donderdag 16 december 2010

Review of Ridley's "The rational optimist"

In “The rational optimist”, Matt Ridley argues that, contrary to widespread perception, the human condition is improving through time, and at an accelerating rate.
Matt Ridley, a PhD in zoology turned journalist and businessman (he acted as non-executive chairman of the UK bank Northern Rock from 2004 to 2007), thus joins the ranks of luminaries such as Julian Simon and Bjorn Lomborg in claiming that the claims of the Cassandra’s of our time are both factually wrong and logically flawed.
The argument of the book rests on two pillars.
First of all, there are the factual arguments. Ridley argues that you can take almost indicator of human well-being, and that you will find that, through time and in most places, these indicators have evolved for the better: food availability, income, life expectancy, child mortality, violence, air quality, water quality, etc…
Second, Ridley suggests an overarching theory of why things got better and why, in his view, they will get even better in the future: of all species, mankind is the only one with a propensity to exchange (Ridley  explains at length why he thinks this is different from the “propensity to reciprocity” that has been discussed in behavioral economics). Exchange leads to specialization according to comparative advantage and, as Ricardo already pointed out more than two centuries ago, this leads to a higher productivity and thus to higher overall welfare. However, Ridley goes on to argue that human exchange is not limited to an exchange of goods, but also includes the exchange of ideas, which are the main source of productivity improvements over time. Mankind progresses because “ideas have sex”: the Internet, for instance, was spawned by the intercourse of the telephone and the computer. Ridley’s optimism about the future is based essentially upon the assumption that ideas will continue to have sex with each other, and lead to a productive off-spring.
A book whose avowed purpose is to row against the tide is bound to prove controversial. Before writing this review, I have briefly looked at some of the discussions on the Internet, and, as expected, there is some heated debate going on out there. As a rationalist who is not naturally inclined to optimism, I will contribute my bit here.
Before proceeding with an assessment of the contents of the book, let me first say some words on the form. Even though “The rational optimist” is not a scientific essay, it does take issue with the work of scientists and aims at influencing the public debate. If an author want this type of work to be taken seriously, rigorous referencing is absolutely required. To be honest, the referencing in this book is dreadful, and would not be accepted in an essay by an undergraduate student. First of all, even though the book contains more or less 60 pages of notes of references, several of the boldest claims of the author are actually not referenced at all! For instance, nice to hear that sperm counts are not falling, but where does the information come from? As these claims can hardly be considered to be common knowledge, one wonders why the references have been omitted. Second, although the author cites extensively from peer reviewed scientific papers, some of the (again) boldest claims come from very partisan secondary sources. I think this seriously undermines the credibility of the book. It is only  in my own field of expertise that I can assess the credibility of the claims made without verifying the original source – for other subjects, I would like to see a reference to a sources with sufficient credibility (to be honest, all the references that I did check out turned out be correct, but, again this doesn’t say anything about the veracity of the claims that have not been referenced).
A related problem is that, as far as I can judge, the author has been very selective in his references. For instance, Ridley argues that, when incomes rise, environmental quality first declines and subsequently improves: the economists amongst us call this hypothesis the “environmental Kuznets curve”. Ridley then cites one relatively old paper on the subject and creates the impression that the issue is settled. Actually, it is not: there is still a very heated debate amongst environmental economists whether or not this hypothesis is correct. One wonders if this selective referencing is a deliberate strategy or is simply the by-product of trying to cover too many fields in a short book.
Now, what about the substance?
To begin on a positive note, there are two points that I have very much enjoyed about this book.
First of all, as already indicated above, Ridley succeeds very well in explaining some essential insights from economics: the benefits from the division of labour and free trade, why the production of knowledge is fundamentally different from the production of material goods, why autarky is a very bad idea, why the failure of real-world markets does not automatically imply that governments will be able to improve upon them…  Despite the biases in the literature he quotes, I think Ridley got most issues right.
Second, he provides some detailed refutations of very concrete doom scenario’s that are fashionable in some circles. In the process, he provides several hilarious examples of past predictions of eminent collapse that have been disproved by time (to put it mildly). Everyone who thinks the end is near should look up the predictions made by the Club of Rome on the exhaustion of natural resources or by Paul Ehrlich on demographics (or on cancer prevalence).
