21 June 2004

The rise and rise of risk

Darryl S. L. Jarvis, The University of Sydney

Bent Flyvbjerg, Nils Bruzelius & Werner Rothengatter Megaprojects and Risk: An Anatomy of Ambition, Cambridge University Press, 2003 (pp 218). ISBN 0-52100-946-4 (paperback) RRP $49.95.

Cass R. Sunstein Risk and Reason: Safety, Law and the Environment, Cambridge University Press, 2002 (pp 352). ISBN 0-52179-199-5 (hardcover) RRP $79.95.

John Quiggin noted recently that risk to the 21st century is what globalisation was to the late 20th century (2003). From Nobel Laureates to the lay person, risk seems such a self-evidently important idiom that few question its utility. In the hard sciences risk is applied to earthquake prediction and preparedness, to metallurgy and materials stress testing, to hazardous waste materials handling, and to assessing threats to the solar system and asteroid collision with Earth. In applied settings, risk management is used in aircraft and automotive engineering, air-traffic management systems, food safety, and environmental protection, to name a few. Even in the realpolitik world risk has become an indispensable tool in defence planning, critical infrastructure reviews, terrorist threats, protection of combat personnel, and in strategic risk assessment in international relations (see Department of Defence, n.d.).

Risk is an outcomes driven analytical tool, providing researchers with measurable outcomes amenable to technocratic and professional application in both public and commercial settings. This proves attractive when research funding is driven increasingly by tangible results; valued not so much for its contribution to knowledge so much as its ability to address specific socio-economic needs. Risk thus helps discipline an otherwise unwieldy academy, directing energies into research with prescriptive and utilitarian outcomes.

Yet the story of risk is a little more complex than this simple utilitarian reading suggests. In fact, the story of risk is rooted much deeper in our cultural psyche than most risk practitioners would care to admit. This derives from more broadly based structural forces endemic in modern societies. Ulrich Beck (1992) was one of the first to note how increasing organisational and technological complexity is linked to increased risks in modern societies. Power failures on the Eastern seaboard of the United States during 2003 show how vulnerable complex modern systems are to simple malfunctions. Some sixteen million Americans and six million Canadians were blacked out by a relatively minor malfunction in inter-connected power grids stretching from Ontario, Canada, to New York.

Modern science has failed to achieve mastery over nature.

In addition to technological risks, Beck also analyses the underlying matrix of attitudes in modern societies that simultaneously—and perhaps paradoxically—precipitates intense risk aversion. For Beck, ‘risks and risk perception are [the] unintended consequences of the logic of control which dominates modernity’ (1992, pp. 215–6). Modernity and modern society are, after all, about the control of nature, about constructing order and achieving regulatory power over nature in manufactured, modern environments (Parsons 1954; Giddens 1990). Modernity is thus about eradicating risk, or at the very least reducing exposure to it. Yet modern science has failed to achieve mastery over nature, and is itself fallible and risk-generating, as the explosion of the spaceship Columbia in 2003 or the widespread prescription of Thalidomide during pregnancy during the 1960s or the power failures already mentioned attest. For Beck the promise—and threat—of modern science explain our fascination with risk and its ubiquitous application.

SIZE MATTERS, OR BIGGER IS RISKIER?

In Megaprojects and Risk: An Anatomy of Ambition, Bent Flyvbjerg, Nils Bruzelius and Werner Rothengatter explore why so many ‘megaprojects’ experience large risk events and fail, or can only be finished and/or continue to operate with substantial (and unplanned for) public subsidies. The authors call this the ‘performance paradox’: more and more megaprojects are being proposed and developed but their record of economic success and environmental impact is mostly dismal (p. 3). With such substantial resources at play in a single megaproject, one would assume that the developers would have the technical competency required, that they would have sophisticated project development plans and financial and economic forecasts of costs and revenues, and that they would have developed workable scenarios for changing circumstances. However, the contrary appears to be true of most megaprojects. In the case of the Channel tunnel, for example, cost overruns of 80 per cent and four instances of near bankruptcy speak to the failure of economic forecasts and risk modeling techniques. Denver’s US$5 billion new international airport was launched in 1995 after a cost overrun of 200 per cent and passenger traffic in the first year of operation only half that projected (p. 3). The authors provide many other examples of equally as disastrous megaproject outcomes.

How can so many experts get so much so wrong so often and in magnitudes that are acutely embarrassing? Flyvbjerg, Bruzelius and Rothengatter argue that in megaprojects risk, democracy, and power inevitably intersect. Political and regulatory authorities set project parameters and goals to suit political ends in democratic environments, but these environments frequently involve changing parameters or less than transparent regulatory structures and pricing regimes that impact the feasibility of a project. Megaprojects cannot be planned and executed in a predictable Newtonian world where cause and effect are self-evident and where order and regulatory stability are the appropriate benchmarks for scenario development. Political interference, changing planning policies and/or needs, changing governments, and other externalities make for grossly imperfect knowledge environments in which to execute megaproject developments (p. 6).

