25 March 2019

Uncertain Futures

Edited by Jens Beckert and Richard Bronk
2018, Oxford University Press, 333 pages,
ISBN 9780198846802

Reviewer: William A Allen, Visitor, National Institute of Economic & Social Research

At least since 1921, when Frank Knight’s Risk, Uncertainty and Profit was published, economists have understood that there is a distinction between risk, which can be measured by a probability distribution, and radical uncertainty, which cannot. Somewhat more recently, the British economist George Shackle devoted his career to exploring the implications of radical uncertainty. While economists have no doubt understood the concept, it has not been in their interests to take it too seriously, because that would prevent them from making forecasts or policy recommendations, and thereby from earning a living. And, it is fair to say, economists are of limited social value if, when faced with practical questions, they can only shrug their shoulders and mutter about radical uncertainty. Nevertheless, the fact that the future is unimaginable can never be ignored.

Uncertain Futures is a stimulating and diverse collection of papers about the consequences of radical uncertainty and how they are managed in practice. As the clear and comprehensive introduction by the editors explains, devices such as narratives, stories, conversations and ‘imaginaries’ give shape to expectations of the future. David Tuckett, in his chapter, dignifies the process with the title Conviction Narrative Theory.

Economic forecasts are a kind of narrative. Usually they are based on models, together with specified assumptions about variables deemed to be ‘exogenous’ – in other words, determined outside the model. But our ignorance is such that in many cases, we simply don’t know how so-called exogenous variables are determined. And the behavioural assumptions represented by coefficients in equations may turn out to be mistaken. The forecasts are accordingly unreliable. Nevertheless, decisions have to be made in the light of some kind of vision of the future; therefore forecasts have to be made, implicitly if not explicitly. It follows that from time to time, forecasting models will cease to be useful because something important has happened which they could not explain. Such an event occurred in 2008.

In such circumstances, decision makers need to be ready to jettison outmoded models ruthlessly. The complexity of modern modelling techniques makes this difficult. Central banks and governments are large employers of economists, and ambitious economics graduates will be anxious not to damage their career prospects. So they will invest considerable time and effort into understanding the models that are currently in use. And, having made the investment, they will be reluctant to abandon it. It took a long time in the 1960s and 1970s for economists in the United Kingdom to liberate themselves from naïve Keynesian models which could not explain inflation, and one can only hope that it will not take them as long to escape from the constraining assumptions of dynamic stochastic general equilibrium models.

Werner Reichmann’s chapter on economic forecasting stresses the social relations among forecasters and concludes that those relations improve forecasters’ ability ‘to see the world through the eyes of others, understand their plans, and discover their influential expectations.’ Maybe, but perhaps discovering a new model to supplant an outmoded one is more easily achieved by a small group than by a large one, just as a small boat can turn around more quickly than an oil tanker. In a similar vein, Olivier Plimlis, discussing forecast errors, notes that forecasters prefer peer-group assessment to assessment by reference to forecast accuracy. In a more positive vein, Andrew Haldane suggests that agent-based models are the way forward, or at least one way forward.

There is a school of thought that believes that monetary policy should be governed by rules, rather than the discretionary decision-making authority of central banks or governments. The avoidance of discretion would mean that there was no need for unreliable models or forecasts. The problem is that no rule has survived the test of time. The Bank Charter Act of 1844, which established automatic control of the supply of currency in the United Kingdom, had within a few years manifestly failed in its objective of preventing financial crises, so that discretionary central banking had to be established. The revised version of the gold standard, which admitted discretionary central banking, though it lasted for about half a century, did not long survive the First World War. The Bretton Woods system lasted only for as long as it suited the United States. More recently, the Federal Reserve embraced a rule for monetary base control in 1979, but abandoned it in 1982 when it became inconvenient.

Narratives, in the form of forward guidance on interest rates, have a longer track record in monetary policy than is widely recognised. In May 1932, the Chancellor of the Exchequer, Neville Chamberlain, issued what would now be called forward guidance about his cheap money policy when he told the House of Commons that:

‘I do not see any reason why, in the absence of any serious disturbances in other countries with which we have close commercial relations, there should not be for some time to come a continuance of the present situation, in which money is both cheap and plentiful…’

The cheap money policy was an important stimulus to economic recovery in the 1930s. It included getting long-term interest rates down to 3½%, and in 1940, supported by extensive economic and financial controls, further forward guidance got long term interest rates down to 3%.

So far so good, but in 1945 the government tried to get long term interest rates down further, to 2½%. It was influenced by Keynes’ view – which might now be called a narrative - that the economy was close to satiation with capital, that investment was likely to be chronically low, and that a new great depression was likely to occur, possibly after a short post-war boom. The fundamental problem was that Keynes’ view was mistaken. The narrative failed to convince bond investors, who became increasingly concerned about inflation, so that long-term interest rates began to rise. Fortunately, there are limits to the power of narrative.

This might be called the first cycle of forward guidance; it is an example of the historical process described by Robert Boyer in his illuminating chapter, which shows how narratives are invented and pursued until a ‘structural crisis’ takes place and the search begins for new narratives.

Two chapters of the book are devoted to central banking. Douglas Holmes examines central banks’ communication with the public, but does not, to my mind, pay enough attention to the constraints that reality imposes on them; nor is there sufficient analysis of what happens when what the central bank has told the public turns out not to be true. Benjamin Braun points out, quite rightly, that the ‘unconventional’ policies of forward guidance and quantitative easing incidentally denude market interest rates of information content about market expectations. This represents a serious drawback, which in Braun’s view raises a question about whether the benefits of these policies exceed the costs. The disabling of the banking system in the financial crisis made some kind of policy intervention imperative; and all such interventions affect market prices and distort the signals they convey.

Nevertheless, it is not obvious that quantitative easing and forward guidance were the best interventions that could have been contrived.
Other chapters describe how radical uncertainty is managed in other fields: the staking of claims in the Arctic, the valuation of options, credit ratings, investment management, forest management and drug development and energy technology. One of the most interesting is Martin Girardeau’s account of the career of the venture capital investor Georges Doriot and his highly successful American Research and Development Corporation, founded in 1946. Doriot tried to assemble as much hard information as he could before making a decision, but recognised its inescapable incompleteness, and the importance of imagination and a sense of the future.

Among the most prominent architects of the future are the designers of large-scale infrastructure projects. For example, Britain’s railway pioneers, who were at their most active in the mid-19th century, have affected the lives of millions ever since. Yet in these days (and perhaps in those), it seems that the only way to get an infrastructure project to fruition is to make misleading but convincing forecasts of its costs and time to completion: cost and time overruns are all but universal.

Radical uncertainty is not a new concept, but it nevertheless receives less attention than it really deserves. ‘Uncertain futures’ is a very welcome and interesting antidote. It left this reader with an enhanced understanding of the expedients that are customarily used as means of either overcoming or else tacitly ignoring radical uncertainty, and of the dangers that a flawed but superficially persuasive narrative can wreak. Mindless positivity is just as dangerous as mindless negativity. For that alone, Uncertain Futures is well worth reading.