However, on the whole, the book is disappointing, precisely because it suffers from exactly the same faults as the pessimists it wants to refute:  it gathers all the evidence that confirm its thesis, ignores largely evidence pointing to the contrary, and then jumps to conclusions.
Let me be more concrete. On the whole, I agree with Ridley that, contrary to conventional wisdom, things have indeed gotten much, much better over the last two centuries for most of mankind. I also think that those who pretend to the contrary either do not know the facts or use benchmarks against which reality will always be disappointing. However, Ridley does not really consider the questions whether these improvements are (to use a work I am not very fond of) “sustainable”.
Ridley seems to assume that, just because all past predictions of doom have turned out to be wrong, predictions of doom are inevitably wrong.  It would have been much more instructive if Ridley had carefully reviewed the assumptions and data underlying these predictions and explained why they have led to erroneous predictions (in the passing, he may have discussed Jared Diamond’s examples of past civilisations that have collapsed under environmental pressures). For instance, the predictions of the “Limit to growth” models may have been spectacularly wrong, but their most important flaw was that they did not leave any room for consumers or firms  to respond to changes in market prices. As argued by economist William Nordhaus, the fact that the authors of these models had grossly underestimated available stocks of non-renewable resources is of completely secondary importance.
Because Ridley emphasizes the faults in the work of other authors, he tends to overlook some important gaps in his own thinking. For instance, Ridley argues that, without fossil fuels, agriculture would never have been able to feed six billion people. As I am not an agronomist, I do not feel well placed to challenge the natural science underlying his calculation and, as I have found no obvious faults in Ridley’s arithmetic, I  am willing to accept this point. However, this does not reassure me at all. As Ridley acknowledges, for all practical purposes, fossil fuels are non-renewable. This means that, at a certain stage, humanity will have to switch back to sources of energy that Ridley has demonstrated not to be able to sustain six billion people. Of course, one could argue (and Ridley does argue) that we will find substitutes by the time we run out of fossil fuels. However, I find it odd that Ridley is so confident in this deus ex machina, after having spent several pages explaining why renewable energy is not an acceptable alternative to fossil fuels at all. This brings us to the fundamental problem with Ridley’s analysis: his willingness to just assume that innovation will continue at an accelerating rate.
Paradoxically, Ridley actually shows that there are a lot of mechanisms that can kill innovation. While I agree with him that the most important culprits in this field are despotic governments and religions, I think he underestimates the capacity of private monopolies to stifle innovation as well. In general, a book that puts such a high emphasis on the role innovation plays in human progress, should have put much more effort in a discussion of the social institutions that make innovation possible (or that can hamper it). As there are many more ways to get things completely wrong than to get them even more or less right, I would not be sanguine about future prospects for innovation, especially if this is the only path to a “sustainable” future.
This lack of serious analysis of institutions is another serious flaw in the book. Things have not only improved thanks to innovation – in some cases, government intervention has really changed things for the better. For instance, the decrease in the emission of sulphur oxides and nitrogen oxides in the past decades have led to a spectacular increase in our air quality, and to substantial health benefits (which have been estimated using rigorous statistical analysis). This decrease has taken place at a very reasonable macro-economic cost, mainly thanks to…innovation. Does this substantiate Ridley? No, for the very simple reason that industry would never have bothered to look for ways to reduce its emissions in an innovative ways, had these reductions not been imposed by government regulations. Optimism alone would not have solved the problem.
Let me conclude. Sir Karl Popper has claimed that optimism is a moral duty. I beg to differ. It is our moral duty to be neither cynical nor complacent, but to use our intellectual resources to make the best possible assessment of the threats and opportunities we face, and to look for solutions. Optimism is an act of faith – a rationalist would not consider faith to be a good basis for action. 

zaterdag 27 november 2010

Review of Paul Seabright's "The company of strangers"

How can we explain that humans are the only living creatures who cooperate with unrelated members of the same species on projects of high complexity and requiring a high degree of mutual trust?
This is the challenging question that Paul Seabright proposes to answer in “The Company of Strangers. A natural history of economic life”.