Megaprojects cannot be planned and executed in a predictable Newtonian world.

Too many risk feasibility studies, Flyvbjerg, Bruzelius and Rothengatter argue, rest on what the World Bank has called the EGAP principle—Everything Goes According to Plan (p. 80). Plainly this is infrequently the case, so why does EGAP hold sway? The authors explain: feasibility studies assume all will go to plan for essentially political reasons: political, institutional, and regulatory stakeholders in megaprojects desperately want their projects to work, or are misled about the true risk exposure entailed in the project because key commercial stakeholders provide inadequate information. To overcome these problems Flyvbjerg, Bruzelius and Rothengatter call for risk analysis to be based on what they term the MLD (Most Likely Development) principle. Public sector organisations should undertake MLD scenario development to identify the riskiest parts of megaprojects and to carrel these risk areas accordingly.

These suggestions are well rehearsed and make a great deal of sense. However, the authors don’t consider the feasibility of implementing decisions arising from an MLD informed process when institutions don’t operate transparently. Megaprojects undertaken in emerging economies, for example, face poor prospects for more transparent stakeholder involvement, and government bodies in emerging economies often lack the institutional capacity and depth to perform proficient risk assessments.

Flyvbjerg, Bruzelius and Rothengatter propose mechanisms to overcome these problems and to generate more robust and realistic MLD risk profiles. One mechanism is to increase the risk exposure of venture capital in megaprojects rather than indemnifying private money with sovereign guarantees (p. 144). If at least one third of the private venture capital in megaprojects was not guaranteed then private investors would be forced to perform more rigorous and realistic risk analyses.

The authors also recommend BOOT (Build Own Operate Transfer) and BOT (Build Operate Transfer) schemes along with other concession approaches (p. 125). These schemes load the risk on the commercial contractor and operator of the project, forcing careful cost and revenue forecasts to ensure the commercial viability of the project. BOOT and BOT schemes do have certain inbuilt sovereign guarantees, but Flyvbjerg, Bruzelius and Rothengatter argue that these schemes increase the likelihood of rigorous project risk assessment and reduce the likelihood of political or regulatory risk by contractually tying governments to long term commitments in specified regulatory frameworks, and pricing and cost arrangements (p. 128).

These proposals make much sense, but unfortunately the authors don’t go on to evaluate whether these concession schemes work or analyse how risk under concession systems has actually been managed. Such analysis would be especially helpful given the near explosion in the use of concession schemes to fund infrastructure development. For example, telecommunications fixed line network rollout and mobile infrastructure in Asian and Latin American economies have extensively employed concession schemes; most seem to have succeeded.

Proposed schemes load the risk on the commercial contractor and operator of the project.

Omissions from the volume are mostly minor. The authors don’t explain how the MLD principle can or should be operationalised, or how scenarios developed under this principle can be ‘stress tested’ to assess the accuracy and soundness of their assumptions. Nor do they give sufficient attention to the practical problems of implementing their proposals for insulating state agencies from political pressures to enable efficient risk assessment of megaprojects, especially in emerging economies. But perhaps these questions are beyond the scope of the volume as intended, and the volume is nevertheless a very welcome addition to the risk literature. It is exceedingly well researched, meticulously argued, and accessible to all. On all these counts the authors are to be congratulated.

ENVIRONMENTAL POLITICS AND RISK

In Risk and Reason: Safety, Law and the Environment Cass Sunstein examines how the United States has responded to environmental risk. He recognises the increasing role of risk in American society and the increasing fixation with risk among interest groups and broader American constituencies. But the fact of risk and public concern about it are not the problem Sunstein tackles. Rather he is concerned about how attempts to reduce environmental harm and risk from environmental degradation frequently result in public policy initiatives that are mostly ‘stabs in the dark’ (p. viii). Well-intentioned though these may be, the lack of successful environmental policy leads Sunstein to explore problems with public policy on environmental risk in the United States. He aims ‘to link people’s ordinary thinking about risk to recurring issues in law and politics’ (p. viii).

Risk and Reason embodies a kind of crusade: Sunstein calls for institutional reform, congressional legislation, and executive branch intervention to create a more orchestrated approach to environmental risk and risk management. He also assesses environmental risk regulation since the early 1970s and looks at the costs of this legislation compared to the outcomes and benefits. Surveying data from the Office of Management and Budget (OMB), for example, Sunstein finds that the OMB has determined that the benefits of regulation have mostly outweighed the costs. (He cites OMB figures for 2000, for example, that count regulatory costs of between US$146 to US$229 billion and regulatory benefits of between US$254 billion to US$1.8 trillion.)

