John Kay and Mervyn King talk about their book, Radical Uncertainty. This is a wide-ranging discussion based on the book looking at rationality, decision-making under uncertainty, and the economists’ view of the world.
These notes have been lightly edited.
-
(4:30) Human urge to quantify and seeking precision; experts pretending to know more than they do to expand and maintain influence; most big decisions are one-off and unique and need to be deliberated and dig deeper
-
(7:20) Case of deflecting responsibility in scientific and policy making - have a single decision maker, accountability
-
(9:10) Models to organise collective thinking and form narrative - Helpful in teaching; not helpful in making quantitative predictions because of black box parameters
-
(10:20) Use of point estimates in policy making
-
(15:30) Use of cost-benefit analysis is a way to organise thoughts and setting a strategy; Wrong to quantify the net benefit at the end; value in unearthing drivers that are certain/uncertain.
-
(18:00) Value at risk - Is there a better alternative? The debate of Best metric vs Best managerial response; Relying on arbitrary figures as regulatory benchmarks and not questioning the scenarios in-depth is the mistake
-
(23:00) In light of the bailouts that happen during crisis, do companies have incentives to be prudent? [To be detailed]
-
(25:00) What is wrong with the current usage of rationality in economics and specifically behavioral economics? Rationality axioms (defined by Econs) plays well in a controlled small-world models but not appropriate in a large-world models abound in radical uncertainties; Biases are evolutionary adaptations to an uncertain world; Decisions in government, business and personal life - are not characterised by small-world settings.
-
(30:50) What has gone wrong with Behavioral Economics? Behavioral Economics Research shifting from Models not conforming with People (work of Maurice Allais, Daniel Ellsberg) to People not conforming with Models (work of Daniel Kahneman, Amos Tversky); Expected Utility - [To be detailed]
-
(41:00) Under what circumstances can a market economy be efficient? [Arrow-Debreu Model] Can the market set be complete ? i.e markets for buying and selling of all contingent outcomes in the future? The (wrong) view that the world would be perfect if there existed a market for everything and then the market economy could self-correct -> more and more financial instruments will complete the set of AD markets; Because the assumptions of AD model did not hold due to imperfect market with asymmetric outcomes.
-
(43:41) Human intelligence designed for large worlds and effective at dealing with uncertainty. Assumption of biases is implausible in light of the evolutionary origins of cognitive ability; Business owners are not optimising and maximising in real life (because information is absent); rather they are finding strategies that are ‘good enough’; Sequential vs. simultaneous option comparison made by decision makers.
-
(46:15) Fast-and-frugal decision making and power of rules-of-thumb in crisis environment by Gerd Gigerenzer vs. simulated setting; Computers are great at solving well-defined, scoped problems at lightning speed whereas human minds have the capacity of making leaps of imagination to deal and solve ill-defined problems
-
(49:56) Decisions/reality as opaque; Do humans have a natural inclination to think of (ill-defined) mysteries as (well-defined) puzzles; or do we despise certainty (groundhog day) and like uncertainty (elements of surprise) in everyday life?
-
(52:21) Managing risk and embracing uncertainty; Keynes and Knight called out the distinction between risk and uncertainty.
-
(56:40) What is wrong with the as-if assumptions made by Chicago school of thought; equating economics to physics as a field with equations and data but not applying the self-correcting behaviour of improving knowledge (found in physics).
-
(1:03:00) Fallacy of dropping as-if assumptions after model is confirmed and not going back and revising/ is prevalent in risk management, policy making.
-
(1:06:40) Robert Skidelski quote - economics is not a progressive science; state of deductive (lot), inductive (little), abductive (very little) reasoning applied in economics. [Charles Sanders Peirce work on reasoning]; deductive and inductive do not have the tools to provide advice to business, policy and society; decisions in the real world involve a significant degree of abductive reasoning.