In case you might think that this is a book about evolutionary biology, Seabright points out right from the start that the time that has elapsed since mankind started living in larger social groups than his close relatives is simply too short for the genetic hardware of homo sapiens to have adapted to these new social surroundings. While he acknowledges that “the division of labour amongst human beings must have piggybacked on a physiology and a psychology that evolved to meet a far different set of ecological problems”, Seabright sets clearly out that this is foremost a book on the social institutions that make it possible for complete strangers, not just to live together in (relative) peace, but to actually enter into mutually beneficial exchange.
The wording “mutually beneficial exchange” will probably have rung a bell amongst those who recollect their “introductory economics” courses, and it will come as no surprise to them that Paul Seabright is indeed a professor of economics at the University of Toulouse.
Is this a book about economics then? Well, if, as me, you think that economics is a subject that does not just cover the analysis of market exchanges, but provides an intellectual framework for understanding all institutions that govern human life (including family, the law, religion, politics…), then this is a book about economics. If, as me, you think that economics and evolutionary biology use basically the same intellectual framework to understand life (note that Darwin himself has always acknowledged his intellectual debts to the classical economists), then this is a book about economics - or is maybe a book about evolutionary biology after all?
As will become clear in this review, it is close to impossible to put a label on this book, whose avowed ambition is to integrate insights from economics, philosophy, evolutionary biology, psychology and political science in an all-encompassing explanation of the human condition.
The actual contents can be summarized as follows.
In a first stage, Seabright shows that even some of the simplest activities of modern societies depend upon intricate webs of cooperation that function without central coordination – an enlightening passage on this issue is Seabright’s report of a discussion with a former Russian communist party official who cannot grasp that no one is actually in charge of the supply of bread in a market economy. Seabright claims that this cooperation is the result of the willingness of individuals to cooperate with strangers “in a multitude of small but collectively very significant ways”.  
In a second step, Seabright argues that this cooperation is possible thanks to a set of institutions that have often grown by experiment or as the by-products of attempts to achieve other ends. For instance, while the social division of labour is central to the human progress, it is only possible if participants can trust each other. Therefore, institutions are needed that ensure that “cooperation not only happens but is reliable enough for others to be willing to take its presence for granted”. Maybe surprisingly for people who have not been formally trained as economists, one such institution is… money. Because we accept money “in exchange for valuable goods in spite of the fact that we may know nothing about the individuals who are offering it to us”, it “is one of the great human inventions precisely because it helps to narrow the gulf between the ingenuity of each individual and the interest of others.”  Seabright goes on to explain that, paradoxically, human cooperation requires the capacity to participate in the creating of prosperity without knowing or even caring about the overall outcome (what he calls ‘tunnel vision’). However, Seabright is not blind to the dangers of ‘tunnel vision’ and illustrates at length that, for instance, codes of professional ethics can make individual acts of local cooperation more reliable, but can also generate blindness to the more distant consequences of one’s actions (just think of the Nuremberg trials). This part of the book also contains a detailed discussion of how “reciprocity”(rather than narrow self-interest or naive altruism) is the basis of successful human cooperation.
The third part of the book takes a look at the global consequences of human interaction: the development of cities, the impact on the environment, the functioning of markets, the development of large firms … While Seabright is not blind for some of the negative effects of human interaction (such as our impact on the environment), he rightly draws the reader’s attention to the capacity of markets “to calculate prices that summarize the information necessary for allocating resources in a world of scarcity” – not surprisingly, markets are one of those institutions that make complex societies possible.  Seabright also points to the (apparent) paradox that, while firms use the market externally, they have eliminated it internally and replaced it by explicit planning and coordination – this issue sheds a lot of light on the question under which conditions planning can be superior to markets (and vice versa).
Finally, part four looks at the institutions of collective action (this is, the political institutions) and discusses how they can deal (or not) with the collective problems of mankind. Seabright emphasizes that the human capacity for cooperation can be also be directed at destructive ends, and that no unduly trust should be put on political institutions.
In case you think that a book with this scope would fill a complete shelf in your home library, be reassured: the book counts merely304 pages, notes and bibliography included.
This is precisely what makes this book outstanding. It does not offer any original contribution to the scientific literature (nor does it claim to do so), but brings together insights from different fields in an accessible summary. Those who are familiar with those fields will find no big surprises.