However, Sunstein contends that, on closer examination, the OMB figures are misleading. He argues that funds have been misallocated, with substantial resources going to relatively small problems; driven by popular over-exposure and undue concern. When broken down, Sunstein shows that the regulation of methylene chloride, for example, will cost US$100 million annually but deliver benefits of only US$40 million and that the regulation of ‘financial assurance for municipal solid waste landfills’ will have ‘monetized benefits of US$0 but costs of US$100 million’ (p. 25). The examples continue. Sunstein concludes that a further 60,000 additional lives could be saved each year if governments reallocated regulatory monies. He goes further, claiming that upwards of US$31 billion could also be saved with budgetary reallocation along with the 60,000 additional lives. Sunstein’s argument is thus not so much about risk and regulation as efficiency in risk regulation, matching resources with clearly identified priorities such that the OMB and the American taxpayer get the biggest bang for the risk regulation buck.

Intuitive perceptions of risk tend to be poorly correlated to real, significant risk events.

Thus Sunstein’s question and task converge on how this reform is to be achieved. He argues that governments should make cost-benefit analysis the central analytical rubric for public policy making. However, cost-benefit analysis presupposes an ability to value and quantify risk and to conduct calculations with relatively refined data. To enable more effective cost-benefit analysis, Sunstein assesses various methods for thinking about risk (pp. 28–52). One particularly insightful argument he makes is that intuitive perceptions of risk tend to be poorly correlated to real, significant risk events and processes. People react to risk in discontinuous, ill informed, and emotional ways, producing regulatory outcomes that consume enormous resources without effectively reducing society-wide risk. He blames the poverty of intuitive perceptions partly on the social amplification of risk through media analysis of risk events and how this influences public perceptions of risk and the consequent hastily constructed regulatory responses to these risks.

Overcoming this problem is, in one sense, simple for Sunstein: better information, better measurement systems, better dissemination of risk information, and thus better, more transparent information to inform people’s objective decision-making about the precise risks they face to their health, safety, and well being. As any risk practitioner recognises, however, this is enormously complex and hard to achieve, and perhaps improbable given the variety of measurement systems, information overload, and the amount of contradictory information available to citizens. Measuring risk in purely scientific, objective ways, even among the hard sciences, is a goal yet to be achieved—if ever.

Sunstein recognises many of these constraints (pp. 260–6) and is acutely aware of the information overload that citizenship entails in a complex democracy like the United States. He is also aware of the practical limitations of developing risk indices or measurement systems to assess risk. Yet Sunstein does not seem to have a way forward through this predicament, apart from affirming the need for more ‘effective communication of risk information’ (p. 264)—but not through advertising because this entails its own dangers— and obtaining better measurement systems (pp. 266–9).

In the end, Sunstein perhaps bites off more than he can chew. The risk universe is big even when delimited to sovereign environmental risk. He seems reduced to analysing numerous specific policies and problems, which, although immensely interesting and deserving of concerted attention, feel as if they are left unresolved.

In the end, Sunstein perhaps bites off more than he can chew.

The book has another fault instantly recognisable by the European risk community. Both the lens through which Sunstein views risk and the recommendations he advances to deal with it betray a strong North American bias. Sunstein celebrates technocracy and the role of objective, value neutral science in achieving risk solutions while also seeing non-interventionist economic incentive-based risk regulation as vastly superior to government command-and-control regulatory responses (p. 269). He presumes rather than demonstrates the superiority of incentive-based regulation in positively distorting market signals and allowing industries and actors to respond in socially advantageous ways. He rejects the heavy hand of the state on technocratic grounds because it produces less than perfect outcomes. This argument remains contentious and the cultural divide toward risk regulation across the Atlantic is far from bridged by Sunstein’s volume.

Sunstein does go to some length to argue that technocratically informed positive feedback market signals will not circumvent democratic processes. He claims that a system of financial incentives allows ‘far less room for interest group maneuvering’ (p. 273). Precisely how this works is not clear, but any keen observer of government subsidy programs (in what ever form they might take) would always assume that such systems remain open to abuse and to pressure from special interests.

Sunstein’s contribution delivers less than it promises. He places too much faith in science, and seems to underestimate just how political politics and policy making can be.

REFERENCES

Beck, U. 1992, Risk Society: Towards a New Modernity, trans. M. Ritter, Sage, London.

Department of Defence n.d., Predict Defence Infrastructure Core Requirements Tool [Online], Available: http://www.defence.gov.au/predict [2004, 13 June].

Quiggin, J. 2003, ‘Managing the Risky Business of Life’, Australian Financial Review, 13 June.

Darryl Jarvis is Senior Lecturer in Government and International Relations in the School of Economics and Political Science at The University of Sydney. He is foundation Director for the Centre for International Risk, a joint enterprise between the University of Sydney, University of South Australia and the Faculty of Economics and Business, the University of Sydney. In 2003 his edited collection, the Handbook of International Business Risk: The Asia Pacific, was published by Cambridge University Press. His Transboundary Risk and Multinational Enterprise: Political, Economic and Country Risk in International Relations will be published later in 2004.

View other articles by Darryl Jarvis: