There is something freeing about being at the end of one’s academic career. Or so I infer from the widely acknowledged but little discussed phenomenon of emeritus professors writing bad books on big subjects. It is freedom from engagement with the academic literature. Freedom to sketch a big thesis with impressionistic vagueness, and then not bother to defend it in any detail. Freedom to acknowledge that the middle third of your book is basically a Wikipedia deep dive, peppered with some thoughts you had while reading. Couldn’t be me – I’d be out of a job.
Anyway, I recently read Vito Tanzi’s book, Fragile Futures: The Uncertain Economics of Disasters, Pandemics, and Climate Change. Tanzi, the book’s jacket tells me, was Professor of Economics at the American University and Director of the Fiscal Affairs Department at the IMF from 1981–2000. He is a macroeconomist, and did important work on public finance, taxation, and the underground economy to name but a few areas. In this book, he turns his attention to disasters. The rough thesis of the book is that we’re bad at handling disasters because of economics’ failure to account for Knightian uncertainty. This is an interesting idea, and I took on this review because I wanted to learn about how macroeconomics handles (or doesn’t handle) severe uncertainty, and to see the argument linking this to failed disaster management.
The book has a few quirks which I found oddly charming, although perhaps this reaction was a self-defence mechanism. First there are the references to Wikipedia, a resource I love but which, as we tell our undergraduates, should be the start rather than the end of the research process. Second, there is the chapter promising a taxonomy of risks, which is 4 pages long and, after loosely observing some differences between kinds of disasters, fails to provide a taxonomy. It also fails to feature in the subsequent discussion, despite the clear differences between climate change and other disasters. Finally, there’s the lack of any interest in modern economic literature on uncertainty, supposedly a core topic in the book. Oh, there are references to Knight and Keynes, and a few economic chestnuts (Akerlof, Coase) are cited. As far as macro goes, for all I know there could be no literature to cite, but Tanzi also doesn’t deign to discuss the large microeconomics and decision theory literature on the various forms of severe uncertainty.
The problem with Fragile Futures is not that there are no interesting ideas. On the contrary, it contains a few very interesting ideas, including a proposal about tax increases in the immediate aftermath of crises. The problem is that these ideas are poorly delivered, barely linked back to the central problem or explanatory hypothesis, and devoid of academic context.
The most interesting parts of the book are the introduction (Chapter 1) and conclusion (Chapters 15 and 16). These chapters sketch out how Tanzi thinks disasters interact with public finances and the constraints of democracy, and they introduce his proposals. There aren’t really arguments here (nor would one expect them in these parts of a book), but the ideas are intriguing. It is all too brief however, and almost seemed to hint at a different, better book that Tanzi might have written.
There are more glimmers of this book that could have been in Part I. Chapter 2 introduces the distinction between risk and uncertainty. It should set up the book’s main argument, but as I’ll discuss below, that never really gets going. Instead, this chapter shows many of the book’s flaws. It is messily structured; for example, it can’t quite decide when it is going to introduce Frank Knight – mentioning him and his 1921 book in asides three or four times before actually introducing his work. It jumps from a discussion of Keynes and uncertainty to one of the subprime mortgage crisis, and then back to the 1920s to discuss Knight. It notes that there are some relevant economic works about uncertainty, cites them, but then moves on without either discussion or explanation of how their contents interact with this book’s. The rest of Part I is similar. Chapter 3 presents the unused ‘taxonomy’. Chapter 4 gestures at some links between democracy, capitalism and uncertain events, but the material is vague and doesn’t amount to a case for the ideas in the introduction or conclusion.Footnote 1
The largest part of the book (Part II) consists of five chapters recounting different disasters and their impacts: diseases, famines, natural catastrophes, nuclear power-related disasters and industrial disasters. The purpose of these long-winded chapters is, apparently, to ‘call attention to the very limited role that governments have played in the past in dealing with exceptional situations … [such as] pandemics and other major disasters’ (9). This is a ridiculous claim to make to a reader in 2023, having just come through a pandemic characterized by major government interventions, including the subsidization of the vaccines which were instrumental in bringing it to its official end. Nevertheless, it is reasonable to argue that governments should do more, and Tanzi has a few suggestions in later chapters for what specifically they should do. But for the reader who trudges through Chapters 5–9, what is on offer is a summary of various disasters, supplemented with brief commentary on the way in which governments failed, markets failed or democracy failed.
Chapter 10 promises to bring together the 90-odd pages of dreary disaster description and present ‘Guiding economic principles for disasters’. This chapter restates the case made in the Introduction, that democratic political systems are short-termist and that disaster preparedness looks like wasted expenditure. This is explained by way of some light speculation. First, Tanzi wonders whether there is an ‘irrational human tendency to simply ignore’ uncertain, non-quantifiable threats. Second, he muses about whether ‘the attention given to the short run may have been influenced by the fact that the economic principles that were developed in the past came about when the average life expectancy of humans was relatively low, say thirty to forty years’. (Given European trends in life expectancy, and some other comments Tanzi makes, I think he has in mind here the economics of Hume and Smith.) Finally, we get a bit of garbled population ethics, in which Tanzi speculates, rather circularly, about our concern for future generations: ‘the more attention a country’s population pays to the individuals who are currently living, and especially to the individual liberties of those individuals, as compared to the interests or welfare of the whole community and of future generations, the less concern there will be, among its members, about the future and about the individuals of the generations that will follow’ (129).
Part III is about climate change. It contains a lengthy discussion of the origins of climate change, its political dynamics, and some basic climate economics. As material on this subject for a wide audience goes, it is actually pretty good, although it is excessively historical and US-focused for my tastes. It fits oddly in the book, however. Climate change is unlike the other disasters in several respects. Some are discussed: such as its wide geographic scope and its need for collective action. Others are not sufficiently reflected upon: climate change is a slow-motion catastrophe, that we can prevent, but one whose harms affect mostly people who do not yet exist. It isn’t obvious to me that the same tools which could enable states to plan for oil spills or tsunamis would enable them to deal with climate change. Nor is it obvious that the democratic challenge of planning for hard-to-quantify but recurring disasters such as earthquakes has much in common with the intergenerational justice challenge posed by climate change.
Uncertainty, Economics and Disasters
The idea that drew me to the book – that our failure to manage disasters is linked to economics’ difficulty with Knightian uncertainty – never really comes to the fore. The promise of the book’s jacket and Introduction is that it is going to be the major explanatory hypothesis. But once Tanzi gets going, it is clear that the role uncertainty plays is much vaguer. Two other features are incorporated into the explanation early on: the short-termism of democratic electorates and the free-market fundamentalism that has dominated public finance theory. The latter is sometimes equated by Tanzi with economics’ failure to theorize severe uncertainty, but this is neither explained clearly nor argued for. In Part II, over-reliance on technology also features as an explanation of various failures.
That explanatory hypothesis is so interesting because Knightian uncertainty is genuinely hard to account for. Named for Frank Knight, this term refers to uncertainty which cannot be quantified statistically, ‘using available, real-life observations’ (19). Keynes’ examples are illustrative:
The game of roulette is not subject, in this sense, to uncertainty; nor is the prospect of a Victory bond being drawn. … The sense in which I am using the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention, or the position of private wealth owners in the social system in 1970. About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know. (Keynes Reference Keynes1937: 213–214)
Tanzi asserts that ‘uncertain events had attracted little, if any, attention because they were difficult to anticipate and to prepare for’ (6) and then clarifies with ‘Generally, uncertain future events, until they become a reality, have not influenced the current tax and public spending levels of governments, and have had little influence on the regulatory budget’ (6). This move between what is included in economic theory and what is done in practice is characteristic of Tanzi and reflects, perhaps, his IMF background. Regarding theory, Tanzi writes that in normal times ‘the tax level and tax structure have been assumed to be those consistent with and sufficient to satisfy the normal requirements of citizens and enterprises in efficient markets in a democratic political system. … Taxes, therefore, are expected to be efficient in their long-run impact, and sufficient to satisfy the routine and immediate needs of citizens’ (5). This thinking is digested in the form of fiscal rules, also targeted at the routine operation of the state. The problem, claims Tanzi, is that public finance theory and the rules which summarize it have simply neglected the Knightian uncertainty which characterizes disasters.
This neglect of uncertainty is reinforced, in Tanzi’s eyes, by the short-termism of democratic governments and the tendency of voters to punish governments for wasteful spending which, he asserts, disaster-preparedness would surely be regarded as. There’s something to this idea, but Tanzi is no political scientist and doesn’t attempt to justify his claims with evidence, nor to differentiate between political cultures and systems. I wondered what he would make of Sweden’s Civil Contingencies Agency, with its network of bunkers sufficient to house 70% of the population in case of war or disaster. (Sweden was last engaged in a war in 1814.)