For the interested lay reader, The company of strangers provides a very accessible and well-written non-technical introduction to the economic analysis of institutions. As each chapter is complemented by a list of notes that refer to the more specialized scientific literature, it can be the starting point for a long journey into this fascinating field.
However, despite its accessibility, the book is intellectually rigorous and specialist readers will also find the book worthwhile. Moreover, it is chockfull of examples that can be used to illustrate principles of economics to a non-specialised audience (including undergraduate students) – I am sure economists will simply love his criticism of Naomi Klein (simply put, No logo contains two important claims that  can obviously not simultaneously be true: (a) through the creation of worldwide brands,  corporations have become all-powerful (b) corporations are engaged in a desperate struggle to survive in competition from each other). Similarly, his demystification of the moral superiority of economies based on barter or gift-giving rather than monetary exchange, should be required reading for everyone.
I must now resist the temptation to reproduce all these examples in this review: besides making me a plagiarist, it would take away any incentives to go out and buy the book for yourself.
Does this mean the book is without faults? Well, of course not. There are some loose ends in the argumentation, and one is left with the feeling that, in the end, the central question (how did we evolve from small groups of hunter-gatherers to complex large societies) remains unanswered. But this is probably more an indication of the size of the intellectual challenge set by the question than of the book’s qualities. Also, I would like to have seen a more concrete discussion of some current global policy challenges, but this probably just a result of my own professional pre-occupations.
Simply put, this is one the books I wish I’d had written. Actually, I am wondering why you are still reading this review instead of ordering it!

vrijdag 12 november 2010

Review of Bernstein’s “The birth of plenty” (2004) and Landes’s “The Wealth and Poverty of Nations” (1999)

Nobel Prize winning economist Robert Lucas has remarked that, once one starts thinking about the determinants of economic growth, it is hard to think about anything else. While there is huge scientific literature on the subject, it is hard to find books that target a general audience but that are not heavily biased ideologically.  
Bernstein’s “The birth of plenty” (2004) and Landes’s “The Wealth and Poverty of Nations” (1999) are exceptions. These books share the ambition to provide an all-encompassing explanation of long term economic development. The time perspective they take is more or less the last millennium (although Bernstein also succinctly discusses ancient Greece and Rome). This approach sets them apart, both from authors who take the ultra-long term perspective (such as Jared Diamond in “Guns, germs and steel”) and from authors who emphasize the quantitative analysis of recent economic growth.
However, apart from their ambition, the books couldn’t be more different.
Let us first consider “The Wealth and Poverty of Nations”. Landes is an emeritus professor at Harvard, and he has clearly intended this book to be his magnum opus (the not-so-subtle reference to Adam Smith’s “Wealth of Nations” takes away any ambiguity on this specific point).
Landes starts his book with a discussion of the importance of geography, and admits that geographical factors (a temperate climate, the absence of tropical diseases, a favourable coastline and navigable inland waterways) do play a role in understanding economic growth. However, this discussion is limited to a dozen pages (out of approximately six hundred), and Landes quickly moves on to a detailed account of economic history, starting with medieval Europe and ending with the financial crisis that struck Southeast Asia in the second half of the 1990s.
This book reflects the result of several  decades of work, and this shows. It is encyclopaedic in nature, with a strong emphasis on cultural factors on the one hand and technological progress on the other hand.
In my view, the detailed accounts of technological breakthroughs are the most valuable part of the book, because they clearly illustrate several important points: (a) until the 19th century, there was almost no interaction between science and technology, and several important breakthroughs were the work of uneducated craftsmen (b) most technical advances are incremental in nature (c) most technical advances are the result of decades of painstaking trial-and-error processes (d) technological progress is path-dependent (e) once a certain threshold has been reached, technological progress can become a self-reinforcing process (f) small improvements can have important repercussions in unexpected areas. My favourite case study  is how the industrial revolution has lead to the replacement of wool by cotton, and how this has in turn affected labour productivity and accelerated industrialisation (basically, because cotton is easier to wash and itches less than wool, the move to cotton has dramatically improved hygiene and improved the health and strength of the labour force – until cotton was introduced, most people actually wore constantly the same pants until these literally fell apart – it does not take a lot of imagination to see what this must have implied in terms of hygiene).
However , even after the second reading of the book, I am still not sure what the central message really is (apart from the fact that Landes does not like quantitative-minded economists). The book has been described by some reviewers as a “ode to Europe” (less kind reviewers have called the book Euro-centric). (To be fair, the book contains extremely gruesome accounts of the colonization process, so I am not really sure whether the book can really be called an “ode” to the colonizers).
Well, it is clear that Landes does think that the seeds of the industrial revolution were sown in Europe centuries before the revolution really took off. Landes provides detailed accounts of little-known technological advances that were implemented in Europe in the Middle Ages. In my opinion, he rightly emphasizes that the Chinese may well  have had superior scientific and technological knowledge until the 15th century, but that it was their decision (and theirs only) not to use this knowledge to any productive use, while the Europeans did constantly push the frontiers of their own knowledge further and further. Thus, as far as the descriptive issues are considered, I have no problems with Landes. However, he fails to convincingly answer the questions why late-medieval European were so much more open-minded than the Chinese. It is too easy to just refer to cultural factors whenever differences between human societies cannot be explained in terms of genetics.
One point that has really annoyed me is the recurring criticism of economics and quantitative approaches to history. Sometimes, one is left with the impression, that whenever quantitative evidence is at odds with verbal accounts of the same issue, Landes just assumes that the figures must be wrong. Moreover, he repeatedly shows that he definitely does not understand the economic theory of comparative advantage (while he is in good company on this issue, I do not think it is very scholarly approach to assume that any theory that you do not understand0 must be wrong).
In short, while “The Wealth and Poverty of Nations” is worthwhile being read, it is more because of the wealth J of facts that are contained in the book, than because it provides far-reaching insights.
Both the tone and the substance of “The birth of plenty”  couldn’t be more different.  Its author, William Bernstein, is a neurologist and (apparently self-educated) financial analyst, who has written the book after discovering the writings of economist Angus Maddison on long-term economic development.
Somewhat disrespectfully, one could thus state that “The birth of plenty” is the work of a hobbyist. This shows in a (by times) very selective reading of the economic literature and a superficial understanding of economic theory.
This being said, the main advantage of Bernstein’s book is precisely that it has been written by a dilettante who is not bothered by an evaluation by his scientific peers: the book is very succinct and readable, and is very explicit in its central propositions.
In short, Bernstein claims that the simultaneous occurrence of the following four conditions is necessary for economic growth: the protection of private property rights, scientific rationalism, efficient capital markets, and low transport and communication costs. The book illustrates with numerous case studies what happens when the four conditions are not fulfilled simultaneously.
For instance, although individual property rights had already been protected by law in Britain since the Middle Ages, and although the rule of law steadily improved through time, it was not until the end of the eighteenth century that the industrial revolution really took off – in Bernstein’s view, because the other three conditions were not yet fulfilled.
It is certainly easy to criticize this simple message on several accounts.
First, the four conditions are not completely independent from each other. Surely efficient capital markets cannot develop without the protection of property rights (which includes the protection of shareholders and creditors)? And is the spectacular decrease of transportation and communication costs in the last two centuries not the result of mobilizing large capitals to bring the results of scientific progress to the market?
Second, the author does not really give any operational content to the conditions he has identified. Enforcement of property rights has never been perfect and never will be – it is a matter of degree. No one disputes that, back in the 18 century,  the rule of law was stronger in Britain than in France, and stronger in France than in the Ottoman Empire. But what exactly where the differences between these legal systems that explain the larger development gap between France and the Ottoman Empire than between France and Britain?
Similar considerations apply to the other conditions as well. The development of modern capital markets has come gradually, and the basis for them (such as double entry bookkeeping, money creating banks) was already there in Medieval Italy; joint-stock companies and central banks had been around for centuries before the Industrial Revolution took off. Actually, as Bernstein acknowledges, the law imposed extremely stringent conditions on the creation of limited liability companies until late in the 19th century, even in Britain – who would even consider nowadays investing his savings in a joint-stock company with unlimited liability of the shareholders?
