problems they actually face suggests they have any hope of calculating the expected probability of outcomes. As Mabel Berezin observed, rational choice, in the restricted economics sense, can apply only in those limited “small world” instances (such as a laboratory) where the choice context is stable or predetermined. Theories based on rational expectations, which require rational choices to co-ordinate according to the Von-Neumann Morgenstern axioms, necessarily recede before a great deal of empirical reality.
In 1955, Herbert Simon wrote that “broadly stated” the task of economics was “to replace the global rationality of economic man with a kind of rational behaviour that is compatible with the access to information and the computational capacities that are actually possessed by organisms, including man, in the kinds of environments in which such organisms exist.” (italics added).
The remarks in italics make clear he was referring not only to the computational or behavioural limitations of human actors. Rather, what he had in mind was his “scissors” analogy. On one blade is the structure of the context in which many decisions have to be made, the conditions for action which are often radically uncertainty, and on the other, is human capability, augmented nowadays by human invented machine support. How is it that 60 years later the message is still not received?
To become more secure, economic theory needs to be ecologically valid. The distinction rational and irrational used in economics is not valid in real world environments. There are no sound grounds for believing that in a human populated financial market ‘correct’ expectations could converge. Its not due to human limitation but to the context decision-makers face. For this reason (see also King, 2016) behavioural economics or finance based on the heuristics and biases movement – i.e. the study of human error in conditions of risk – is unlikely to make a significant contribution to macroeconomics.
Rather, the future of macroeconomics requires a major mental shift away from models based on normative rationality towards models based on incorporating the implications for their decision-making of the conditions of action that face economic agents. The shift might be, perhaps, “nudged” by institutional design, that is an organised shift in incentives implemented by research councils and macroeconomic consumers like central banks. If so, it could come about either by a movement towards investing in significant inter-disciplinary team-based research or by a substantial re-tooling of economic expertise and methodological sophistication.
You may also like…
17 October 2017