Tanzi’s proposals for fixing this situation are fourfold: (1) macroeconomics and public finance theory should incorporate Knightian uncertainty; (2) taxes should be temporarily increased, sharply and progressively, to fund disaster relief; (3) quantitative easing should be deployed; and (4) broader global governance should be developed to tackle climate change amongst other disasters. There is also some vague gesturing at the idea of rainy-day funds.
There is no real proposal for (1); (2) and (3) are sensible but they are of course reactive rather than proactive moves to plan for disasters; and (4) is hardly a novel insight. Rainy-day funds are exactly the kind of thing Tanzi seems to think democratic electorates won’t tolerate. At this point in the book I wasn’t surprised that Tanzi didn’t engage with the literature on democracy, global institutions and climate change which might have offered exactly the resources needed to develop some of these ideas (e.g. Caney Reference Caney2022).
The vagueness of the first proposal, about macroeconomic theory, was the most frustrating to me. To begin with, Tanzi doesn’t provide a compelling diagnosis of the problem with current economics. Instead of providing the relevant details of public finance theory, or a discussion of how macroeconomics handles uncertainty generally, he simply states that Knightian uncertainty is entirely neglected. This comes via a mixture of assertion (with a ‘trust me on this’ tone) and some quick citations without explanation, e.g. ‘the expected behaviour of the government, as … established by public finance theory (see e.g. Musgrave’s influential 1959 book), … ignored the possibility of uncertain or unexpected future developments and needs’ (5).
If Tanzi was not going to discuss how macroeconomics and public finance theory treats uncertainty, he might at least have discussed how decision theory and environmental economics handle it. Caution, robustness and the precautionary principle could all have featured heavily in a book about the ‘uncertain economics of disasters’. So too could concepts such as ambiguity, ignorance and awareness – to give shape to the problem, if nothing else. But they do not feature. Tanzi calls for a macroeconomics of uncertainty, but when the time comes to suggest what this might entail, he offers little more than a shrug.
Perhaps practice can help where theory does not. Fragile Futures does have some discussion of the way that private individuals and firms manage risk, that being insurance. It was thus strange to me that the book contains so little about natural catastrophe insurance, a $2.5bn market dedicated to uncertain events.Footnote 2 The insurance industry even underwrites contracts covering terrorism risk,Footnote 3 which comes perilously close to Keynes’ original example of uncertainty concerning war in Europe. All Tanzi offers in this regard is a short discussion of a Swiss Re report on their attempt to create markets for volcano risk. What he concludes from it is that ‘the premium was not based on what could be defined as statistical risk, but largely on information-supported intuition, on information about the future occurrences of a highly uncertain event, such as volcanic eruption. Therefore, the premium did not reflect and could not reflect an optimal price in a well-working market. It could not make the market more perfect’ (22). The last line is more dogma than analysis and does little to confront the realities of decision-making in uncertain environments.
One topic ripe for exploration is that the evidence we have about disasters differs quite significantly across the categories that Tanzi considers (and some others which he doesn’t, but which have insurance markets today). Volcanoes and hurricanes have limited historical records, but well-understood physical bases which support some attempts to model their frequency. Nuclear disasters have exceedingly little frequency data – even in this company – and very little understanding of the potential extent of the harms, but do have a lot of engineering understanding of modes of failure. Terrorism, by contrast, has complex social causes which make its prediction an entirely different exercise. Famine sits somewhere in between: part social, part climatological; with some data on both fronts but which I imagine are difficult to amalgamate into a predictive model. Despite the long-winded discussion of disasters, Tanzi doesn’t make anything of this difference. He also doesn’t reflect much on the role of the government as opposed to private insurance. What should we think about the fact that hurricane risk in Florida is a private insurance market, but flood risk is underwritten by the US Government, through FEMA? What should we make of the British government’s recent attempt to foster a private terrorism reinsurance market? Does the fact that these are uncertain rather than risky events imply something about where they ought to be handled? For answers to these and my other questions, we must look elsewhere.
In summary, Fragile Futures is a disappointment. While non-academic readers might benefit from the chapters on climate change, they will gain little in the way of understanding how we tackle it or other disasters. For readers of this journal, the book is best skipped.
Joe Roussos is a researcher at the Institute for Futures Studies, where he leads the project Expertise in crises. He works on the management of severe uncertainty, especially in situations where policymakers aim to use uncertain science to inform their decisions. His work combines decision theory, philosophy of science, and social epistemology, and aims to be relevant and accessible to policymakers.