As regards scientific rationalism, recall that until the 19th century, science and technology developed more or less independently: what really mattered for technological progress was a mind that was open to new knowledge, and this state of mind already existed in Europe for centuries, as long as this knowledge did not touch upon issues of religion (although Peter Watson has argued in “Ideas” that even in supposedly “dark” Medieval Europe, the intellectual debate on theological issues was much more open than in some areas of the world nowadays). And, to be honest, I have not understood the relevance for economic growth of the long chapter on astronomy, although the discussion in Bernstein is indeed “enlightening” on other issues (with the observational tools at hand at the time, the empirical support for Galilei’s system was not as strong as we think, and Copernic’s system was actually complete nonsense – it is not because the Church was wrong, that its opponents were right).
A third point of criticism that can be levied against the historical analysis is that the central thesis relies very heavily on the quantitative work of Angus Maddison. We easily forget that more or less reliable economic statistics are a recent phenomenon – Maddison’s work is thus based mainly on indirect evidence, and has been criticized for being very speculative.  To be fair to Maddison, he has always acknowledged this specific point, and challenged his critics to find better data. However, Bernstein sometimes gives the impression that he takes Maddison’s figures at face value. This is not an innocuous assumption, as one of the central assumptions in this book is that the growth of income per capita only really took off in the beginning of the 19th century.  Bernstein explanation for this is that, until then, technological progress was always offset by the population growth it induced – in other words, that the world was caught in a Malthusian trap.
In general, Bernstein does not sufficiently acknowledge that, taking into account their quality, we should be very careful in drawing sweeping conclusions from the data that are available. For instance, the absence of hard data on how income inequality has evolved during the industrial revolution means that we have to mainly rely on verbal accounts and indirect evidence – this is quite unsettling, taking into account that this issue was at the heart of the development of Marxism!
The best chapters in this book are the ones that discuss institutions in Britain, France and Spain from the late Middle Ages until the industrial revolution. Bernstein describes vividly how the legal system in Britain improved through time and how it affected incentives for industry positively, while the French and the Spanish kings devised policies that were so daft that one would think that they had been designed with the only purposes to stifle growth.
Bernstein also discusses several contemporary issues, such as the (lack of) the correlation between income levels and indicators of happiness, the link between development and democracy, the trade-offs between income growth and redistribution and possible constraints on future growth. Both left-wing and  right-wing ideologues will find elements in these chapters that they will profoundly dislike – I take this as a strong indicator of Bernstein open and independent mind.
This does not mean that these chapters are without flaws. For instance,  Bernstein heavily quotes from the work of economists Barro and Sala-I-Martin when discussing the relationship between economic development and democracy. Barro and Sala-I-Martin are admittedly reputable references, but, when discussing a highly controversial issue, other sources of information should have been discussed as well.
The weakest point in the analysis of contemporary issues, however, is the cursory treatment of natural resources. I agree with Bernstein that both the predictions and the intellectual framework of the Club of Rome are seriously flawed, but this is not enough to dismiss concerns that the (un)availability of natural resources can constrain income growth in just a few lines.
However, the sloppiness in the argument is hidden by a very fluent narrative – and, as this is not a scholarly book, I think the author should be forgiven for his faults. After all, most of the claims have indeed been vindicated by economic research, especially those concerning the crucial role played by the rule of the law.


All in all, I think anyone interested in a broad discussion of long-term economic development will find the books by Landes and Bernstein interesting and challenging. However, if one has to choose between the two, I would recommend Bernstein.


maandag 1 november 2010

Review of Jared Diamond's "Collapse"

In his 2005 book “Collapse”, Jared Diamond provides interesting case studies of (more or less advanced) societies that collapsed under environmental pressures (that were at least to some extent of their own making). The historical case studies are followed by a discussion of some contemporary societies that face important environmental challenges.  In the three final chapters of the book, Diamond tries to distill some broad policy conclusions for our present world.
In my opinion, the best part of the book are the chapters on environmental archeology.  In these chapters, Diamond discusses first the collapses of the following societies: Easter  Island, the Greenland Norse, the Maya, the Anasazi Indians and the Polynesians of Pitcairn Island.  This analysis of collapses is complemented by an analysis of societies that have proved able to reverse unfavorable environmental trends (including the Pacific Island of Tikopa, central New Guinea, and Japan under Shogun rule). From a scientific point of view, it is fascinating to learn how archeologists derive their knowledge of illiterate past societies through the reconstruction of past environmental conditions. The book contains for instance a discussion of the “quantitative” analysis of factors predicting the survival of isolated societies that I would count amongst the better uses of statistics in social science.
