According to the receipt that I kept as a bookmarker, I acquired this book during a visit to the University of Stavanger in February 2010. I had read some of Aven’s other work and the title of this book sounded extremely appealing and interesting. Enthusiastically I started reading, but after a relatively tiresome first chapter I put the book temporarily down and prioritized reading other stuff. A lot of other stuff… In the end, ‘temporarily’, turned out to be for about 5 and a half years! Recently I decided to make an effort and work myself through the book.
I already criticized Aven slightly for his style of writing in my review of “Foundations Of Risk” and I was a bit disappointed to find out that this book is even worse at times. Occasionally there are too many formulas. This coupled a rather academic style of writing (including the obligatory references, quite a lot of these refer to previous work of Aven himself, which looks a bit odd, but okay, he has done extensive work in the field of risk) result in a book that isn’t always accessible. I’ve found out that I don’t have the patience when I read a book like this to try to follow all the formulas and calculations. But don’t despair, the strategy that I used for “Foundations…” also works here: sometimes I just skipped forward to the short (but rather good) summaries that each chapter has and that works fine in order to get the message.
Chapters 4, 15 and 16 provide ‘good’ examples of the at times gruesome technical approach with formulas and statistical stuff that makes me doze off (and, as said, subsequently skip fast forward to the summary). The material is theoretically, scientifically and mathematically correct, I suppose, but I myself find words often clearer and easier to understand (mind you, not always, sometimes numbers can have enormous communicative value, but formulas seldom have) and I am a major fan of storytelling - especially with regard to risk!
In the preface Terje says that “The book has been written for professionals in the risk field, including researchers and graduate students”, but I have to say that I feel that the book appears to be written mainly for the latter two, while many risk professionals would have benefitted from a much more accessible and less academic approach. It wouldn’t surprise me if others like me pick up the book because it looks interesting but then give up because it’s tiresome.
The preference for the academic audience also betrays itself slightly in one of Aven’s conclusions: “The scientific basis of the risk assessments needs to be strengthened”. I wonder if the focus shouldn’t rather be on a need to strengthen and improve the practical application, but maybe a theoretical foundation is seen as the necessary requirement there…
Having this criticism out of the way, the subject is fascinating and Aven says a great many smart and wise things about the subject. I’m not going to do a detailed description or summary here, but an overview of the various chapters is rather informative about what to expect:
- Risk is Equal to the Expected Value.
- Risk is a Probability or Probability Distribution.
- Risk Equals a Probability Distribution Quantile (Value-at-Risk).
- Risk Equals Uncertainty.
- Risk is Equal to an Event.
- Risk Equals Expected Disutility.
- Risk is Restricted to the Case of Objective Probabilities.
- Risk is the Same as Risk Perception.
- Risk Relates to Negative Consequences Only.
- Risk is Determined by the Historical Data.
- Risk Assessments Produce an Objective Risk Picture.
- There are Large Inherent Uncertainties in Risk Analyses.
- Model Uncertainty Should be Quantified.
- It is Meaningful and Useful to Distinguish between Stochastic and Epistemic Uncertainties.
- Bayesian Analysis is Based on the Use of Probability Models and Bayesian Updating.
- Sensitivity Analysis is a Type of Uncertainty Analysis.
- The Main Objective of Risk Management is Risk Reduction.
- Decision-Making Under Uncertainty Should be Based on Science (Analysis).
- The Precautionary Principle and Risk Management Cannot be Meaningfully Integrated.
Finally there is one more chapter that sums up the main messages from the book and presents a proposal for a structure of the risk assessment framework according to Aven.
The book is relatively compact with only 252 pages and consequently the chapters are relatively short. Some extremely so (4 to 5 pages), most are in the 8 to 12 minute range and a few reach 20 pages. That at least helps to consume the book in small packages. Each of the chapters has a list of references and suggestions for further reading.
To just pick some cherries, I think that chapter 18 on decision making and the role of risk assessment might be the most important chapter. So if you have limited time do yourself a favour and read through that one (much of that message is also found in “Foundations…”, by the way).
I would also recommend taking a look at chapter 17 which includes a good explanation of the ALARP process and at chapter 19 with some good discussions of the precautionary and cautionary principles
Other important messages from the book (hard to pinpoint because they are kind of threads throughout the entire book) include frequent discussions of knowledge based probability (instead of the frequentist approach that some in our profession strive after) and understanding of and working with the concept of uncertainty.
All in all interesting and thought provoking, but not easy to digest.
I also want to recommend the fine and insightful review on Amazon by Meine van der Meulen.
(Wiley, 2010, ISBN 978-0-470-68388-0)