The facts that are revealed are often baffling. It still remains beyond my understanding how illiterate  Polynesians developed navigational skills with a precision close to that of a GPS. Also, why have we never been taught at school that the Norse arrived in Greenland centuries before the Inuit, maintained for centuries a catholic society with close trading ties with Europe but eventually all starved to death with a sea full of food within the reach of their hands? And where else could you learn that Japan is actually a country that is mostly covered by forests, “thanks” to the preservation actions started by the Shogun?
The main interest of these chapters however lies in their clear demonstration of the central theme of the book:  there are indeed historical examples of once flourishing societies that have more or less completely disappeared due to environmental factors (mainly the deterioration of soil quality and the overexploitation of renewable resources).
After the discussion of past societies, Diamond moves on to a discussion of modern societies. He covers the recent genocide in Ruanda, the different environmental fates of the two states sharing the Island of Hispaniola (Haiti and the Dominican Republic), and finally the environmental challenges faced by China and Australia, respectively. These chapters certainly are chockfull of interesting facts. I was already aware of the work of my fellow countryman Platteau on the economics of the Rwandan genocide. I was still amazed to find out that, in some villages, the  killings were purely intra-ethnic. As a democrat fiercely opposed to any sort of authoritarianism, I almost dislike the idea that the forest in the Dominican Republic were saved by an odious strongmen, but I have to accept this fact. The eagerness with which the Australian government has provided incentives for non-sustainable environmental practices looks almost suicidal for an advanced society.
However, a grand unifying theme is missing in these chapters, and one is left with the impression that these case studies were chosen simply because these are the ones that Diamond happens to know, for instance because he lived or worked in these countries. What is also missing is any attempt to put figures in perspective: when one reads the chapters on China and Australia, for instance (these were written in 2004), one wonders why these countries have not already collapsed under their own mismanagement. Surely something is missing in the picture that Diamond has drawn.
Finally, Diamond attempts a general assessment of the 12 main environmental threats this world is facing, and how to face them. I think the list of threats is relatively non-controversial, but it is still worth repeating them over and over again. His discussion of the role played by big business is very balanced, and does not get caught in the trap of left-wing one-liners. However, his general analysis of contemporary policy issues remains very superficial – I suppose the main reason is that Diamond is simply not familiar enough with concrete high-level modern day policy issues. The arguments used to refute “one liners” of those who  dismiss the importance of environmental problems are mostly also of the “one liner” type (although I strongly agree with his refusal to just assume that scientific progress will eventually solve all our woes). The discrepancy with the depth of the analysis of past societies makes these chapters a disappointing reading, and one wonders whether the book would not have gained without the last three chapters. Alternatively, Diamond should have invested the same efforts in discussing with high level experts as he has done for the previous chapters. To take just one example from my own field of expertise: it is comforting to see that he has talked with Economics Nobel Prize winner Elinor Ostrom, but Diamond’s grasp of economics remains very superficial (an undergraduate would fail his exams with sentences such as “the price in those markets is not determined by production costs but by demand and supply”).  
What is then the overall evaluation of this book?
I suppose that the biggest problem for “Collapse” is simply that it was published after Diamond’s ultimate masterpiece “Guns, germs and steel”, which I regard definitely as one of the most essential books I have read in my life: reading this book was one of those rare occasions in my life where all my existing views on the world have been shaken upside down. In this perspective, “Collapse” is indeed not as revolutionary.
However, I still think that “Collapse” is required reading. Too many people still simply assume that environmental issues are unrelated to our own material welfare. “Collapse” provides a thoroughly documented antidote to this type of thinking – and, in the process, contains a lot of fascinating (if not exactly enlivening) stories. I just fear that most people who will read this book are already convinced of all this.
Maybe just a last words on stile. Diamond is a scholar, and this shows in the book. Although I definitely prefer  a thoroughly documented and detailed argument, I think the level of detail in this book might put some people off – some of the details could have been moved to footnotes or a technical annex without affecting the strength of the analysis.